Here’s what we know: There’s a pair (father and son) of Russian scientists trying to resurrect (or rather, “rewild”) an Ice Age (aka Pleistocene era) biome (grassland) complete with (gene edited, lab-grown) woolly mammoths (derived from elephants). In Arctic Siberia (though, not at the one station there that Amazon Prime delivers to!).

Here’s what we don’t know: How many genes will it take? (with science doing the “sculpting” and nature doing the “polishing”)? How many doctors will it take to make? (that is, grow these 200-pound babies in an artificial womb)? What happens if these animals break? (given how social elephants are)? And so on…

In this episode of the a16z Podcast — recorded as part of our podcast on the road in Washington, D.C. — we (Sonal Chokshi and Hanne Winarsky) discuss all this and more with Ross Andersen, senior editor at The Atlantic who wrote “Welcome to Pleistocene Park“, a story that seems so improbably wild yet is so improbably true. And while we focus on the particulars of what it takes to make this seemingly Jurassic Park-like story true, this episode is more generally about what motivates seemingly crazy ideas — moving them from the lab to the field (quite literally in this case!) — often with the help of a little marketing, a big vision, and some narrative. And: time. Sometimes, a really, really, really long time…

Show Notes

• The problem of melting permafrost and climate change [1:06]
• Why scientists want to create a modern woolly mammoth [7:45], and the gene technology they may use to do it [15:09]
• The importance of grass in the development of early humans [17:51]
• How researchers market this project and others like it [25:02]

Transcript

Sonal: Hi, everyone. Welcome to the “a16z Podcast.” I’m Sonal. Today, Hanne and I are doing another one of our on-the-road shows from Washington, D.C. And today’s guest is Ross Andersen, senior editor for The Atlantic’s science, health, and technology coverage. And he wrote a story earlier this year, in the April issue, called “Welcome to Pleistocene Park,” which you don’t have to have read to follow this conversation. But here’s what you do need to know. A small group — a very small group, in fact — of Russian scientists in Arctic Siberia are trying to resurrect an Ice Age biome, complete with lab-grown woolly mammoths, through a scheme for rewilding grassland instead of forest. 

And while we focus on the particulars of all that, in this episode — in a hallway style riff beginning with the connection to climate change, and then moving to gene editing, to discussing the science of paleontology, and the sociocultural and economic aspects of radical geoengineering — this episode is really more broadly about what motivates seemingly crazy ideas, moving them from the lab to the field — quite literally, in this case — through marketing and narrative, which is where we end and begin the conversation.

The problem of melting permafrost

Ross: So, when I landed on the website and I see that these guys are trying to rewild all or a great part of Northern Siberia, and Alaska, and the Canadian Yukon with this Ice Age grassland biome — and that they wanna put woolly mammoths there — you know, I had the same reaction that everyone listening to this has, right, which is, like, “What?”

Hanne: “Jurassic Park. Is it real?”

Sonal: Yeah, “Jurassic Park,” totally. The Ice Age.

Hanne: “Is this a joke?”

Ross: “Who are these crazy people?” Yeah, yeah, yeah. Totally. Yet, I was excited to write the piece. And then the other thing about this project that was really compelling is that it’s not that these guys were only just romantic about bringing the Ice Age back to this huge stretch of the Earth. Their primary motivation for doing it is to act as a climate change mitigation strategy, which is to say that the Arctic is warming very fast, and under the surface in the Arctic is what’s called the permafrost. That’s ice that has been there for, in some cases, tens of thousands of years.

Sonal: And, in fact, very deep. I read in your article, like, up to, like, a mile deep in some places.

Ross: Yeah. That part of the world was so rich in grass and in large animals at that time. It’s got lots of, sort of, organic matter, which has lots of carbon in it, in fact, more than, like, the entire output of the United States right now.

Sonal: Let’s take a step back for a minute. First of all, what’s the connection between the permafrost and climate change? Like, how can a grassland steppe with some fluffy, furry animals stop climate change, bluntly?

Ross: Totally. Okay. So, most of that part of the world up in the Arctic is covered with tundra. You might think of it as the Arctic desert. Like, very little grows on it. It’s kind of, like, scrub. And what’s neat about grasslands is they actually keep the earth underneath them colder. First of all, they reflect away more sunlight than the darker, kind of, tree regions. You’re already hedging against the warming, right, by having grasslands out there. And in the winter…

Sonal: Shade.

Ross: Shade, boom.

Sonal: They like wearing white on a hot day.

Ross: And then in the winter, you have — the snow cover is, like, on the grass, is really thin, such that, like, the Arctic cold in the winter when it’s really dark, and it’s just the auroras up there, can really penetrate the ground deep, and keep the permafrost even more frozen.

Sonal: Well, you actually use the language that it’s, like, locked in some thermodynamic vault…

Ross: I did. I didn’t want to roll that out.

Sonal: …which I think is, like, the best way of, “I’m rolling it out for you.”

Ross: Yeah, thank you. Thank you, thank you.

Sonal: That’s such a really good way of describing it. And so, what happens when those — because isn’t that a good thing to have all that organic matter? I mean, that creates oil, it creates, you know, this rich ecosystem that fertilizes our grass. I mean, what’s wrong with that melting?

Ross: What’s wrong with that melting is that bacteria will get at it, and through the process, they will decompose it and release carbon as part of that process.

Hanne: And it’s melting, not just because of the warming, but isn’t there an ecological contribution to the grass going away?

Ross: What’s so important about the animals being there is that the animals help to maintain that grassland ecosystem. And the woolly mammoth is involved because woolly mammoths, like many of their elephant cousins, are really good at knocking down trees. In fact, they were excited about it. Like, it was like one of their favorite things to do.

Sonal: But we could just, like, knock down trees ourselves. Like, why do we need the animals to do this? Why don’t we just raze a shit ton of forest trees, you know, pine trees, whatever, and just create grassland? Why do we need these woolly mammoths to be there?

Ross: In the absence of mammoths, they’ve just had, like, a huge Russian military transporter out on the plains that they’re literally just, like, slamming into trees with to knock them down.

Hanne: They’re weeding with their, like, military vehicles.

Ross: As you’d imagine, throwing out, like, a fleet of tractors that can knock down the trees of the taiga and, like, the entire Arctic region, would be a pretty carbon-intensive activity.

Sonal: So it’s, like, actually, making the problem worse and trying to solve it versus…

Ross: Like, we need all the world’s oil.

Hanne: But wait, can I back up and ask a question? Like, what I was trying to get — why do trees grow up that now are a problem? You know what I mean? That we need to, like, if you — why is the problem starting?

Ross: Well, one theory is that trees took over. First of all, you had the end of the Ice Age, which created a whole bunch of warming, right? And so the trees, kind of, that helped them spring up out there, but also, in the absence of large herbivores, like the woolly mammoth, it’s easier for trees to, like, spring up. And so, lots of people think that, when these animals went extinct — and we can talk about how they went extinct and some of the really interesting debates around that — that paved the way for these forests.

Sonal: Actually, one of the things that struck me and I feel like I reference “Sapiens” a lot on this podcast — the thing that just blew my mind is, Yuval Harari paints this picture of how humans are basically the worst predators in Earth’s history, and we’re so tiny relative to this huge megafauna, both on land and in water — from, like, huge woolly mammoths to whales in the ocean — and that everywhere humans move, you can immediately see a decline drastically in the number of large mammals that would walk the Earth.

Hanne: Yeah. It was so interesting when you talk about this birth period, and also, like, in quick succession, right, just ravaging…

Ross: That’s the word, yeah.

Sonal: It is. It is the right word.

Hanne: …absolutely, like, the wildlife and, you know.

Ross: Yeah. Yeah, it’s really interesting. A lot of that science has crystallized as our timelines for where humans have shown up in the world have gotten more refined. So, from very early on in paleontology, the consensus was, everyone noticed these large animals had died out at the end of the Ice Age, and they thought, “Well, the end of the Ice Age was just a period of warming, and these animals didn’t adapt.” And then as time went on, it’s like, well, glaciations — like, the Ice Age, was not 3 million years of glacial cold. It was, like, 10,000-year bursts of glacial cold and then interglacials, as they’re called, where things would warm again. And these animals had weathered, like, 30 of those.

Hanne: These tsunamis. You called them, like, ice tsunamis. Yeah.

Ross: Yes. And they’d been fine coming out of the other side of them. So why this one, did all of these megafaunas die? Humans show up, everything dies.

Hanne: Well, not everything. A specific kind of thing, right?

Ross: Yes.

Sonal: Grassland played a big role, because you no longer had this advantage where big animals could hide behind trees, or rocks, or big things. And so, humans had to adapt by becoming very good at hunting, like, shooting with spears or fire, in order to attack these animals and essentially learn coordination as they got out of trees.

Ross: Well, one interesting question around that, that I didn’t get to in the early 14,000 word draft — there was more on this — is that it’s always a mystery why Africa has kept a lot of its megafauna.

Sonal: Why? Why is that?

Ross: Yeah. So, one of the running hypotheses is that the megafauna of other continents were what’s called naive prey, because, like, humans show up, harmless little thing — whereas, in Africa, the megafauna there had grown up alongside us evolutionarily.

Sonal: Right, co-evolved sort of.

Ross: They saw, like, “Oh, these guys appear to be quite dangerous.”

Creating a modern woolly mammoth

Sonal: Yeah. So, back to your piece in The Atlantic — reading it, your use of this form of narrative journalism that gets you attached to the characters, the human characters — and I was actually more fascinated by the scientific characters. And that is, the grass, the mammoths, the role of, you know, elephants. And so let’s break each of those down and talk about, you know, what they are and how they connect to this.

Ross: Oh, interesting. That’s interesting. I never even thought about it that way. I mean, I obviously thought about the human characters. Sergey and Nikita, it’s these two guys, you know, this father and son in the Siberian Arctic, in the very far east, and they’re trying to rewild that part of the world into an Ice Age grassland with extinct woolly mammoths to fight climate change.

Sonal: Okay. So let’s break down the first character that I think is the most obvious and important one — is this idea of manufacturing mammoths, and specifically the woolly mammoth. Talk to us about that.

Ross: First of all, one of the other things that really attracted me to this story was — the woolly mammoth, when you talk about animals that are no longer with us, short of the dinosaurs, the woolly mammoth is the most romantic one, right?

Hanne: That’s so tied to this idea of, like, the first man, kind of. Like, it’s, like, how we have this idea of codependence on this, you know, animal from a very early age, even in popular culture.

Ross: We can tell, yeah, that’s right.

Sonal: Even if you think of things like “Clan of the Cave Bear.”

Ross: Yes, exactly.

Hanne: Yeah, exactly.

Sonal: That series by Jean Auel.

Ross: Although that’s, like, a huge Ice Age mythology, right, like “The Clan of the Cave Bear,” exactly. Yeah, they show up in cave paintings, right? They’re so resonant with, like, this kind of emergence of humans.

Sonal: And the woolly mammoth, just to give us a visual picture, basically is a big fat Snuffleupagus with tusks.

Ross: You got it. It’s a furry elephant. And that’s actually quite central to this piece, because if you do want to manufacture woolly mammoths, which is a crazy phrase, you want to do it the same way nature did, which is, you know, elephants were in Asia, in the temperate parts of Asia, before they were up north in the Arctic. As they slowly moved, nature modified their genomes through natural selection so that they had longer fur, and smaller ears, and you know, an extra layer of fat so they could stay warm in the Arctic. It’s nothing more complicated than that.

Sonal: Except, in this case, it’s happening through CRISPR, and scientists are manually modifying the genes to essentially edit in these characteristics from elephants, which are in the same family.

Ross: That’s right, yeah. They want to take, you know, basically, an Asian elephant genome and just make really a small number of tweaks. The guy who is really at the forefront of this is George Church, who is a geneticist at Harvard, and kind of has his hands on any number of, sort of, eccentric schemes like this. But I mean, when I first heard about this, I thought, you know, “Really?” But then I started talking to people in the field, and they were like, “Look, he’s out there.” Not “he’s out there” like he’s crazy. George is really at the forefront of this. I mean, like, he has the right approach, which is to make, like, again, as few tweaks to this genome as possible — just so you get these basic features — and then let nature do the rest. Get to, you know, 5, 10 generations of these and that will refine it.

Hanne: I love when you say you realized the idea isn’t why — how crazy this is to do it as actually, like, “Well, it’s actually not that crazy.” The reason is, like, why wouldn’t it work, right?

Sonal: I love it too.

Ross: Right. Right, right, right.

Hanne: Do we know exactly what the woolly mammoth was? Do we know exactly what we’re aiming for, or are we guessing?

Ross: We have used several DNA fragments to sequence, like, the entire woolly mammoth genome. However, we are not trying to make — so, I’m speaking out of two corners of my mouth here because I’m saying we’re gonna manufacture mammoths, but what we are actually gonna do is manufacture a furry, fatty Asian elephant. Like, we are not aiming…

Hanne: A mammoth look-alike.

Ross: …for the original genome, for the exact genome of the original mammoth. We’re just looking to remodify Asian elephants.

Hanne: An Asian elephant with the characteristics of a woolly mammoth in certain key areas.

Sonal: Just to give some textural feel, you described that Church and his group are adding cold-resistant hemoglobin, a full-body layer of insulating fat, they’re shrinking the ears.

Hanne: Why are they shrinking the ears?

Sonal: Why are they shrinking the ears?

Ross: Good question. Well, imagine, you know, in the Arctic, you get, you know, 70 below during the winters.

Hanne: Frostbite.

Ross: The African elephant has these huge ears, and those are not needed in the Arctic. And then you said cold-resistant hemoglobin. I wanted to call it antifreeze blood.

Sonal: Like a new version of True Blood, like, “Drink this antifreeze blood.”

Ross: That’s right. And they wouldn’t let me get away with it.

Sonal: Hanne, you asked an amazing question about, you know, is it actually doing it from truth or not, but is there a truth? Because you also pointed out, we have this dead DNA problem. Like, you think of DNA as this thing that lives on for ages and eons, but in fact, those DNAs decompose and [are] not really available even to draw from.

Ross: That’s right. One reason that we’re looking to just modify Asian elephant genomes instead of, like, doing the Jurassic Park style, like, “Oh, we found it in the amber,” is that, look, even after a few thousand years, DNA gets really decayed, and by cosmic rays, and by microbes, and by any number of — nature is really — you know, the universe is a really harsh place. Oh, yeah.

Hanne: So it sounds like you’re sort of saying like it almost doesn’t matter. As long as an elephant can live there, it’s okay. But once we start giving them these different — and we’re introducing a new animal into this very complicated ecosystem environment, like, does it maybe matter that they’re not exactly the woolly mammoth?

Ross: My view is that it’s worth what will probably be some considerable suffering on the part of the first few, if not more, generations of these mammoths. And like, I’m alive to that, and I actually try to talk about, in particular, the social suffering. I mean, elephants are really social animals. They hang out in matriarchal herds. Their grandmothers are around, like, teaching them, you know, all of these behaviors. They grieve their dead. They have, like, a really rich communication with, like, you know, these little rumbling sounds, many of which are inaudible to the human ear. They’re some of the most social animals on the planet.

Hanne: How do we even know, you know, these unformed, untaught — these poor difficult new things, dropped into this new landscape?

Sonal: And by the way, all at the same age.

Hanne: How do we even know they would know to do what we want them to do? I mean.

Ross: I suspect that — have you ever seen in the zoo, they have the guy who gets in the mama tiger suit, you know?

Hanne: Yeah, yes.

Ross: I think there might be something like that happening early on. I mean, I can’t imagine.

Sonal: We think of these as purely biological things, and we forget that there’s a transmission of culture that has to happen as part of it. And in fact, even the language you use in the piece — I actually was a little taken aback. You have this language, and it’s beautiful. As an editor, I’m like, “Oh, gorgeous diction.” You talk about how we sculpt them to survive the winter but let natural selection do the polishing. It felt more like playing God, just bluntly. Like, it’s like creating a Galatea clay. I don’t know, Pygmalion and Galatea, like, you know, whatever.

Hanne: Well, I think, yeah, it reminded me — it feels to me like making a golem, kind of, right, because we’re shaping the outside, and we’re not doing any of the — and when you’re describing all the complexity of, like, you know, the biology of the gut to eat the tundra and, like, all of that complicated, you know. And then we’re just, like, shaping this stuff out, the exteriors and then plopping them down.

Ross: Well, the other thing — I mean, I think this really gets to one of the philosophical tensions that I wanted to confront, to your point about playing God. Another thing that’s like playing God is removing 95% of the megafauna from the surface of the Earth.

Hanne: That’s right.

Sonal: Yes.

Ross: We have natural human biases around things like gene editing that, like, get us all prickled and, like, “Oh, we’re playing God.” But in fact…

Hanne: But in fact, we’ve been editing everything.

Ross: …we have this tremendous effect on the Earth.

Sonal: So let’s break down some more of the science on playing God. So we talked about CRISPR, the gene-editing tool, and let’s talk about the genes. So we described some of the characteristics and features that we wanna add, but by my count, there are 95 genes to do the job. 15 that were completed, 30 that are being tweaked, and he says George Church was guessing that we need maybe 50 more.

Ross: He actually was saying, you know, a total of 50. Beth Shapiro, who I regard as sort of the world expert on this stuff — she was, like, you know, “Not so fast. You have to see what those changes do to the rest of the body, and how they interact with each other.” So like, sure, maybe 50, but it’s too soon to say.

Sonal: Right. Well, the other thing that I found very fascinating, especially in the tales of that recent news about the artificial womb in an animal being able to be incubated, is that you essentially grow these mammoths in an artificial womb. So what’s that process?

Ross: Yeah. And I’m glad you brought that up, because actually, that is the most science-fictional aspect of this whole thing.

Hanne: That’s the biggest stumbling block.

Ross: That’s the biggest leap, yeah.

Sonal: Interesting.

Ross: Gene editing, you know, it’s a known technology, it’s a matter of trial and error. It’s like, “Let’s, you know, keep spitting out embryos with, like, different changes, and eventually, we’ll get there.” Growing an embryo, especially this is the animal with the longest gestation period.

Sonal: Which is what, 22 months or something?

Ross: Two years. Yeah, yeah, almost two years. That’s right. And it’s, you know, 200 pounds at the end of it, and you’re gonna do all that, like, really complex fine-tuning, maternal fine-tuning, like, hormonal work in this huge closet-sized tank. Like, that’s more than 10 years away. George Church thinks that you can make a mammoth, like, genetically, within five years. And he said to me, “Just like there’s uncertainties on the pessimistic side,” like, “Oh, actually it’ll take 20,” he’s like, “It could take shorter, you know.” But growing an actual elephant, a furry elephant, in a tank — we’re not there yet, technologically. That is a thing that, it’s like, no one is working on even as hard as these guys are with the gene itself.

Sonal: I hear you when you say it’s the most science fiction of this whole piece, but when I heard the recent news about the artificial womb, it actually gives me great hope, because you think about all, you know, the collateral good things that come out of this kinda science and work. Like, will we be able to have true artificial wombs for human beings as a result of this work, or other things that we can essentially let women have kids? Like, that’s just a beautiful idea to me, that we can actually manipulate that on some level.

Ross: It’s completely lovely. But just to put that in context and to illuminate the challenge, if you were to make it analogous to human beings, women have, like, a 40-week gestational period. These are, like, preemie lambs. Like, they were born at, like, the equivalent of 22 human weeks. And they stuck into these artificial wombs, and they were able to go to term.

Why grass was essential to early humans

Sonal: Let’s go back to breaking down the characters one by one. We need to talk about grass. You mentioned that Ice Age is actually really a grass age and, by the way, that the formal name of Ice Age is the Pleistocene Age. I actually didn’t connect [that] all three of those things are actually the same thing.

Hanne: Is it exactly what we think of as the Ice Age?

Ross: It is the Ice Age, it is. So it’s 3 million years, and the really interesting thing about it is, it’s kind of, like, the nursery period for human beings. Like, this is where we sort of, you know, discovered fire, learned to harness fire, developed language, developed advanced tool use, and then, all of a sudden, we kinda pop up, history starts, what, like, you know…

Sonal: Accelerates out of there.

Ross: …5,000, 6,000 years ago where you have, kind of, genuine writing. But all those behaviors really incubated in the Ice Age, so I’ve always been kind of fascinated with that period.

Sonal: And timescale-wise, that ended 12,000 years ago.

Ross: Yes.

Hanne: Can I just have a moment of fan mail here?

Ross: Oh, God, please.

Hanne: I love when you looked at one blade as, like, this little soldier fighting this grand army, you know, of the wages of, like, the planet.

Ross: I went down deep in this ice cave with Nikita, the son in the story, like, walking around in a geode. Almost every surface is covered, you know, with sparkling ice. And we get to, like, the bottom of this little chamber, and you know, he sort of, like, scratches at the ice wall, and he pulls out this, you know, pale dead blade of grass from the Ice Age from 30,000 years ago. And at the time, I will confess to you guys a little sort of writerly craft. I watched…

Sonal: I thought you were gonna confess fear, because I was thinking about that whole thing, and I was, like, “Holy fuck, claustrophobia, cave, freak out, cold.”

Ross: Totally, totally. Fair, fair. So going into the piece, I really thought that the kind of reigning mythology that people will have in their mind reading this article is Jurassic Park. And so, how can I, kind of, subvert that, right? When they’re kind of explaining how they do the resurrection of these dinosaurs, there’s a moment where they’re in a cave, and they hold up to the light this amber, and there’s an ancient mosquito trapped in it. And I thought, like, “Is there a way I can get an image like that?” And so, then, at the bottom, when he pulls out this piece of grass, I was like…

Sonal: Here it is.

Ross: “That’s my zip line into the deep past.” I’ll have to admit, I had always been much more romantic about forests than grass going into this piece. Sergey was talking about grass and its importance in the rise of humans, in particular. That really captured my imagination, and was an idea that I felt like was not out there in the world.

Sonal: And what is that? What is the connection between grass and humans?

Ross: Well, grass is, like, kind of the newest big plant-based biome on the planet. Like, forests have been around for, you know, 300, 400 million years, and grass is, like, less — well, big grasslands are less than, you know, 60, 70 million years old. And they’re really neat, like, grow really fast. They just, like, erupt out of the earth, and they make food very easily for animals. And they’re not — a lot of them are not afraid of being eaten. They love to be eaten. So you have trees, you know, well, like, or other plants, will invest all this energy into thorns and into poisons because they’re, like, “Get away from me.”

Hanne: “Back off.”

Sonal: To repel people from eating them.

Ross: “I don’t want you to eat me. Let me do my thing. I wanna grow.” And grass is, like, “Eat me. Eat me. Eat me. Eat me.”

Sonal: They’re sweet. They’re like, “Yeah.”

Ross: “And just poop me back out, so then I can grow even more, and you can eat me again, and you just go, go, go.”

Sonal: Into this feedback loop. You have this line actually — had so much packed into it. By allowing themselves to be eaten, they partner with their own grazers to enhance their ecosystem’s nutrient flows.

Ross: Yes. The animals poop them out, and they poop — you know, the great thing about poop, while we’re, you know, talking about things that we didn’t know were so great, like grass, is that it’s really sort of warm and kinda seeps into the earth very quickly, and it’s been processed by microbes. It’s like, kinda, you know, juicy.

Hanne: It’s ready to go. Natural fertilizer.

Ross: Yeah. It’s just fertilizer, right? We know, right, so what do we use for fertilizer? And so, it makes these grasslands just, like, cycle, cycle, cycle really quickly.

Sonal: I agree, this idea of the grass is so counterintuitive, and I first came across it in “Sapiens,” and one of the things he says is that humans tamed — created humanity, because it allowed us to use wheat to, like, drive our lives, and there’s all these different forms of grass that exist now. You’re describing rice, wheat, corn, sugarcane.

Hanne: I thought it was really interesting how, like, this is a portrait of all these, you know, cutting edge, sort of, science and tech discoveries and capabilities. And we’re using it to, like, reach deep into our, like, no longer accessible past. Like, you described this moment of solastalgia, right, like, this yearning for what once was. That’s kinda part of the human condition.

Sonal: And by the way, solastalgia, as in an existential grief for a vanished landscape — because that was the first time I ever heard that word.

Hanne: Yeah.

Ross: Yeah.

Sonal: I didn’t know what the hell that was.

Hanne: No. Me, too. Yeah.

Ross: Yeah. It’s a very minor philosophy. Yeah.

Hanne: Good. I was hoping you would define it. Yeah.

Ross: Yeah. So I’m really drawn to stories that show humans interacting on long timescales, which is a thing that I think we’re doing more and more now.

Sonal: By long timescales, you mean like cliodynamics or just anything that’s, like, the arc of history? What is that?

Ross: Yeah. I mean, like, when we think about what it’s going to mean to be human beings now and in the future, that we’re taking into that context 10, 20, 30, 40 millions of years into the past, and perhaps 10, 20, 30,000 years into the future. And this is, I should, again, give a shout-out to Stewart Brand, who obviously has had many fertile thoughts along this path, but…

Sonal: Stewart Brand who is the father of the “Whole Earth Catalog” and now runs The Long Now Foundation.

Ross: Yeah. But this idea of looking at our existence in a way that really zooms out from our current moment, which is certainly a relief in this particular historical moment we find ourselves in.

Sonal: There’s this interesting juxtaposition between past and present that’s so fascinating, both mechanically and then historically. But even down to some other random details, like, you mentioned — the first most popular Arctic station besides this one is the one in Alaska, and that’s one place that Amazon Prime delivers to.

Hanne: I know.

Ross: I know.

Hanne: I totally was struck by that too.

Sonal: I was just, like, “What the?” That was, like, “Wow.”

Hanne: It was unbelievable.

Ross: Didn’t that sound awesome?

Hanne: Yeah.

Sonal: That is so awesome. And it’s so funny because the other Arctic station is, like, “Okay, we don’t have Amazon Prime, but we have alcohol. Lots of vodka.”

Hanne: Like a little competitive, like.

Ross: They really go all-in on it too. The town that’s close to Pleistocene Park is, like, a really depressed mining town, and so I was wondering, like, “You must have poachers.” And he said, “Well, no.” You know, they hunt in all the forest around it, but they don’t hunt in the park. And I was, like, “Well, why not?” And he said, like, you know, personal relationships. And then he says to me, like, you know, “When the leader of the local mafia died, you know, I gave the opening remarks to his funeral.”

Sonal: I mean, it is an interesting thing about science meeting society. Like, when you have science not in a lab and playing out in the physical environment, you are gonna bump into things like cultural realities, poachers. One of my favorite things I’ve ever done in my life was go to this Jurassic Park of India. It was just a few years ago that I went. It’s called Balasinor. And it’s the world’s most ancient enclave of dinosaur eggs.

Hanne: Whoa.

Ross: Whoa.

Sonal: Yeah. And I’d found it by accident, because I was doing, like, some local research. And I rented a special truck. It took us forever to get there, even though it’s so close, because it’s on these down, windy roads. And the thing that was so amazing is you see these dinosaur eggs fossilized in the rock, but all the dinosaur pieces — the whole way that Balasinor was found is because some local women in huts nearby were using it for plates and bowls.

Hanne: Oh, my gosh.

Ross: Oh, my gosh.

Sonal: And they had no idea of the value. And they actually then put it on the market, some scientist came across it, and then all these scientists descended. But you have the government, you have the locals, you have the scientists, and you have all these characters. And one thing that did strike me in your piece is — that was kinda left unanswered is — who’s paying for all this?

Marketing long-term study projects

Ross: They’ve got NSF funding, and funding from the Russian government at the moment, and they do that partly because if you wanna study the permafrost or the Arctic, in general, you need to have these various outposts. And so it’s worth their money to do that. The more interesting question even than the funding, to me, which you’re kind of getting to when you’re talking about this lovely story about the dinosaur eggs in India, was that, for this to expand. Like, Yellowstone right now, which is a thing that everyone loves, right, like, you can’t get people to say bad things about Yellowstone. People universally acknowledge it as being an amazing thing in the world. But, like, its expansion impinges on real people’s lives, you know, because all of a sudden, big predators are showing up in their backyard, etc. And so, for something like Pleistocene Park to be successful, it’s going to have to interact with and make peace with the human world on, like, quite a grand scale, if they are gonna do all of Northern Siberia, and Alaska, and Yukon, etc., etc. And that as being representative of the larger tension we have of trying to figure out how we coexist with wild animals and with the wild, in general.

Sonal: There’s a socioeconomic component too, because you think of these towns that don’t have a lot of money to survive. They don’t have a lot of economic opportunity. Why wouldn’t you wanna sell, like, ivory, you know, from these tusks and make some money for yourself to survive or support your family?

Hanne: Or dinosaur egg china.

Sonal: Right. And so it’s really striking when you do think about this question of who funds it, because there’s a lot of science and money that goes into this. And there’s just a lot of tradeoffs that people have to make. And anyway, another open question is, like, this project is so radical in scheme and scope that — is anyone else doing anything this ambitious in the world anywhere?

Hanne: Well, you compared it to one other major climate project, right?

Ross: Oh, yeah. They’re geoengineering projects or proposals. Also, the American Prairie Reserve is another large grassland rewilding project. It doesn’t have, sort of, sexy extinct creatures to sell it or, like, a major climate change mitigation strategy to sell it, but it’s really interesting, and it’s, like, part of Montana.

Hanne: Tell us, I would love to hear the story behind the stories.

Ross: Funny story is, going up there — this is, like, a protected area, and so you have to get official Russian permission, not just like a regular visa, just to actually go to this region. So we get there, and I had a really good friend of mine, Grant Slater, who’s an amazingly talented documentary filmmaker. We’d kinda worked together. I knew that he would have this, sort of, deep time sensibility alongside me. And so I was really excited to see what he would do with it.

Sonal: Visually.

Ross: And there’s also a really interesting creative tension, being out with a filmmaker, because, like, he has things he needs to get, I had things I need to get. Anyway, I’m rambling.

Hanne: Yeah. It’s a different kind of storytelling.

Ross: But Grant’s paperwork, his, like, official permission had not come in on time, and so we had to, like, go get — we went and got questioned at the military base by, you know, these Russian soldiers who are, like, in full fatigues, pretty big dudes. And what was funny about it was, Grant had lost one of his suitcases in Moscow. He had to buy clothes, like, at the airport. And the shirt he was wearing during our interrogation was this shirt that said, in Russian, “Russia is a great power.”

Hanne: No.

Sonal: It’s like a scene out of a comedy movie.

Hanne: And didn’t they — they went, like…

Ross: I was devastated when he got caught. Devastated.

Hanne: They thought he was a spy, right? Like, they were, like, “You’re obviously…” and he’s wearing this t-shirt that says Russia is great.

Sonal: No, no. Even worse, they asked him if he was a spy.

Hanne: No, that’s right.

Sonal: Like, a spy is gonna say, “Yes, I’m a spy.” That’s just crazy. So just to close, I think the most striking thing about this piece, that this idea sounds so crazy at first. The thing that really struck me is that the region that you were in was once famous for beaming propaganda throughout the country of Russia, and at the same time, there’s an element of marketing that has to happen in this idea, for someone to convince other people, to drive people towards their vision, to get them to believe it.

Ross: I’m also captured by this question of how, you know, when you have these really esoteric science projects that are tied into questions of human meaning in all kinds of different ways, how you can present that.

Sonal: And sometimes cults of personality as well.

Ross: And cults of personalities, and how do you, kind of, make that — I mean, something that Elon Musk is really adept at, right, taking ideas out.

Sonal: I remember you did that Q&A with them at Aeon, a long time ago.

Ross: Yeah. Like, he’s really good at packaging crazy-sounding ideas and, like, getting lots of governments, investors, to throw lots of money into them, while managing to keep control of them. Part of that is the narrative, right? He does hook it into, like, larger questions and existential concerns in a way that I don’t think is just manipulative, I think he sincerely believes those things.

Hanne: And I also think a lot of it is, like, just saying, like, “This is happening now.” Like, sort of making us realize, like, “Actually, this is happening now.” You know, that’s a lot of turning it around to feel possible, basically.

Ross: Yeah. People are working on it. It’s a thing. You can go there. Yeah.

Hanne: Yeah. It’s a thing. Yeah.

Sonal: Well, also that it takes time, because one of the most telling anecdotes in your piece — because you know, there’s a whole debate we don’t have to go into in this podcast nor do we have time, about climate change deniers, climate change science, what’s legit, what’s not — a whole other conversation. But what I found fascinating was that science initially rejected Sergey’s paper about the dangers, you know, in the warming…

Ross: Of the permafrost. Yeah.

Sonal: Right. And in 2006, the journal then asked him…

Ross: Yeah.

Sonal: …he didn’t have to approach them again, to resubmit it. And it was published later that year.

Ross: Yeah.

Sonal: And that just goes to show you, there’s also a right time for some of this.

Ross: Right.

Sonal: Like, there’s a readiness that has to happen. Thank you for joining the “a16z Podcast.”

Ross: Well, thank you for having me on.

Hanne: Thank you.

• 

  Ross Andersen
• 

  Hanne Winarsky
• 

  Sonal Chokshi is the editor in chief as well as podcast network showrunner. Prior to joining a16z 2014 to build the editorial operation, Sonal was a senior editor at WIRED, and before that in content at Xerox PARC.

  Follow Twitter

This episode of the a16z Podcast takes us on a quick tour through the themes of economics/historian/journalist Marc Levinson‘s books — from An Extraordinary Time, on the end of the postwar boom and the return of the ordinary economy; to The Great A&P, on retail and the struggle for small business in America; all the way through to The Box, on how the shipping container made the world smaller and the world economy bigger.

In this hallway-style conversation, Levinson and we (with Sonal Chokshi and Hanne Winarsky) touch on everything from productivity growth & GDP to the “death of retail” — to finally connecting all the dots through logistics, transportation, infrastructure, and more. How are supply chains changing? How does all this, taken together, affect the way we work? And what can — or can’t — policymakers do about it? Perhaps, Levinson argues, a lot of the improvement to our living standards really comes out of “microeconomic improvements at the private sector level rather than as a matter of great policy”. But that’s a bitter pill to swallow for those seeking solace in easy answers from governments, whether at a national or city level. Maybe it’s just a matter of managing our expectations — or resetting our clock for when the new normal begins… and ends.

Show Notes

• Why the post-WWII era was a golden age for economic growth [0:48]
• The current state of the retail economy [8:43] and how the history of A&P relates to today [12:43]
• Discussion of shipping and supply chain issues [15:50], and issues that will face future generations [27:38]

Transcript

Sonal: Hi, everyone. Welcome to the “a16z Podcast,” I’m Sonal. Today’s episode with me and Hanne is another one of our podcasts from our recent road trip with Voices from the Ground in Washington, D.C., though this one actually takes us all around the world. Our guest is the economist, historian, and journalist who was last at “The Economist,” Marc Levinson, the author of the beloved book, “The Box,” which is about how the shipping container made the world smaller and [the] economy bigger. But this hallway-style conversation is actually a quick tour through all his books, starting from his most recent one, “An Extraordinary Time,” where we touch briefly on the topic of the golden age of productivity and beyond, to the topic of the death of retail in his book, “The Great A&P,” to finally wrapping up on logistics, transportation infrastructure, supply chains, and touching very briefly on the future of work and where government comes in policy-wise in all of this or doesn’t.

The golden age of economic growth

We’re so excited to have you. Welcome, Marc.

Marc: Thank you. Glad to be with you.

Sonal: So, those seem like really different topics. What’s the big idea that drives the thrust of your work — that kind of connects all the dots?

Marc: I’m really interested in the connections between economics and the world we live in. A lot of my work starts out at a microeconomic level, looking at particular companies, looking at particular industries, and tying the developments there to broader trends that really affect how we live, affect our standards of living. More recently, I’ve been focusing on some of the trends in productivity growth, because I believe that a lot of the improvement in our living standards really comes out of these, kind of, micro-improvements at the private sector level, rather than as a matter of great policy. And what that means, and this is a frustration for public officials, is that there are no easy government solutions. We’ve now been through generations in which politicians and the economists who advise them said that they had the cure for productivity growth.

Hanne: Right.

Marc: I argue in “An Extraordinary Time” that, actually, this was what was behind the political swing to the right in the late 1970s, early 1980s, when we got Margaret Thatcher and Ronald Reagan. Because the more social democratic types of governments before that hadn’t been able to restart productivity growth. And so, voters turned to people with other ideas, but the people with the more free-market ideas proved no more successful than the people with the more statist idea.

Hanne: What are we actually comparing to, as we’re thinking about these, like — “Well, this is not good enough?” What are we holding up as, you know, something that we’d prefer it to be?

Marc: The end of the post-war boom and the return of the ordinary economy. The story I’m telling is that the quarter-century after the war was an unusual period of very rapid economic growth. The period from 1948 to 1973 was probably the period of the fastest economic growth in the history of the world. GDP around the world grew at more than 5% a year.

Now, at 5% a year, something doubles in 14 years, quadruples in 28 years. So, even with some population growth, people’s incomes were growing very rapidly. People’s living standards were rising in a way that was visible to them. They were able to buy houses for the first time and cars for the first time, and send their kids to high school, and maybe even college, and we had all kinds of very rapid advances in living standards.

Hanne: What was that due to? What was the big driving force?

Marc: We had an unusual confluence of factors in the post-war period that people have really forgotten about now. One is that there was a great deal of underused capacity — underused resources in the economy. I’d like to remind people that at the end of World War II, we still had 3 million mules on farms in the United States.

Sonal: Wow. It’s such a technical, post-industrial revolution time. You don’t even realize it. It’s that…

Marc: You have millions of people, and not just in the United States — European peasants and Japanese farmers owned half an acre of land — who could be moved from very low-productivity jobs into very high-productivity jobs in the cities. And we had a lot of that in the ’50s and ’60s. So that was one big boost to productivity. We had very rapid increases in education levels, and we know that education is associated with productivity. In the United States, at the end of World War II, going to college was not common. It was — just a few percent of the population of 18-year-olds actually went on to college, and the average education level was around eighth or ninth grade. So, in very few years, governments spent a lot of money building a more educated workforce and it paid off.

Sonal: The government was the one that seeded that, or was that just a shift in the fact that adolescents existed and childhood changed?

Marc: No. This was heavy expenditures, building universities all over the place. Take a look at how many universities in the United States started after World War II, okay? That’s when a lot of government money started — it was no longer an elite thing to go to university.

Hanne: And how about women entering the workplace?

Marc: Well, women entered the workplace. The other thing I think that was really consequential in this quarter-century I’m describing was that we had the growth of motorways — the interstate highway system in the United States.

Hanne: So, public infrastructure, like, the transportation infrastructure?

Marc: Public infrastructure, and think about what that does. If you are a manufacturer or a retailer, that lets you sell over a wider area, lets you operate your facilities more efficiently. You don’t need a warehouse in every town. You can have one that’ll serve a large area. If you’re an employer or a worker, it’s changed the size of your labor market. I mean, in Silicon Valley, San Jose and San Francisco are now part of the same labor market, right? That wasn’t the case after World War II. These were very different cities and they were a considerable drive apart.

And so, that creates a better fit between people and jobs, and then leads to higher productivity. They can’t be repeated. Once you’ve moved those sharecroppers to the cities to take jobs in industry using heavy machinery, they’ve moved — and you don’t have those underused resources again. There were countries around the world that literally went 25 years from 1948 to 1973 without a single year of recession. We had countries that had less than 1% unemployment back then.

Hanne: So, how did this burst of productivity, this golden age, actually come to an end?

Marc: Well, in 1973, we really saw a trend change. That was the year of the great oil crisis that some people may remember. What we have moved into, since 1973, is really an environment in which economic growth has been slower. The improvement in living standards has been slower, the unemployment rate, in most countries, has been permanently higher. We have not been able to recapture the very unique good times that we had in this golden age, and I think that we’re not going to be able to. What we’re experiencing more recently is actually normal. This is the way most economies work most of the time.

Hanne: And how it worked before this golden era?

Marc: The golden age was actually the exceptional time. It’s not normal that economies grow at a rapid pace. It’s not normal that incomes double, or triple, or quadruple in a matter of just a few years, and I don’t think we should expect that to recur.

Sonal: I view this as analogous to child development and how a human body, an adult develops. Because there’s a rapid development that happens in the birth of a child, and then another big rapid shoot that happens in adolescence, and then you continue to grow, but it’s a little slower. And, in fact, thinking about the natural conclusion of your argument — is that that growth has now shifted to other countries like India, China, where they are now experiencing the kind of boom that you were describing that happened pre-1973. 

Marc: That’s a great analogy. Take Japan which in the 1960s or early ’70s was growing at 7% or 8% a year, and then it downshifted, and then it downshifted some more. More recently, China went through a period where it was growing at 10% a year.

Sonal: By the way, in China’s case, we do have to take the numbers with a big grain of salt.

Marc: Even so, people were extrapolating and saying, you know, when China grows at 10% a year for the next century, its economy is gonna be twice as large as the rest of the world put together.

Sonal: Right.

Marc: But China’s not going to grow at 10% a year for the next half-century. It’s becoming much more like a normal, mature economy in which the growth rate is a couple of percent a year, and that’s all they’re going to be able to expect.

The current retail economy

Hanne: But it doesn’t mean necessarily that, like, in adolescence or growing as a human being, you only get one burst. These things can come in waves, there can be other, kind of, confluences to these factors. When you were talking about the Japanese farmers with their mules, and the, sort of, move towards the way automation and the move to cities increased productivity, are there any inklings that you’re starting to see of possibilities, like, with the automation we’re starting to see happen today, and maybe even with autonomous cars, new infrastructure might, you know, for city infrastructure — are there things that give you any sense of maybe a new era might be coming at some point?

Sonal: Or just even a way to juice the body on steroids? Like, just inject some more steroids into this economy?

Marc: You’re asking great questions here, and the answer is maybe. I think that these are things we really can’t predict. If you look at past episodes of fast productivity growth, in general, they weren’t predicted very well. For example, we had a spurt of productivity growth which translated into faster income growth in the late ’90s and the first years of the 2000s. This is the famous internet boom, you may remember.

Sonal: Oh, we remember it.

Marc: In 1992, no one predicted this. What happened was that there had been investments in infrastructure, there had been developments in technology decades earlier — and, finally, during this period of time, they all came together.

Hanne: But I think a lot of people would argue that we’re in a moment like that again now.

Marc: Well, I think that’s a question which we can’t answer. So, if you take a look at a technology, will it actually revolutionize the way certain industries work? I don’t exclude the possibility, but you have to keep in mind that there are also a lot of complications. You’re seeing this right now as we go through this rather brutal shakeout in retailing. Yes, everybody knows that you can order goods on the internet. That’s not news these days, okay? But the reality is that for a lot of retailers, there’s a problem here, because they’re maintaining an internet business, they’re also maintaining a retail store business, because some customers want that. So, in some cases, their costs have gone up. They have not become online retailers — they have become bricks-and-mortar/online retailers.

Sonal: Right. And they’re <inaudible> for the online sometimes.

Marc: They’ve got multiple channels that they’re having to service, and that’s actually made their operations less efficient in a certain way.

Sonal: I would actually say there’s a flip side to this, though, again — which I think is really fascinating. Because when you think about the internet economy, birth of Amazon — which is, let’s face it, the behemoth in everything, the everything store, the everything everything. And they recently, as we know, started doing physical brick-and-mortar bookstores. The difference is that they started online and they went into physical, using data to help stock and think differently about how to create their store in an internet-native way in the physical world. So, I also wonder if while the death of retail might be on the horizon, if after that there might be an entire new post-boom — a new boom around retail that’s completely reshaped by new technologies. We don’t know.

Marc: That’s entirely possible. But just to give you something to think about, Amazon’s problem in terms of getting its books into its physical stores is entirely different from its problem getting its books ordered online to you, the customer, okay?

Sonal: Yes.

Marc: So, now it needs a different kind of logistical system. It needs to figure out how to distribute to retail stores like the ones it, apparently, is building. That’s going to have a lot of costs attached. It may have some inefficiencies attached, at least while they’re developing it. So, my point is to say that the path of — you know, sometimes people who are involved in the tech industry, kind of, get very romantic about how quickly these great technologies are getting absorbed. But, in reality, life is messy, and some of these technologies take a while to be used efficiently, and some of them will never be used efficiently.

Sonal: That’s right. There are more failures than there are successes. There’s no question about that.

Hanne: So, does it remind you at all of the, sort of, death of the supermarket that you talked about in your book, “The Great A&P?”

Marc: In “The Great A&P,” I was writing the history of what was, for about 50 years, the largest retailer in the world. People forget this now, but The Great Atlantic & Pacific…

Hanne: Is that what A&P stood for?

Marc: Yeah.

Hanne: I had one in my town. I didn’t even know that.

Marc: Yes, it was The Great Atlantic & Pacific. It was so named in 1869 for the transcontinental railroad.

Hanne: Wow.

Marc: And at one point, it had more than 16,000 stores in the United States, so it was a behemoth. It was the Walmart of its day. But one of the things that kept it so vibrant is that it remade itself continually, because shopping trends change, consumer expectations change.

Hanne: From what to what?

Marc: It started out as a seller of coffee and tea and spices. It made itself into a small grocery chain. And then, in 1912, it developed the idea of having an economy grocery chain, which is to say, it would have a very bare-bones store and sell products much cheaper than the competition. And that’s what drove its growth in a small period of years. It integrated vertically, so it made its own chocolate, its own macaroni, canned its own salmon — and then had a huge network of manufacturing plants, and, again, we’re in the 1920s here. And then it started building supermarkets. It was not the innovator in any of these things. A&P did not develop the idea of supermarkets, but once it saw how supermarkets would work and how they would fit with its business, it started building supermarkets all over the place, and by the end of the 1930s, was the biggest supermarket operator in the country.

Hanne: So, what ended up being its downfall?

Marc: The company stopped innovating. The company stopped remaking itself. It was big, it was fat, it was happy. The two brothers who had controlled it for decades both died in the 1950s, and it was then run by people who had been with the company for decades, and whose idea was to preserve it rather than to keep it changing.

Sonal: You know, the beat that keeps coming up in this is that, basically, the tension between this idea that you can innovate but then you get too good at what you do, too comfortable, too complacent. Okay, so what’s the big, then, lesson or takeaway, you know, from your work on “The Great A&P” to this narrative around the death of retail today?

Marc: Retailing is full of dead bodies. People like to talk about how unfair competition is sometimes, because the big companies have more power than the little ones. But when you are a big retailer, you can’t change so easily. Okay, if you own one store and you think you need to do something else, you go in there with a hammer and some plywood and you can do it. If you own a thousand stores, you’re stuck. Okay, you’ve got your locations, you’ve got your product line, you’ve got your brand name, and you can’t change easily. It’s a really difficult situation, and so a lot of stores end up dead.

Sonal: It’s the innovator’s dilemma, classic case.

Marc: It really is.

Shipping and supply chains

Sonal: So, another theme that’s come up and that connects all the dots with this entire conversation — that this post-boom world was one of the drivers — was this, like, rapid development of infrastructure. Amazon exists because of logistics and infrastructure. Like, innovations and being able to ship things and deliver things fast. You know, we talk about the supermarket and the growth of suburbs around railroads and transportation. Transportation on logistics and infrastructure is, like, the thread that connects and drives all economies. So, let’s talk about “The Box,” which is all about logistics and infrastructure in the form of container shipping.

Hanne: You chose one very specific thing in “The Box” to talk about. It had this massive effect on the global economy. Give us a little bit of sense of what that story was like. I remember Tim describing, when he pitched your book, this incredible scene of just the giant mountains of peanuts in the ships. Like, how quickly did shippers see this possibility and start using it? Was it fast or slow?

Marc: Well, let me give you just a quick history here. The idea that you could save money by shipping goods in containers came along in the 1700s, okay? This was an old idea, and nobody had ever figured out how to make money out of it. Because what would happen was that, well, you would make a container out of wood and nails, and you’d put your goods in it, and then at the other end of the trip, somebody would break the container apart and use it for firewood. That was a pretty inefficient system, and nobody ever found this to be viable. It actually cost more to ship goods and containers.

What made this whole thing work was the arrival of a guy named Malcolm McLean, who was a trucker — so he didn’t come from the shipping industry — and he understood that what was needed was not particularly a container but a new system for moving freight, okay? And the container was just a piece of what…

Sonal: It was just a container.

Marc: It was just a container, and a lot of people back in the ’50s and ’60s who were in the shipping industry were very enamored of their ships, and they thought they were in the shipping business. And McLean’s basic position was, nobody cares about your ship — they just wanna get their goods from here to there, and let’s design an efficient system for doing that. So, shipping containers first came into use in the United States in 1956. They started being used internationally across the Atlantic in 1966, and the industry was pretty substantial by the 1970s. By that time, most of the older vessels had gone out of service. But it was really in the 1980s when modern logistics took off.

There were a couple things that happened. You had, in this country, freight deregulation — which meant that you could actually sign a single contract to import products that would cover delivering the goods to a port and moving them inland by rail to a final destination, or having a truck pick them up and move them to a final destination. So, you could actually integrate all these modes of transportation and have some assurance that the goods would get there. And then you had improvements in communications. This was referred to as electronic data interchange.

So, all of a sudden, it became possible to run an international supply chain in the 1980s. Now, you could send instructions across the ocean quickly about how you wanted something shipped or how you wanted something made. And so, this innovation — the container — that had really come about in the ’50s, started to make a substantial difference in the world economy in the 1980s when we had the birth of modern supply chains.

Sonal: The most fascinating thing to me — the idea that astounded me most about the box was the idea that the containerization of moving goods allowed things to travel multimodally. That because of that — this modularization — you could now bring things across ship, to train, to plane across the world. And that is, like, a really eye-opening idea. And it’s really interesting that you referenced the EDI, because the analogy that I was thinking of was actually packets and moving data across lines, like the ethernet. And that packetization of data also led to this thing where you can move things across phone lines, ethernet lines, other computer lines, broadband, etc., and essentially reassemble them at the other end.

Marc: That’s a good analogy.

Sonal: That’s a really, really mind-blowing idea. So, I guess the question I have is, what’s happening next and now that you think is interesting in the next evolution of supply chains that is along these lines?

Marc: Well, there are a couple of things that are going on in supply chains, and they’re not necessarily good. International trade and manufactured goods actually grow more slowly than the world economy for the past six or seven years. That’s a big reversal from the previous trend. Why is that? One of the reasons is that supply chains have become less reliable. The ship lines went out and purchased very, very large vessels, and I’m sure your listeners have seen these vessels can carry…

Sonal: I’ve actually seen them first-hand, because I went to the Panama Canal, and it’s incredible.

Marc: Okay. The Panama Canal doesn’t handle the biggest one.

Sonal: That’s right. Because they’re actually rate limited by the Panamax ships.

Marc: The biggest container ships now at sea can carry more than 10,000 truck-size containers.

Sonal: Wow.

Hanne: That’s amazing.

Marc: These are enormous vessels. So, what has happened? Well, imagine you’ve got a port, but instead of having a ship carrying 2,000 containers showing up every day, now you’ve got a ship that carries 10,000 containers showing up once a week. You’ve got a mess on your hands, because you’ve got this enormous load of traffic which you need to get all of these containers out of the port.

Hanne: It’s too much.

Sonal: It’s like a bottleneck congestion.

Hanne: Yeah.

Marc: Yeah, that’s right you’ve got a bottleneck. And this has come at a time when growth and trade has been pretty slow, so there’s considerable overcapacity in the industry, and even so, the reliability has fallen. So, what you’ve seen is, actually, manufacturers and retailers contracting their supply chains. They would like to make things closer to where they’re used, because they think there’s less risk.

One of the things that I think happened in the growth of these international supply chains is that companies paid a lot of attention to cost. They said, “You know, our hourly labor cost in China is a lot cheaper than it is in Detroit.” They didn’t really pay much attention to risk. And risk is a cost factor. There were a number of U.S. companies that failed or came very close to failure because of supply chain disruptions. Key merchandise wasn’t available when they needed it for their factories or for their store shelves.

Sonal: I mean, this is the story of hardware startups. The problem isn’t that they can’t plan out, and predict, and build, it’s that they need to lock down that supply chain inventory at the right time. But, yet, they have that issue — that they don’t know how many products their customers are gonna buy, so they don’t know how much to make. So, there’s a sort of chicken-egg problem.

Hanne: It’s the same thing in book publishing.

Sonal: Yeah, yeah.

Marc: So, you’ve got the container ship lines that essentially created their own crisis.

Sonal: It’s like a success disaster of sorts.

Marc: They got bigger and bigger ships, because that was more efficient for their purposes. Their own costs running ships went down per container as the ships got bigger. They didn’t devote too much thought to the problems of the ports, or the railroads, or the truck lines, and all of them have had a lot of difficulty coping with this flood of containers. And so, I think one question facing this industry going forward is whether these long-distance supply chains will continue.

Sonal: Well, I’d like to ask a question, though — is that necessarily a disaster? Because isn’t that also the inevitable sort of cycle of things aggregating, un-aggregating, lengthening, and contracting, etc.? And also, in that same context, one of the arguments I’ve heard for — there’s actually advantages to shorter supply chains. For example, in the case of, like, hardware and software innovation, there’s this rapid iteration and back and forth that happens.

So, if you have a components manufacturer in Mexico, and you’re designing a chip or some piece of hardware, you could rapidly iterate on your designs without the long delay that happens when you have a big time difference and a bunch of other logistical issues with someone, you know, doing the same thing in China. So, there’s some argument that it’s actually not a bad thing, because it actually speedens innovation almost in some cases.

Marc: In some cases, it may speed innovation. In general, I think that manufacturers and retailers are expecting that it’s going to reduce risk. Another trend that you see is that many manufacturers and retailers are now looking to multiple sourcing. Now, in many industries, it’s cheaper to have a single source, right, because you’ve got huge economies of scale. One factory makes a ton of stuff, and that’s great so long as it’s working.

Sonal: It’s cheaper because of the China stuff, right.

Marc: It’s cheaper, but maybe it’s worth paying a little bit more and have an extra warehouse. We’re seeing a lot of that now. We had, for example, a work stoppage out at the port of Los Angeles — actually, the West Coast ports in general — in the early part of the century, a lockout by the port employers. How did that affect companies?

Sonal: With a lockout, you mean, like, it was, like…

Marc: They locked out the union workers as part of a labor dispute.

Sonal: Oh, right, right, right.

Marc: And so, a lot of companies said, “Well, maybe we ought to redirect some of our traffic to ports on the U.S. East Coast.” Okay, so they still send their goods to Los Angeles, or Long Beach, or Oakland, but they also send a portion of them now to Savannah or New York, because they want to have options. They want to not have the risk that their supply chain will be shut down.

Sonal: There’s another fascinating analogy with the digital world here, because it reminds me of the time of the early days of the internet, when as the internet became super popular and more multimedia started coming online, there were tremendous bottlenecks in data traveling through pipes. And so, they had to figure out new methods to essentially reroute and decentralize it from these central choke points. So, it’s kind of a fascinating thing.

And it also now, by the way, explains when I was in Panama, I was a little struck by this thing where every single ship that goes through spends a day being inspected before you can even put it through, and it costs, like, $1 million per ship to put it through or some — I forgot the amount, but it’s some significant amount. And it just blew my mind, like, there’s so much extra work, but now I understand, because if one ship bottlenecks that entire thing, nothing gets through for, like, the entire day, and that’s a huge blockage.

Hanne: Right. You gotta make sure this is gonna get through.

Sonal: Yeah, exactly. It’s kinda fascinating.

Marc: So, what we’re seeing in container shipping now is a lot like what we saw in the United States with the railroads in the 1870s, 1880s, 1890s when many railroads went bankrupt. We went from a country that had hundreds of railroads, each a few miles long, to a relative handful of large railroad networks. We went through something similar when we had airline deregulation, starting in 1978.

You may remember, we used to have lots of regional carriers around the country, had a number of national airlines, and only two international airlines that were heavily protected by the government. Now there’s a lot more potential competition, and, of course, the carriers have dealt with that by merging. So, you’ve actually now got a situation in which you’re supposed to have cutthroat competition, but they’ve tried to find a way around it by merging and reducing the number of airlines.

Hanne: It’s the inevitable cycle.

Marc: We’re headed in the same direction with container shipping now. Many container carriers are in financial distress, a lot of them have merged into the big carriers. There are now, probably, three so-called alliances of container carriers that kind of dominate world trade. So, we may be in an environment in which there are few enough players that they’ll be able to have a better handle on prices on shipping rates, and that will mean less competition. That will be good for their shareholders. It probably won’t be good for shippers.

Future of the world economy

Sonal: Okay, so to wrap up, we started this with your view that connects the dots between all of these books, and you have this perspective of this economist-historian. And, you know the question is — is it good to know that this is an ordinary economy, or are we just talking about cycles of things that are just gonna inevitably decline and grow? How do you know it’s just not a typical waning, and that this actually really is something different?

Marc: I think this actually has a lot of political implications. For decades and decades, we’ve trained people to believe that the government can provide a very steady income, can provide low unemployment, can provide rapid economic growth, and I think the government’s ability to do this is limited. What we’re seeing, I think, and not just in the United States, is somewhat of a crisis of expectations.

Hanne: It’s a reckoning.

Marc: If you take a look at what’s going on now in Europe, or in Korea, or in Taiwan, people expect more of their public officials than their public officials can deliver. People want their incomes to grow quickly. Their public officials promise, “Yeah, we’ll bring back the good old times. We’ll make your incomes grow quickly.” But, in reality, we’re in a normal age in which people’s living standards rise slowly.

Hanne: I feel like while you’re saying this, all I can think is, like, “Well, only really time can teach us that that’s not…” You know, because I’m trying to think, like, “Well, what can we do to reset those expectations,” but we can’t really. It’s just time and not…

Sonal: Understanding, possibly?

Hanne: Well, and not having the same growth, right, getting used to not having the same growth, which is sort of disheartening, when you think about it.

Marc: It is. On one level…

Hanne: Because my question is, like, so how did we move beyond that?

Marc: On one level, it’s disheartening. On another level, I think we have trained people to believe that the government can deliver things it really can’t deliver.

Hanne: So, how do we start to undo that?

Marc: There is a question about how the available income is distributed, which is really quite separate from the question of how fast productivity is growing, how fast the economy is growing. And I think we have to have a real discussion about how income is being distributed and how automation is going to affect our workforce.

Sonal: This is top of mind for everybody, including us, our guests, everybody.

Marc: And I bring this up not really in an economic sense, but almost more in a psychological sense. There’s a lot of concern about where the jobs are going to be for people in the future. I’m not too concerned about that. I’m pretty confident that we’ll have ways in which people can earn livings. But so many people get some degree of satisfaction from their work, and if what we’ve got is a world in which people are doing part-time work, occasional work, unsteady work…

Sonal: I mean, work itself has become containerized.

Marc: That’s right. What are people going to be moored to, in this sense? And we can do anything we want to guarantee your income, but the fact is, if all you do is wake up in the morning and watch television, because you’ve got nothing else to do, you’re not gonna be very happy. So, I think we have an issue here that really goes beyond economics to…

Sonal: To finding meaning.

Marc: Finding meaning in our lives as human beings.

Hanne: To the human condition, yeah.

Sonal: One last thing. You talk about a crisis of expectation, but the reality — and this is actually another thing that connects the dots — is that these problems are multifactorial, multimodal. There’s crisis expectations around the world, and they’re playing out in ripples and waves in different ways across, you know, France, Europe, the U.S. in so many different ways. So, I also wonder if there’s some change in the geographic or governance structures that we have to think about. Do you have any thoughts on, sort of, the geopolitical implications of this? Is there any, like, parting thought on that front?

Marc: We’ve seen, obviously, a big shift away from faith in the nation-state, to people wanting power and control closer to home. I think we’ve seen over the centuries cycles in that — that that sort of comes and goes. Will city governments be able to deliver satisfaction in a way that national governments can’t? I’m not convinced of that, but in some issues, some areas related to people’s quality of life, they can be very important. I think another issue that we face is that there are things we know that will improve productivity over time.

Sonal: Yeah.

Marc: We can’t predict how that’ll work, and we still need to make those expenditures. So, for example, we know that improving education levels is, in general, good for a country’s productivity. So, we need to invest in education, but can we say if we spend an extra billion dollars in education now, that it will improve productivity three years from now? We can’t say that. We’re doing this somewhat on faith.

Sonal: Especially if the skills and jobs of the future change.

Marc: Well, that’s correct. We know that as a general proposition, it’s been important for economic growth that we’ve had scientific research going on. Does that mean that if we put more money into scientific research now that we will be able to benefit from the consequences at any predictable time in the future? The answer to that is no.

Hanne: Right, and these are the reasons exactly why we tend not to make these decisions.

Marc: That’s exactly right. That’s exactly right.

Hanne: Or we can’t.

Marc: These are very tough choices, because you’re asking for our tax money to be spent, and you really can’t promise the return.

Sonal: It’s like investing in the future, difficult to do.

Marc: Exactly. Exactly.

Sonal: Well, Marc, thank you for joining the “a16z Podcast.”

Marc: Thank you so much for having me. It’s been great fun.

Hanne: Thank you.

• 

  Marc Levinson
• 

  Hanne Winarsky
• 

  Sonal Chokshi is the editor in chief as well as podcast network showrunner. Prior to joining a16z 2014 to build the editorial operation, Sonal was a senior editor at WIRED, and before that in content at Xerox PARC.

  Follow Twitter

Is a network — whether a crowd or blockchain-based entity — going to replace the firm anytime soon? Not yet, argue Andrew McAfee and Erik Brynjolfsson in the new book Machine, Platform, Crowd. But that title is a bit misleading, because the real questions most companies and people wrestle with are more “machine vs. mind”, “platform vs. product”, and “crowd vs. core”. They’re really a set of dichotomies.

Yet the most successful systems are rarely all one or all the other. So how then do companies make choices, tradeoffs in designing products between humans and machines, whether it’s sales people vs. chatbots, or doctors vs. AIs? How can companies combine the fundamental building blocks of businesses — such as network effects, platforms, crowds, and more — in a way that lets them get ahead on the chessboard against the Red Queen? And then finally, at a macro level, how do we plan for the future without falling for the “fatal conceit” (which has now, arguably flipped from radical centralization to radical decentralization) … and just run a ton of experiments to get there?

We (Frank Chen and Sonal Chokshi) discuss all this and more with Brynjolfsson and McAfee, who also founded MIT’s Initiative on the Global Economy — and previously wrote the popular The Second Machine Age and Race Against the Machine. Maybe there’s a better way to stay ahead without having to run faster and faster just to stay in place like Alice in a tech Wonderland.

Show Notes

• The basic economics of networks and the concept of complements [1:08]
• Discussion of whether “the firm” will survive [10:08], and our inability to simulate all possible outcomes [15:34]
• The ability of crowds to problem-solve [20:47], with caveats related to biases in AI [23:34]
• Advice for businesses today [30:45]

Transcript

Hi, everyone. Welcome to the “a16z Podcast.” I’m Sonal. Today we’re doing one of our book podcasts around the new book just out, “Machine, Platform, Crowd.” The authors previously wrote the popular book, “The Second Machine Age.” And before that, their book was “Race Against the Machine.” Sensing a bit of a theme here. So in this episode, we cover those themes, first starting with a bit of Econ 101 around network effects, complements, and other key concepts. Then we discuss how this all plays out organizationally, especially given trends like machine learning, blockchain, and crowds — and tackle the tricky question of whether networks can replace the firm. And where are we in the classic question around the future of the firm? And finally, what can companies do more concretely?

Frank Chen joins the conversation in between, as well, to share his perspective on what he sees, given his role as head of investing and research at a16z. But our main guests on the episode, both from MIT, are Erik Brynjolfsson and Andrew McAfee, who I’m gonna call Andy, is that okay?

Andrew: Otherwise, I’m gonna mistake you for my mom.

Sonal: Good, I don’t wanna be mistaken for your mom.

Andrew: That would be weird.

Sonal: I’m way too young to be your mom. We kind of go way back in the sense that I met you years ago and…

Andrew: Not as far back as I go with my mom.

Sonal: No, no. Let’s be very clear about that.

Andrew: So, you and I will do Andy. Okay?

Networks and complements

Sonal: All right. Good, we’re doing Andy. So, this is your third book together. The real thrust of your work is that this is unprecedented in the speed at which we’re changing and what the effects are. And I think a great theme for this conversation is to sort of break down how those changes are going to play out, and where they’re happening.

Andrew: Yeah.

Erik: Yeah. Well, but let me just push back on that first part a little bit, because in Silicon Valley, everybody agrees with that. And we agree with it. That’s very clear. But we were reading people who didn’t. One of the things that got us writing our first book, “Race Against the Machine,” was there were people who were talking about “the great stagnation,” and how there were no good inventions anymore. Nothing good was invented. In particular, Tyler Cowen, he was spot on that median income had been stagnating. And that was kind of troubling for us. Because, you know, I had been taught the slogan that productivity isn’t everything. But in the long run, if we just have tech progress, everything else takes care of itself. And when Tyler showed us that evidence, we were like, “Oh, this is a real problem.” But we refused to give up on the idea that technology was just doing amazing things…

Andrew: We weren’t gonna let a little evidence get in the way of all the <inaudible>, for God’s sake.

Sonal: Yeah, no. No way. No dammit, we’re not letting that happen.

Erik: But fortunately, we figured out a way out of it. And the way out of it is that even though technology is making the pie bigger, there’s no economic law that everyone’s going to benefit from it. It’s possible for some people to get left behind. Now, to be clear, that’s not what happened for most of the past 200 years. But the past 10, 20 years, there really have been more and more people being left behind. And so you could get stagnating median incomes, even as some people, maybe in the top 1%, got fabulously wealthy. And that helped us reconcile these different perspectives. And it led to a whole broader set of discussions about the way that organizations, and society, and business processes aren’t keeping up with these amazing technologies, and some of the dysfunctions that can create and some of the opportunities that can create.

Sonal: So what are some of the big — well, I think we should break down the fundamental building blocks of a lot of the arguments that you make throughout your work. So let’s talk about networks. And one of the biggest questions I had reading your book was, is a network going to displace the firm in the future? We talk a lot about network effects in our business.

Erik: So networks, sometimes economists call them demand side economies of scale — and it’s basically the idea that a product or service becomes more valuable the more other people that are using that product or service. A classic example is, you know, a telephone, or a fax machine, WhatsApp, Facebook. And you can have supply side economies of scale, just to distinguish that — that’s when the costs get lower as more people use it. And both of these things lead to the big companies winning.

Sonal: And just for shorthand, we tend to describe supply side economies of scales as just economies of scale, the demand side economies of scale as network effect.

Erik: That’s the more common way.

Sonal: More generically.

Erik: And we use both sets of terminology. It’s sometimes useful to talk about supply side and demand side, because a lot of the economics become more intuitive once you understand that there’s the demand side and the supply side, and they both can get better as you get bigger. And then to add a little more layer of subtlety to it, you can have traditional single-sided network effects, like other people using the same telephone, or you can have a two-sided network. And that’s what — really the platform revolution, a lot of that has been triggered by the growth of the so-called two-sided networks.

And the idea there is that it’s not necessarily people using the same product as you, but it could be people on the other side using a different product. So, like, drivers and users are using slightly different apps. And me as a user, I don’t really benefit when more users are also, you know — I want more drivers. And the drivers want more users. So you care about the people on the other side of the network.

Sonal: Except when you’re pulling, because then you do care.

Erik: That’s right.

Sonal: And that’s a case where you do want them on the same side.

Erik: Exactly. And then to make it even more complicated, you can have two sided and one sided at the same time, you can have economies of scale. So you can layer them. You mentioned the word building block. Let’s start with these primitives. And then you can start combining them in different ways.

Andrew: This really starts to turn into three-dimensional chess, because the right way to think about the app ecosystem in Apple is not any kind of one or two-sided network. It’s an n-sided network.

Sonal: Multi-sided, yeah.

Andrew: And lots of different groups of people who value things on the other side, but we don’t decide what the sides are, and we let the self-selection happen. And you just watch the vortex form around that ecosystem. And the only way to understand that is by doing what Erik just did — start with network effects, one-sided, two-sided, two goes to N, value goes to N.

Sonal: Okay, that’s great. And then let’s probe on one big thing, which is we talk about network effects. But let’s quickly define complements in this, because that’s a term that’s frequently used. And I think it has a lot of misconceptions around it.

Erik: Sure. One of the key economic building blocks that we talk about is complements. And a complement is a very simple concept. It’s the idea that one product is more valuable in the presence of another. So my left shoe is more valuable if I also have my right shoe.

Sonal: Well, that’s an obvious example.

Erik: Yeah, yeah. Software is more valuable with the right hardware. And so, complements can be physical, they can even be organizational. Well, so you may have a system that taps into the crowd that’s more valuable when you have a global internet that allows you to do that. So you can have organizational, or technical, or physical complements. And you can sell products that are complementary to each other.

Sonal: The razor blade is the classic example.

Erik: Yeah, razors and blades. And sometimes when you have products that are complementary to one another, it actually can be profitable to give one away to increase the demand for the other one. So people famously gave away razors to sell blades. And this can interact with the network effects and the scale economies. It’s not a good strategy if you don’t have those other things. One of the things that, you know, makes us tear our hair out is that, you know, when MBA students are like, “Oh, yeah, we’ll just give it away.” Like, where is the underlying strategy?

Sonal: Oh, so if you’re just saying like, “I need to do freemium,” without really understanding the underlying strategy that we’re trying to accomplish.

Erik: Exactly, disastrous.

Andrew: And complements are weirdly subtle. And Erik just explained…

Sonal: This is why I wanna ask about it. Because it’s a very nuanced concept.

Andrew: Erik just explained them super clearly. The Econ 101 example that I always fall back on is hamburger meat and hamburger buns. And so, if the price of hamburger meat goes down, demand for buns is going to go up, even if the price of buns doesn’t change. That’s the key thing. The price of one good can stay the same and demand for it will go up. The complements are so tricky that they actually tripped up Steve Jobs really badly. This is not lore, this is fact. He did not want to open up the app store to any outside developers. He thought he had to maintain super tight control over that digital environment. And when the iPhone first released, it did not have any external apps on it. He fought boardroom battles for about a year with people who said, “No, you need to open this up.”

Sonal: What made him cave?

Andrew: Pressure from really smart people inside and outside the company, people on his board and executives at the company. What he didn’t fully realize is that if you open up the app store and you curate successfully, you have just opened the door to this massive number of complements, each one of which is going to nudge out demand for the iPhone. And even if each one only nudges that demand outward…

Sonal: Like 99 cents worth.

Andrew: Yeah.

Erik: Oh, no, even less.

Andrew: Even less.

Erik: Just to be clear, we’re not talking about the literal money…

Sonal: Yeah, I know. I know. Exactly.

Erik: Yeah, we’re talking about the fact that it makes the…

Sonal: It’s a relationship that makes the entire…

Erik: It makes the phone…

Sonal: It makes people want the phone. Remember the early days of the iPhone — I still don’t have an iPhone, I have an Android. But I still remember to this day, the first thing people would say I’m like, “I don’t really like the iPhone that much.” And they’re like, “Oh, it’s not about the phone. It’s all about the apps. It’s all about the apps.” That was the line all the time.

Erik: Angry Birds.

Andrew: Yeah. And the only way to understand the value of opening up that app store is to understand that you are unleashing this tidal wave of complementary goods that were priced at all different price points, including zero, which is awesome. So zero is a really great price. But the more fundamental thing, I think, is that it shifted out demand. It nudged demand upward for the other complementary good, the iPhone itself. And once you grok into that, then you say, “Oh, I got to find all kinds of different ways to do this and play three-dimensional chess with my platform.”

Sonal: Is the corollary of all this that “closed” will never win then?

Erik: No, it’s not nearly as simple as that. But it does show you that if you can leverage these complements, you can create not just a one-time win, but in a whole ecosystem, because Andy’s story turns into a virtuous cycle where the more demand for the iPhone…

Sonal: Right, flywheel.

Erik: Exactly. It’s a flywheel. So that can work very well. But it’s not like you always open up, or you always build complements.

Sonal: Right. Because I was gonna say, a lot of the winners until now have been closed companies.

Erik: Yeah, absolutely.

Andrew: Yeah. And Apple was comparatively closed against Google and the Android ecosystem. One of the things we say is, there is not one right answer. There is not one recipe that you follow for success with machines, platforms, or crowd.

Erik: There are principles.

Frank: And for entrepreneurs who are listening, understanding complements, and the way the people who are creating these ecosystems that have complements is super important. So we’ve been talking about complements where the more apps in the app store, the more attractive an iPhone. So think about that when you’re thinking about development tools for these platforms. Xcode, Visual Studio are so important to Microsoft and Apple, because they’re creating these complements and therefore the desirability for their iPhone. That’s where they make all their money. So if you think, “Hey, I’m going to create a better development tool. I’m going to create a better Xcode.” Like, think again, because Apple is going to spend as much money as it needs to defend a complement universe.

The future of the firm

Sonal: Crushing you. The question that comes to mind for me is what this means for companies.

Frank: So one thing that — conventional wisdom now is, we fund companies whose defensibility is a network effect. In other words, we’re in Lyft and Airbnb precisely because once you have all the hosts, you’re going to get all of the renters, right? And so, one thing to think about is, maybe in the future, even the firm that creates the network effect gets decentralized. Who needs a firm? Why don’t people just come together and we’ll create the right set of incentives for the network to behave? So you can imagine an eBay where there is no company. There’s just a network coming together with the right set of incentives.

Andrew: That was how we wound up the book, is trying to grapple honestly with this question of in the universe that can be turbocharged by the fact that everyone’s got a device, that we’ve got this completely decentralized cryptocurrency system you could pay people with, that we’ve got these technologies of radical decentralization.

Frank: Like the blockchain.

Andrew: Like the blockchain.

Frank: Like the blockchain. Public distributed ledger. Every transaction…

Andrew: Where you could stick…

Frank: …everybody’s <crosstalk>

Andrew: …contracts and code into those things. You can do a lot of the stuff that we used to need a company for. The question gets teed up, are we still gonna have companies in the future? And as Erik and I started to think about all the stuff that we’d learned and tried to digest, our answer was an unequivocal yes. And the main reason for that is that ownership of a thing matters, simply because almost — well, every economist, I think, that we’ve talked to would agree that you can never write a complete contract that will specify exactly what everybody is going to do in all future states of the world.

Sonal: Every possible contingency cannot be accounted for.

Andrew: And the reason for a firm is it gets to make the decisions that are not contractually specified elsewhere. And it gets all the value that’s not apportioned elsewhere in the network.

Erik: It starts with Ronald Coase…

Sonal: Of course, the classic “Nature of the Firm,” 1937 or something.

Andrew: ’37. He’s a hero. He was 9 years old when he wrote that. He was in his 20s or something…

Sonal: Did you say 90 or 9?

Andrew: No, he was in his 20s.

Erik: Yeah, he was in his 20s, 26 I think he was. But then and then more recently, Oliver Hart, who was my thesis advisor, and Bengt Holmström, one of our other colleagues at MIT, elaborate on that, as Andy was saying, with this so-called incomplete contracts theory. One of the blinders that a lot of people, especially technologists have, is they say, “Hey, we can just write everything down under an engineering mindset. We’ll write a complete contract that covers all contingencies.” And the reality is, the world is just too complicated to cover every possible contingency. So when you own a car, you can sell that to someone else. And whoever owns the car gets to have all of what are called the residual rights of control, everything that’s not specified in the contract. You want to change the color of it, that’s what ownership means. And ultimately, you take that to the level of the firm. A firm is an aggregator of a bunch of assets and owns certain things. And that means, that gives them a certain power, that gives them certain incentives of how those objects are used.

Andrew: As Erik and I were trying to reason our way through this and convince ourselves to one view of the world here, this amazing real-life experiment happened, which was the Dow.

Sonal: Yeah. And let’s do a quick terminology thing. When you say the Dow, you mean the corporation that was formed, but that’s very different than a DAO which is a decentralized autonomous organization or decentralized autonomous corporation. This is the Dow, the entity.

Andrew: This is the thing called the Dow.

Sonal: The proper noun, not the generic noun. Yes.

Andrew: The Dow, which was intended to be a completely owner-free, completely decentralized organization along the lines that you just described. And it got hacked, and somebody found out how to treat it like an ATM essentially. So, to the extent there was a group of people, kind of, behind it, they collectively freaked out and thought about what to do. And then they made this fairly autocratic decision — looks a lot like an ownership decision for me, to reset the clock on the entire Dow.

Erik: They became de facto owners, they asserted those rights in a way.

Sonal: That’s right.

Andrew: A de novo. They said, “Okay, we’re gonna do this. And if enough of you go along with this, then this is what’s going to happen.” It was extraordinary for a very decentralized organization, it was kind of heavy.

Sonal: I mean, I love you’re saying something counterintuitive, which is a firm is not going to go away. It’s gonna actually look the same as it does now, then. But when we talk about the transaction cost of all this coordination, and why you need management — or even you have this incomplete contract theory, and people — you can’t predict every contingency. What if we have an algorithmic AI who’s able to then account for every one of those contingencies versus — we’re basing our theories right now on what we know already. We don’t know how it’s gonna play out in the future.

Andrew: Amen.

Erik: Well, we’ll never say never. And yeah, if there’s an AI that has magical properties that we can imagine, you know, all bets are off, of course. But we’re talking about a world right now, where the blockchain and related technologies are allowing radical decentralization of lots of types of decisions. And that’s really important. It’s changing, creating a lot of new opportunities, but it doesn’t change everything. And there are still some core things like this concept of incomplete contracts. Anything that’s not explicit that you can’t write down, maybe you can’t anticipate, and maybe the current AIs can’t anticipate, then those are the residual and that’s where ownership actually…

Andrew: That leads to something like “company” being an enduring part of the economic landscape.

Sonal: I mean, I would even make it more basic, which is, it’s human nature that people — at the end of the day, systems of networks that are online, or in a company, or any other form, are made up of people, and people are fallible and are emotional.

Andrew: And fractious, right?

Sonal: Yes, they wanna fight.

Andrew: If we look at the breakdown in the Bitcoin community and the civil war going on there. Okay, one reason you have management is to say, “Gang, we’re going to go this way and not that way.” And disagree and then commit, as opposed to disagree and then disagree.

Erik: And we all have bounded rationality. Friedrich Hayek called it — was the fatal conceit, the idea that we could plan everything in excruciating detail. The world is far too complicated for any one person or any one group of people to do that. There’s even a, kind of, a Red Queen phenomenon, that the more sophisticated you are, the more sophisticated your competitors are, your customers are, your suppliers are.

Sonal: Why is it called the Red Queen phenomenon?

Erik: Oh, so Alice was…

Sonal: From Victoria Aveyard’s novel, or…

Erik: No, from “Alice in Wonderland.”

Sonal: Oh, from “Alice in Wonderland.” Of course.

Erik: You have to run faster and faster…

Andrew: Faster and faster just to keep up.

Erik: …just to stay in place.

Sonal: Got you.

Erik: So if you get more sophisticated, all those other parties are getting more sophisticated too. You’re not going to be able to completely anticipate what they all do, because they’ll be even more clever.

Andrew: But think about how crazy this is. Hayek brought up the term “the fatal conceit” to demolish this idea that we could centrally plan an economy. And at the time, when a lot of intellectuals in the West were excited about Soviet-style central planning, Hayek wrote one paper and just demolished it. There’s an almost 180-degree reverse — perhaps “fatal conceit” going on — among the fans of radical decentralization as opposed to radical centralization.

Sonal: Right. So you’re saying the same phenomena is at play, just in a different direction. But I wanted to add something, too, because I was gonna say, there’s now some claims out there that the power of simulation has gotten so good that we might be able to actually move to that fatal conceit of being able to centrally plan an economy, because of all these data and machine learning, you know, sort of, signals and whatnot.

Erik: So, Alan Greenspan, of all people, I asked him about computers and the ability to simulate the economy. And he was a chairman of the Federal Reserve, you know, set interest rates and everything. And he said, “Well, yeah, we can understand a lot, lot better. But all the companies are reacting that much faster as well.” And so it’s exactly this Red Queen phenomenon, that however much the Federal Reserve advanced, each company advanced, all the other guys are doing the same thing. If you could freeze the rest of the world, and you were the only party that had access to cloud computing and Moore’s Law, etc. Yeah, maybe you could stay 1 to 10 steps ahead of them. But that’s not the way the world works.

Frank: There’s a great story from the early days of AI on this fatal conceit idea, which was in the late ’80s, Japan tried to organize their entire industrial policy around creating artificial intelligence.

Erik: Fifth generation.

Frank: The fifth generation. Supercomputer.

Sonal: Like what’s happening in China right now.

Frank: Built around expert systems, optimized all the way down in the silicon. So you can imagine, silicon optimized for Lisp, right, so that we can build apps. <crosstalk> And it was a complete failure, precisely to this idea of — you actually can’t plan anything, right? What happened out of the ’80s was more the rise of client server computing and Microsoft Windows. Nobody anticipated that.

Andrew: And the idea that we’re out of that world because of Moore’s Law, because we have much more computational power now, I find that ludicrous.

Sonal: Well, tell me why? If we have this accelerating, growing, fast-happening thing — and I don’t want to make it a crutch to say, like, we can’t predict the future, dot-dot-dot, blah, blah, blah. We already know that. But why not? A lot of things that were tried before didn’t work because it was the wrong time. Why wouldn’t that be possible now? Like, can’t simulation work there?

Frank: Yeah. I mean, speaking as an investor, you know, who’s trying to predict the future and often gets it wrong.

Sonal: As you should.

Frank: You know, it’s hard to imagine a better system than the one we have, which is, let’s spend a little money and run a ton of experiments on businesses to figure out what people want. Because until you have it in the world, you’re not sure what the people will want.

Andrew: And that’s not called simulation in the face of massive computational power. That’s called entrepreneurship and capitalism. It’s a very different approach.

Sonal: I agree with you guys. I find that.

Erik: So, if anything, the data is going the opposite direction.

Sonal: Which is?

Erik: We’re seeing less planning and predicting, less five-year plans, we’re gonna do this, and a lot more experimenting, testing, fail fast. That seems to be a model that works a lot better.

Sonal: But the other thing I was gonna say is, like, I look at countries like China and their incredibly coordinated efforts. And while I agree that past central industrial planning efforts have failed, for various reasons, I don’t know, I think there might be something to it this time. I just want to make sure you guys really disillusion me of that. Help me let it go.

Andrew: And our colleagues, Daron Acemoglu and James Robinson wrote this amazing book called “Why Nations Fail.” And their answer was really straightforward. Nations fail because they have extractive…

Erik: Extractive institutions.

Andrew: Extractive institutions, where an elite grabs power and they just suck up the value of more.

Sonal: Arguably, that’s why companies fail too. <crosstalk>

Andrew: Exactly. And they make sure that their descendants…

Erik: Yes, that’s a good analogy. You should write the next book.

Andrew: And they hand down power to their descendants, and they just make sure that they pervert the rules of the game to benefit themselves. That’s as opposed to inclusive institutions, where you have an honest shot of making the most of your human capital. Now, which one is China? They took big steps in the direction of inclusion by opening up to a market economy. Would we call that authoritarian state, one of actually inclusive institutions? I would not.

Sonal: I think that’s the legitimate thing to say. Okay, so just going back to this idea of extractive institutions. So, I do think it’s interesting that there are now networks that are coming up that are letting people participate differently as owners…

Erik: For sure.

Sonal: …in different ways. And that is where I think this topic of ICOs and token launches is really interesting.

Erik: Part of the power, as Hayek would have said, is that you decentralize some of the local knowledge. They have information that nobody else has. And if you could…

Sonal: That’s right, or resources, like if it’s a computing power…

Erik: Yeah, they have skills. Exactly. If you can move the decision rights to where that knowledge is, you’re going to be better off. And one of the great things that technology has allowed us to do is move around decision rights, move around ownership. So hopefully, if you do it right, you get a better match between the incentives and the decision rights.

The power of crowds

Andrew: The entire third section of our book is about this rebalancing necessary between the core institutions of a company, and the crowd available over the internet now. How much more room there’s very likely ahead of us, with crowdfunding, with crowdsourcing, with different ways to tap into what people can do to give them an ownership stake, to get them bought in and pointing the right direction. Have we scratched the surface of that?

Erik: Let’s talk a little bit about Joy’s law, that no matter what company you work for, most of the smart people in the world work for somebody else. It used to be limited what you could do about that, because there’s only so far you can communicate. But now for the first time in history, a majority of the world’s people are connected with a digital network. So they can access all the world’s knowledge. And part of it isn’t necessarily that they’re smarter out there, part of it just comes from the raw variety, the diversity, the variance. Within a company, you tend to have people who are like-minded, they’ve trained the same way. That’s who they get hired. And maybe the way to solve a problem is with an entirely different approach. And that may be somebody from a different culture, a different way of looking at the world.

And you’re very unlikely to have that diversity inside of a company. It works against it. But if you can find a way to tap into it. One of our colleagues, Karim Lakhani is now at Harvard Business School, he was a Ph.D student at MIT, has done just case study after case study of examples where tapping into the crowd blew away what companies were able to do internally.

Andrew: He worked with the National Institutes of Health to try to improve the speed and accuracy of sequencing human white blood cell genomes, which are really complicated but important to sequence. The National Institutes of Health, which I would call the core of the medical establishment.

Erik: Core in the sense of core versus crowd.

Andrew: They had an algorithm that could do a run in about four hours with about 70% accuracy. There was a faculty member at Harvard Med School who made a big improvement to that algorithm. He developed one that got them up to about 75% accuracy. Karim then worked with the NIH and Topcoder to make this an algorithmic challenge and open up to the crowd. And the best solutions got down to about 10 seconds and about 80% accuracy.

Erik: From 4 hours to 10 seconds.

Andrew: So we called up Karim and he goes, “About average. When I run a crowdsourcing tournament, this is the magnitude of improvement I expect to see.” The last part of that story that continues to blow us away is that they interviewed the best performers who submitted the top-performing algorithms. None of them had a life sciences background. There was not a geneticist…

Sonal: Oh, that’s the best part of the story.

Andrew: …there was not a biologist among them.

Sonal: So crowds and prediction markets are similar. What’s the difference?

Andrew: I would say a prediction market is one way to harness crowd wisdom. Markets do a really good job, overall, on aggregating knowledge.

Erik: Markets tap into the crowd. Google taps into the crowd because their search algorithm basically exploits the link structure that all of us contribute whenever we make pages. There are lots of ways of tapping into the crowd, but being clever about how to reach them, motivate them, aggregate them — still a lot of work to be done on that.

Frank: Let’s talk about the nature of work. Because I think what people do in that firm, either inside or outside, probably changes a lot. So, we have this idea that human decision-making is sort of fundamentally flawed in that, like, there’s biases that you bring to your decision-making that you don’t even understand. So when you’re thinking through, you’re still going to make the same mistake because you don’t understand that you have that bias.

Andrew: After all, walking you through your decision-making process is your brain that came off that flawed decision-making process in the first place. It’s not going to catch its own mistakes typically.

Frank: Right. So it’s a permanent blind spot. And by contrast, you would sort of assume that a machine learning algorithm, trained with a carefully selected broad set of datasets, will have a decision-making efficiency or effectiveness better than, you know, flawed humans. So if that’s the case, what do people in firms do? Like, how do you prepare for this world, where there’s going to be machine learning algorithms that can, in general, make pretty good decisions. And then there’s this idea that, like, maybe the talent is better outside your company than inside your company. So what should you do? Should you join a company?

Erik: It’s just breathtaking what it can do. But it is far, far from being AI complete, being able to do everything that humans can do. There’s a certain class of problems that it’s kicking butt on, but that’s a tiny sliver of what human decision-making is. Even just defining what the problem is, exactly what needs to be done, that’s half the battle. But you need humans to do that. There’s a quote that we had from the book from Picasso, “Computers are useless. All they do is give you answers.”

Sonal: I was a little shocked Picasso was alive when computers were…

Andrew: He actually said that. We went and wholly investigated that one. He said that.

Sonal: I know. I just never associate Picasso and computers. It’s amazing.

Erik: Well, he’s a brilliant guy in a lot of different ways. And obviously, he didn’t know much about the latest neural network systems. But his understanding was spot on, that simply giving the answer isn’t necessarily the most interesting or important part of solving problems.

Sonal: Kevin Kelly actually makes this argument in “The Inevitable.” We had him on the podcast, that the number one job of the future for humans that humans preserve — and this is I think what you’re getting at — is that we ask the questions and computers answer. But I have to say, I actually disagree with that a little bit. Because I’m seeing a new class of generative AI that makes me wonder if they’re going to be asking you questions that make us want to answer differently. I mean, there’s all kinds of interesting things.

Erik: Our brains are made of atoms and so are computers. You know, I’m not going to say that there’s some things that they just can never touch.

Andrew: But I agree, which is that on average, our wetware is amazing. But it’s got a host of bugs, and biases, and glitches in it, that machine learning systems, and properly-configured algorithms in general do not have. So, if you could only pick one of those two entities to help you — the good news is, that’s a false choice. We don’t have to make that choice. And I think the art going forward is being more clear about, “What are we actually good at?” versus what the machines are actually good at. The happy news is that they have very different failure modes.

Erik: Yeah. And I think that’s exactly the key point. It’s a matter of how we can leverage each of them. Because machines have biases as well.

Andrew: Yeah, algorithms are biased by definition.

Erik: It’s not just [that] somebody designed them, but also the training data that they get. I mean, if you decide to give loans based on all the loans that have been approved or rejected in the past, that could have some biases built into it. And some of these neural nets could have billions of connections. Getting it to sort out how exactly — it’s not gonna be one of those — says, “Okay, discriminate against women.” But there may be some very subtle interactions that are hard to anticipate or explain. That said, at least the machines can be tested and improved. And it’s often easier to do that than it is with humans.

Andrew: We are really resistant to having our wetware tweaked. We really just don’t like to be told that we’re glitchy, and here’s the fix and just go do that, no. There’s a concept…

Sonal: That’s the story of most marriages.

Andrew: Yeah, most marriages and most everything, right? It’s really, really hard to do. There’s a concept from linguistics that I find incredibly helpful for helping understand what I think some of the most durable human advantages in a world full of machines will be. And it’s a concept called the intuition of the native speaker. And what they mean by that is, if I look at any English language sentence, I can immediately tell if it’s grammatically perfect or not.

Erik: You just hear it in your head.

Andrew: Yeah. We are the native speakers of the human-created world. Computers are doing this as their second language. I believe we have a massive advantage. We are the native speakers about this reality around us.

Erik: Rather than trying to build a system that does everything from soup to nuts, you get some kind of a division of labor. Sebastian Thrun described a system to us recently that was just fascinating. He’s at Udacity. And a lot of…

Sonal: Another a16z company.

Erik: Yeah.

Andrew: Rock on.

Erik: All right.

Sonal: We make good investments, hey.

Andrew: Listeners at home, we don’t have a list of a16z companies that we’re ticking off.

Sonal: Yeah. I was gonna say, this is all natural, organic. Nobody’s planned a thing, I was just gonna say.

Erik: We do have a list of cool companies, you know, which seem to overlap for some reason. But, you know, Sebastian described how they get incoming traffic in their chat rooms of people asking about their offerings. They decided, “Let’s take this data and we’ll see which of these conversations lead to sales, which ones don’t lead to sales, and label them that way. And then train a neural net about which replies were successful.” And then what they took with those replies, they didn’t try to have a standalone chatbot that then talked to customers. Instead, they had the human salespeople keep interacting. But when they saw one of these more common error modes, they would gently prompt the not-so-good salesperson, you know, “Maybe you want to give them this set of answers or this other set of responses.”

Sonal: So it’s kind of getting their argumentation idea.

Erik: It’s absolutely argumentation. Because there’s a long tail of other questions that the bot had no clue what they were about. So it could help with the most common sets of queries. And this is, I think, a pattern that you see lots and lots. You see it among radiologists. You combine the two and you end up having fewer false positives and fewer false negatives.

Frank: Yeah, I love this idea of sort of machines and humans working together. And I think it’s only a matter of time before we walk into a doctor’s office or a lawyer’s office, where that isn’t the fundamental interaction, and we’ll just be horrified like, “Where’s your AI companion? Why are you trying to do this yourself with your biases?”

Sonal: Oh, that’s fascinating.

Andrew: I couldn’t agree more. Why on earth would I expect my GP, who’s a really good doctor, to be on top of the accumulated mass of human medical knowledge and keeping up to date with the latest developments in all the fields that might relate to what I walk in the door with? That’s an absurd request on a human being. Now, I want that person to be well trained. Even more, I want them to be able to empathize with me, and get me to go along with the course of treatment and get me to buy-in to what’s going on. Because that AI in the background that’s got access to my test and my lab results, again, assessed jaundice in my skin and, you know, how white the sclera of my eyes are, that’s going to be the diagnostic expert in the not too distant future at all.

Sonal: That everybody wants, right?

Frank: That’s exactly right. And I want AI not in the backroom, I want it in the room with me when I’m doing the conversation with the doctor.

Sonal: A seat at the table.

Andrew: You’re right.

Erik: Well, with a seat at the table. It’s a theme that comes up again and again. We talk about mind and machine, product and platform, core and crowd. And we don’t want to give people the mistaken idea that you just cross off the first words of each of those lists and only do the second one.

Andrew: The mantra that I’ve learned is that tech progress rewrites the business playbook. And what the two of us believe is that the way the playbook is being rewritten these days is in favor of machines, platforms, and crowds. So the balance needs to shift more in those directions.

Sonal: So the playbook is in favor of machine, platform, and crowd.

Erik: As opposed to…

Andrew: Mind, product, and core.

Erik: Right. So each of them is really a dichotomy. And the most successful systems are rarely all one or all the other.

Andrew: That’s right.

Applications for today’s businesses

Sonal: A couple of threads that we didn’t get to pull. One question I had when we were talking about not all the talent is inside your company. And, you know, a lot of people talk about open innovation as a way to kind of get around that, like, open source communities, etc. What does that mean for business concretely? What does that mean for core, in the way that you’re defining core, and deploying the power of the crowd? Like, does a business whose main strategy is their core business, does that mean that all their innovation is now outsourced to the crowd? Or is it the other direction? What’s the ideal framework?

Andrew: I think way too many, even successful companies today are overweighting their core. They’re probably spending too much of their total budget on it, way too much of their managerial bandwidth on it. And at the risk of being a little bit cute, I think a core capability for most organizations going forward is going to be interfacing with the crowd, harnessing its energy and its abilities, and then finding out how to bring that back into the organization without setting off all kinds of antibodies, and resistance, and nonsense.

Erik: It’s part of the same lesson we learn from the mind machine trade off — is that defining the problem is important. Whether you define it for the machine, or whether you define it for the crowd, understanding what the problem is you’re really trying to solve. If you can define it well enough, then these contests work great. The contests don’t work great if you just say, “Hey, guys, you know, tell us stuff.” You give them a really precise…

Sonal: Which is what people used to do with the olden days. Remember when companies used to do these crowdsource and innovation boxes.

Erik: Yeah, and it never worked.

Sonal: Yeah, it never worked for a reason. So then that begs another question, though, for me, which is, if you take the innovation from the crowd, and you said earlier that there’s this escalating effect where everyone has access to the same tools, and they’re all catching up really fast with each other, and you can’t — it’s always the Red Queen. You have to run faster than everybody else. But if everyone has access to the same crowd, how does the company get advantage in this space?

Andrew: Then, honestly, it’s a matter of where your leadership throws its attention, how firmly you believe in these new kinds of energies out there. Not how willing you are to open the checkbook and spend money on technology, but how willing you are, forgive me, to open up your brains and rethink your business model in the face of this craziness.

Erik: Who can use these tools more effectively? Just like who can use the cloud more effectively? I mean, it’s a matter — it’s like what it always is. It’s just a set of weapons out there. And some people have a better strategy. Some people have better techniques.

Andrew: The companies that failed during the transition from steam power over to electric power, almost none of them failed because they refused to invest in electricity. That was not the failure mode. The failure mode was, they refused to rethink what a factory could be.

Sonal: And how to really absorb into the core of their business, yeah.

Andrew: And they refused to take seriously the idea of an overhead crane, or an assembly line, or a conveyor belt.

Sonal: Yeah. I’m just thinking about the statistics. When you said this thing about this antibody that organizations naturally have, which is essentially — they just immediately reject this not-invented-here syndrome, basically, a disease.

Andrew: Yeah. Look, and those antibodies are the best news possible for your industry.

Sonal: Research has shown over, and over, and over again, that it is practically impossible for big companies to absorb startups successfully unless they keep them isolated. And one of the questions I have is — the next follow up, basically, of what happens when you leverage this crowd. How do you then really bring them into the company so that you don’t have these antibodies? Do you have any concrete advice?

Andrew: I would look to do that in some of the most forward-thinking parts of the organization. As Erik said, in parts of the organization where the problem can be most clearly defined, and where you’ve got people at the helm of that part of the organization who are willing to take the innovation, the algorithm, whatever that the crowd comes up with, and slot that into the work of the organization.

Erik: There’s a role for the core to be able to define that.

Frank: In our world, a perfect example of the core leveraging the crowd is the classic enterprise software company. So, in the old days, basically, you wrote software, it was all proprietary. You won Gartner Magic Quadrant, then you sent your Rolex-wearing direct salesperson to go sell it to someone. The new enterprise company is, “Let me create an open source project. Let me get a lot of contributors. Let me get contributors to get downloads. And that’s my path to market.”

Sonal: Right. The open source becomes, like, the…

Frank: And the core needs to be there because they got — what’s the project? And what problem are we trying to solve? But the crowd comes into it to basically lend legitimacy, and support, and enthusiasm for the project.

Erik: So if you can be that scarce complement to the abundant crowd, you can create a lot of value. Then you become the linchpin that is capturing a lot of the value as well as creating it. Ultimately, we are an economy of creative destruction. And one of the strengths of the United States and other dynamic economies is that we have this constant turnover. And one of the things that discourages us is that there’s actually fewer startups, less innovation, fewer young firms in America today than there were 10 or 20 years ago.

Sonal: Oh, yeah. We talk about this phenomenon. That worries us too.

Erik: Absolutely. We are all for trying to make the bigger companies more nimble, understand this.

Andrew: Amen.

Erik: But the bigger way that the economy innovates is by having this innovative set of new startups that rise and adopt some of the new technologies. You got to have both. And we’d like to see progress on both dimensions.

Sonal: That’s great. Thank you for joining the “a16z Podcast.”

Andrew: Thanks for having us on. This is fantastic.

Erik: It’s a real pleasure.

• 

  Erik Brynjolfsson
• 

  Andrew McAfee
• 

  Frank Chen is an operating partner at a16z where he oversees the Talent x Opportunity Initiative. Prior to TxO, Frank ran the deal and research team at the firm.

  Follow Twitter
• 

  Sonal Chokshi is the editor in chief as well as podcast network showrunner. Prior to joining a16z 2014 to build the editorial operation, Sonal was a senior editor at WIRED, and before that in content at Xerox PARC.

  Follow Twitter

Here’s what we know about open source: Developers are the new buyers. Community matters. And there will never be another Red Hat (i.e., a successful “open core” business model … nor do we necessarily think there should be).

Yet open source is real, and it’s here to stay. So how then do companies build a viable business model on top of open source? And not only make money, but become a huge business, like the IBMs, Microsofts, Oracles, and SAPs of the world? The answer, argues James Watters, has more to do with good software strategy and smart enterprise sales/procurement tactics (including design and a service-like experience) than with open source per se — from riding a huge trend or architectural shift, to being less transactional and more an extension of your customer’s team.

Watters, who is the SVP of Product at Pivotal (part of VMWare and therefore also Dell-EMC), is a veteran of monetizing open source — from OpenSolaris (at Sun Microsystems) to Springsource (acquired by VMWare) to Pivotal Cloud Foundry — with plenty of failures, and successes, along the way. He shares those lessons learned in this episode of the a16z Podcast with Sonal Chokshi and general partner Martin Casado (who was co-founder and CTO of Nicira, later part of VMWare before joining Andreessen Horowitz). These lessons matter, especially as open source has become more of a requirement — and how large enterprises bet on big new trends.

Show Notes

• General strategies for open source companies [0:44] and differentiating in the marketplace [9:26]
• Dealing with competition from large firms [13:34]
• Why open source is growing [18:20] and a discussion of Red Hat, OpenSolaris, and other case studies [21:26]
• Advice for entrepreneurs on open vs. closed source [26:48]

Transcript

Sonal: Hi, everyone. Welcome to the “a16z Podcast.” I’m Sonal. We’ve talked quite a bit about open source in the podcast already, from the topics of open versus closed, to managing community and identity, to selling to developers. And a few years ago, partner Peter Levine put out a piece arguing why there would never be another Red Hat, which is one of the only open core business models to survive. But given the current and coming wave of companies built on top of open source, the tricky question left to discuss is, how do they make money? And joining us to have that conversation, we have James Watters, who’s the SVP of product at Pivotal, a cloud platform company that runs software in multiple clouds, and they’re part of VMware. And then also moderating this podcast, we have general partner Martin Casado, who himself came out of VMware, which had acquired the company he co-founded and was CTO at previously, Nicira. And he is the first voice you’ll hear.

Open source strategies

Martin: One of the paradoxes in this entire space is, there’s been a ton of money that’s been invested in open source, but almost no examples of successful companies built around it. Silicon Valley had a spidey sense that there was an opportunity there, like, nobody pulled it off. Yet now, there’s a number of examples of open source companies doing very well. James is one of the few people on the planet that’s cracked that, where they’ve figured out how to monetize and build a big business out of open source. And just to give a sense of how real this is, Pivotal Cloud Foundry went from 0 to $270 million in license, not support.

Sonal: I mean, what’s that mean, in license?

Martin: In license and in software sales.

Sonal: Oh, so it’s not professional services…

Martin: It’s not professional services, exactly.

Sonal: …because that’s the stuff that has like — that’s the thing we talk about all the time, in terms of building businesses, is that you don’t necessarily wanna rely on professional services because it doesn’t give you a lot of margin on your business.

Martin: Exactly. So this is, like, legitimate software sales.

Sonal: So, James, tell us how you went from 0 to 270 — million, we’re talking about, not seconds. How did you do it?

James: Yeah. And I think it’s fair to describe, kind of, my maturation and my thinking about this through failures, too. I worked on OpenSolaris at Sun, and then, for a while, at VMware, we were looking at how do you monetize SpringSource in and of itself. SpringSource is, like, the most popular Java programming framework in the world, by VMware 4, pretty famous, money over $400 million as an open source project. I had worked on that, you know, both of those projects, and had gotten my bumps and bruises along the way. And so I had some very particular opinions coming into doing the third round of PCF, Pivotal Cloud Foundry. I think there’s, kind of, two dimensions. One is the basics, which is, if you look at IBM and Oracle, Microsoft, SAP, etc., what they get right in this very basic thing is they understand how to cater to enterprises.

Sonal: Yeah.

James: And I think the number one temptation that most open source companies fall into is the first thing they do is they cater to their users.

Sonal: And by users, you mean the developers?

James: The developers that, you know, fork it on GitHub, start on GitHub. And the misassociation, as a first thing between the people that use it and then the people they need to cater the selling to, is kind of what you might call the first false horizon of open source monetization.

Sonal: Interesting.

James: What I mean by cater to is, like — I think this is the number one thing that I see open source companies maybe get wrong — is those are the people that show up and talk to you, those are the people that you’re interacting with. When you’re creating an open source community, you have to, first, get these unpaid users excited. But ultimately, IBM and Oracle are not out there mining their mailing lists of end users to go do deals.

Sonal: Yeah. How do you sort of straddle this? Because theoretically, you should be able to do both. But isn’t sort of the mantra, I should say it the Indian way, of open source, that you have to take care of your open source community and then also figure out how to become a business? Like, how does a company that’s trying to become a real big company straddle that?

James: It’s super important, right? I’m not saying that that’s not. What I’m saying is that your commercial strategy and your community strategy are not the same thing.

Sonal: That’s a great point.

James: It’s tempting when you put so much heart and soul effort into building that proof of validation around the community first, to say, “Oh, well, then, I’ll go upsell them.” And I remember meeting MongoDB when they were early in their rise, and they were selling, you know, a $4,000 support contract to thousands of users with a lot of expense. And when we built PCF, I was like, “I don’t think that’s the right way for us to go be big.” Just because you’re open source doesn’t mean you only sell to your open source users.

Sonal: So what did you do in getting across this first horizon?

James: The next thing you have to get right is you’ve got to get a major trend that affects enterprise buyers. So if you go back to enterprise as being the source of the money, you know, Oracle didn’t magically just exist one day. They caught the relational database market at the right time and built a franchise. And that relational database market really changed how enterprises were building applications, what they could do on applications. And in the same way, we’ve tried to catch the microservices trend as the major enterprise change that’s happening. That’s the kind of change that you need on your, you know, commercial strategy to generate CIO interest and enterprise purchases. It can’t just be a tool that someone’s potentially using [at] lower levels.

Martin: There’s been, [for a] long time, this thought that open source is primarily a commoditizing force. There’s an existing market, say, Unix. It’s an existing market with an existing buyer. Then you create an open-source version, which is differentiated by being open on open source, and then you go commoditizing that existing market. And you’re saying something quite a bit different. You’re actually saying that you can enter a new market using a shift and sell into that. So, I guess two questions. One, do you believe in the commoditization? Is that still a business plan? And then the second one is, like, is there any difference in open source to identifying these shifts, or is it just like if you’re doing a closed-source product?

James: Let’s look at history a little bit. I think Zen and OpenStack both felt like they could go commoditize VMware, right? They both said, “We’re gonna go commoditize VMware.” The last time I checked, Oracle and IBM did not get to be the size they are by having commoditized a previous generation of suppliers. So if you just look at the record — and this is what I call, like, the false horizon of open source strategy — you might be tempted that your only play is to be a commoditization play. But history tells us that the largest software companies in the world that get into the hundreds of billion in valuation, that’s not what they did. They caught megatrends.

Sonal: They built something new. They didn’t just commoditize something old.

James: They did. And I know that that sounds, like, obvious…

Sonal: I’m glad you’re pushing on it.

James: I’ve learned, over the years, that open source is now both a necessary part of almost any major software company strategy, because it’s a buy-in criteria that, you know, enterprises have for major new initiatives. But — that I think the open source got a little bit of a false start by only being a low-cost commoditization as a business model.

Martin: From my conversations, it’s almost somewhat of a contrarian view. Because traditionally, you’re like, “Open source projects chase after, basically, the sales dollars that have matured to market,” like MySQL, you know, like Android, like Linux, like JBoss. Every one of these you can point to a closed source incumbent that they were chasing after. So it’s nice to hear that you believe that open source is not somehow relegated to this commoditization. Because I agree with you, that basically limits the upsides you’re gonna get.

James: It does, because it limits your business model. It doesn’t let you invest. We have a $100 million a year R&D budget. When you’re chasing a new big strategic trend, and you’re getting the kind of checks that we are and the trust we are, you can build a big R&D team. If you’re chasing commoditization… 

Martin: Well, I mean, you are trapped to a fraction of the market you’re chasing after, by definition, right? Like, if it wasn’t a fraction of it, you wouldn’t be commoditizing it, and it’s very unlikely you’re gonna get it 100%. And so you necessarily, you know, have a ceiling over you. Well, it sounds to me that your recommendation is a lot like closed-source software sale. However, is it more than just, kind of, software sales, or is it just that, like, the industry is ready for open source now when it wasn’t before? Like, what has happened in the last two years to make this viable? Or is there more to the puzzle that you haven’t talked about?

James: And so the basics are strategy and segments. If your strategy and segment looks wildly different than every other big software company that’s existed before, that’s probably a pause. And then the second thing is, I think that cloud has started to affect open source, and we’ve taken a model of continuous delivery of our software, inclusive of cloud API. We don’t do what I call shipping you the tarball and the support contract and say, “Good luck.” We actually automate the continuous deployment and update of our software. And so we have an additional point of leverage other than just support, which is that, if anything doesn’t work in that large-scale update of thousands of nodes, you just blame us and call us. And that’s a different vector. That’s almost like a cloud vector in terms of value-add of software packaging.

Martin: That’s exactly right. We hear this all the time, which is, more and more, the customer wants to consume things as a service.

James: Right.

Martin: And this is often conflated with whether it’s deployed off-prem or on-prem. Like, to me, this is a total conflation, right? Like, whether or not it’s a service does not mean it’s necessarily deployed off-prem, right? These are two things. So one decision, off-prem or on-prem. Like, that’s actually a decision often, like, bound by regulatory compliance and security, etc. But there’s an entirely different decision, is — my consumption model as a service. That could be, I pay for it as a service, but also could be, somebody else basically manages it, does the update, manages the lifecycle, I don’t deal with the tarballs. That’s very much in line with what we’re seeing across the industry too. And what makes this discussion particularly relevant to open source is that it seems that once you’re talking about something as a service, questions around open source kinda diminish, because they’re not actually dealing with the code itself, they’re dealing with the service.

James: Big success stories sometimes have contrarian bets in them. And I would distil our two contrarian bets to, we did not go for the commoditization play, we went for the sea change and app design play.

Martin: That’s right.

James: And then we also tried [our] best to deliver a service-like experience even with software.

Differentiating in the marketplace

Sonal: Are there any challenges, though, in sort of bringing a polish to open source work? Because when I think of traditional software companies, they have baked in design for user — like, really client-facing versus developer-facing. So how do you sort of navigate that part of that, as building a real business on top of open source, if you don’t have that experience natively?

James: One of the decisions we made is we made our UIs closed source. So everything about the infrastructure of the platform is completely open source. And we chose to make the UI in that last mile of experience built off the APIs closed source.

Sonal: Yep. Okay.

James: So I do think there is some room to differentiate there, and you know, when you go to monetize, you’re going to need some small checkboxes. I’ve learned a lot, you know, growing this, about the importance of packaging for procurement. And one of the things I’ve observed is that procurement is exceptionally good at pricing a certain kind of thing, which is labor per hour. So, I’ve seen very large software contract orders, and procurement will actually pick on the labor per hour, because that has a nearer comparable. And so when you start to think about procurements looking to compare things to exact replicas to price it, having a little bit of, you know, closed-source UI or things that are maybe even immaterial to the product but are still there to say, “Oh, well, this is different,” is somewhat important.

Sonal: Interesting.

Martin: That’s really interesting.

James: And thinking about dynamics of navigating the politics of procurement are important.

Sonal: What are some of the other dynamics of navigating the policy of procurement, your lessons learned that you can share with us here?

James: I mean, you guys are asking for the goods now.

Sonal: Yeah, we are. We want the secret sauce. Give it to us for free. Share it with our listeners.

James: I mean, I think that is — generally, in an enterprise, when you get to procurement, somebody wants you to win, and you’re actually, then, in a political process to get through the last mile. Like, they’re kinda, like, the guardians there.

Sonal: You mean the internal champion type of person?

James: Internal champion wants you to win. Like, if you’re in procurement, they want you to win, and they wanna work with you. And this is why one of my other rules about open-source monetization is, try to avoid dogpiles. So a dogpile is a little bit, like, everyone’s doing X project, we’ve got to distro of X project. Think about the dynamics when you go into procurement then, and 20 people show up with an offer for X project. Like, no matter how much your champion loves you, the procurement officer is gonna have 10 comparables to compare you against. If you look at MongoDB, for instance, one of the things that protected them was that they are the only supplier of Mongo. So while they might not have had the high-end strategy right when they started, they at least were the sole supplier. So they could still somewhat dictate prices. If you look at OpenStack distros, I don’t know that anyone made it out alive [out] of that gunfight. 

Martin: This is such an important point. And I think it’s actually worth restating, which is, if you’ve originated a project and you’re bringing that to market, you should retain the ability to be the sole supplier, if only to set pricing and brand awareness in the market. Companies live or die by this type of thing. Another thing that I like to hear, if this is true for you, having spent a lot of time dealing with procurement — that more than anything else dictates the life of enterprise sales. Before going into procurement, I always expect procurement to need their pound of flesh. Like, this is how these guys are comped often. It’s like, “Okay, you know, whatever the discounting is.” And so, you know, we expect that going to procurement, you go — a bit of a Kabuki show, and then, you know, you end up with some pricing. And that pricing is actually for the vertical has been set in the market somehow. So it’s pretty well understood for forecasting by the business, pretty understood by the customer, and it’s, kind of, motions you have to go through. Do you find in open source that that discussion is different because you are open source? Like, do you have more pricing pressure because it’s open source? Do you have more of a push-forward service component? Or can I think about it the same way I think about closed source?

James: I think you can think about it the same way as closed source, if your model is, you’ve sold someone on a big new trend, you’ve helped train their organization to get there, and then you’re working on a fresh demand forecast of the transaction together. If you’re coming in and upselling support only, and they’re already installed, and all you’re upselling is a support model, that’s gonna be a lot more difficult. Because they already have it running, they already have it working at their scale, and all they want is, essentially, professional services by phone.

Competition with larger players

Sonal: What happens when a big company comes in and competes with you as a smaller, growing open source-based company?

James: This is something that open source companies have to navigate, which is that sometimes larger companies will adopt the same software by name, if you’re not careful about how you position yourself.

Sonal: Wait, what do you mean, like, they can literally just take your name even though you…

James: It’s very popular for large companies, say, IBM or somebody else, to say, “We support X.”

Sonal: Oh, I see what you mean. Okay, got it.

James: Larry Ellison famously did it with Larry Linux. He tried to do it to Red Hat. They often don’t have that much capability behind it, but they can, at least, push a little go-to-market on it.

Sonal: But wouldn’t your internal champion in the procurement office be able to tell the difference? Like, why would they even pick something else when they can get your product, which is what they want?

James: This is why having a unique market position is important regardless if you’re open source or not, because I’ve seen cases where very large orders were suddenly stopped because another large company just quoted something that sounded similar for the same amount. Now, that can happen open or closed, but if you’re in an open-source dogpile, as I mentioned, word of caution, it’s especially rampant. That can actually drive your entire upside of ever being a half-billion-dollar a year software company — like, your probabilities go down really hard.

Sonal: So how do you get out of it?

James: The key is keeping a unique offer in the market, right. You don’t wanna be yet another distro of, you know, a certain famous technology that provides support. I think you wanna have a fully packaged differentiated approach.

Martin: So here’s probably the most basic question in this space. If you have an open source project, it’s probably available online. So if you’re walking to somebody and you’re about to sign a $10-million ELA for license, what’s stopping them from just downloading and running it themselves?

Sonal: Yes, I have that same question. I’m glad you asked that. I wanna know the answer to that too, James.

James: You know, some of the origin of this discussion was, we were speaking on Twitter around this $100,000 or $200,000 contract value, and even some sub that, as sort of a valley of death. I think, if you go back to why the big software companies are, like, successful, they’re an extended part of those companies’ teams. So, I don’t really like the $100,000 a year relationship, because it’s very transactional. When you’re part of a big change like microservices in the enterprise, they’re gonna want advisory. They’re gonna want you there. They’re gonna want a team of two to three people that just live there. A key thing is when that $10-million transaction comes, they’re as much voting with, you know — they want you to be part of their extended team, part of their strategy, part of their relationship with you.

Sonal: Yeah. 

James: Because you have an expertise that’s been hard-won that they do not have.

Martin: So I spent, you know, quite a while doing enterprise sales, and I’ve got this view, it caused this conversation that you and I had on Twitter, which I had claimed that, you know, in enterprise software sales, there seems to be this valley of death between, like, say, 30k and 150k. If you’re 20k or below, you know, you can call somebody up, and they can pay for it. And let’s say if you’re 200k and above, then you have hopes of supporting a direct salesforce and still have good margins. And then, in between that, you’ve got this valley of death, where you can’t really support a direct sale because you can’t pay the people. But, you know, if you’re trying to do something new and innovative, you don’t have account control and it becomes very transactional. So, do the dynamics of that change with open source sales?

James: I think what’s happened is that open sources become how large enterprises wanna bet on big new trends. And then that opens the capability for the first time of open source companies — I’ll use the word — being high end. Meaning that IBM and Oracle are parking a bunch of good platform architects, as we call them, technical architects, at the account that explain the new things to that account. If you wanna get a $10-million relationship going, you’ve got to become that extended part of their team. That is why I jumped in on our original Twitter discussion around the $100,000 a year deal, and I was like, “Hey, actually, I think 100 is too low. Like, I would try to get to 400 to 1.5.” And the critical thing that happens at that [amount] is not that you extract more money. It’s actually that you can be a better advisor. Like, you can actually put talent on the ground, and you’ll find that these large enterprises, during times of big change, really value that talent on the ground.

Martin: Interesting. So, this is a nuance, a new way of looking at the discussion. So, how much does your ACV have to be to support a salesforce and say, “It has to be at least 150, 200k?” You would say that it actually should be higher than that, because the goal isn’t just the margins on a direct salesforce. The goal is really strategic account control.

James: Correct.

Martin: And to get that, you need a deeper engagement. That’s very interesting.

James: Yeah. And I think so many open source companies have died because they’d transact around the edges with, like, director level or technician level for support, and they never got to the CTO or the CIO and true strategic account control like the big players had.

Why open source is expanding

Martin: So, I’d love to get back to this question of, like, why now? Like, I’m just so curious. Which is, like, is it, like, have you figured out something that nobody else has, or is it that the enterprises now realize that they need to pay for this software, or is it something else going on? Like, why are we starting to see the Elasticsearches? Why are we starting to see the Mesospheres? Why are we starting to see these companies be successful?

James: Well, if you look at the clock cycle of how often big new software companies get built, it’s not every day. Enterprise architectures only change so often, right? Like, if you were to try to sell a middleware offer in 2008, I don’t think you could have raised money. You would have had a really hard time, because none of the design patterns were changing. I think what’s happening now is that cloud is disrupting a lot of people’s approach to software infrastructure, to how they build applications. And so, this crop of open source companies is getting a chance to be the leader of a new thought, versus purely being a commoditization play.

Martin: And so, you’re saying that it’s not inherent in open source, as I see inherent in the trend that’s going on, and open source is just becoming, basically, a requirement because of customer expectations in community.

Sonal: But I have to push back on this one. Isn’t that inherent in open source, because that has to do with the nature of a community, the speed of development, the fact that you don’t have to necessarily go through a waterfall type of development process? I mean, isn’t there something here that’s inherent to open source, by definition?

Martin: Fixing problems is fixing problems, whether you do it in open source or a closed source. And the reason that we’re seeing this rise in open source companies is not endemic to open source. It’s the fact that we’re seeing, actually, a transformation around developers and their aesthetic and the technologies.

Sonal: Totally. I totally buy all of that.

Martin: And then the question is, well, where does open source come into the play with it? It seems like there are intrinsic benefits to it, and there’s a level of expectation from the customer.

James: Yeah. And I think that the commodifying open source companies did a really good job of normalizing open source as something that people did, and even creating an expectation that that’s the right way of going. And then, this wave of what you might call, like, the microservices generation, or the cloud generation open source companies, are actually catching enterprises with a new need, and a new, you might call it, hundreds of billions of dollar need. And suddenly, IBM and Oracle’s earnings are missing every time, and suddenly, these new companies are getting green shoots.

Sonal: Okay, that’s fair. Just wanted to make sure that there wasn’t some inherent experimentation baked into an open source-based project that then leads to this more innovative type of…

James: You’ll keep getting me to talk open source strategy, I’ll keep talking software strategy.

Sonal: That’s great.

Martin: I love it.

Sonal: That’s great. That’s, like, what we need to hear.

James: Sometimes people say, “Oh, well, why is enterprise software so expensive?” It’s expensive because, you know, the trust model of enterprise buyers is, you’re really gonna take care of them, advise them, ensure outcomes. You’re not just providing support. So what you’re really fighting for is the chance to be one of those new trusted architectural advisors, I believe.

Sonal: Red Hat lived on the x86 shift.

James: Red Hat kinda got lucky and found a commoditizing change, which was Unix to Linux. That’s a rare event to have a major, you know, chip system re-platforming sea change.

Martin: So, let’s talk about Red Hat, because Red Hat seems to not follow — there’s a big transformation. Open source is a great way to do it because it gives you the edge with community, all the things, you know, all the bromides that we normally talk about. So, you take advantage of the transformation, you become a strategic advisor, etc. But if you look in the back, that’s kind of, like, the one success case doesn’t really follow that.

James: For all of the time it’s been in market, Red Hat is still a $2-billion a year software company, right? If you compare it to what IBM, or Oracle, or SAP, or Microsoft, or any of the megas, they still haven’t caught that size of a trend, even though their prices, frankly, are similar to Microsoft’s now, like $2,500 an OS. So it’s not that they’re just down-market, it’s [that] they just didn’t catch a trend big enough to become a company that big.

Sonal: Yeah.

Martin: I see.

James: Like, I don’t think they — you know, in the same way that the mainframe captured all of the world’s transactional processing, or Oracle captured all relational database apps and enterprises.

Sonal: I mean, you’ve described this now a few times in this podcast, catching this trend, this wave, this need. That’s just product-market fit. Like, there’s a real market need for this.

James: Yeah. My thesis is that, just saying “We’re an open source company” can distract you from software strategy.

Sonal: Yeah, exactly. That’s great.

Martin: So, what did OpenSolaris get wrong? I mean, like, you’ve seen this from many angles. Oh, there’s a pained look on his face. I’m not sure if I’ve kicked a wound.

James: You know, it was a great learning experience for me because I think Jonathan Schwartz had a fairly early and simplistic open source strategy. And he was one of the first CEOs of a mega enterprise company to do this. So I watched this firsthand. And he was like, “We’re an open source company.” The problem was that Sun was still on the wrong side of all the trends. Like, it was actually late to x86. It was late to everything that was happening around Webscale. And so, open source actually only hurt us, in a sense, because, you know, working on monetizing OpenSolaris, one day, Jonathan said, “It’s all free. It’s no longer licensed. Go sell support contracts. Everyone will buy support.” But what happened was that, overnight, like hundreds of millions of dollars of revenue evaporated, and then it was a tactical challenge to go get people to sign up for it again. We didn’t have some strategic initiative to go talk to them about.

Sonal: Which goes right to building a strong, good software company, to your original thesis.

James: So we worked on that. And then the next thing I saw was, within the Spring group, there was something called tc Server, which is just cheaper middleware. And it only got to, say, tens of millions of dollars in sales and an average selling price of, say, 80,000, somewhere in that range, and it never caught a strategic new design point either. So when we had PCF, I really said, “Hey, this microservices refractory is actually a once-in-a-generation change. We could take a different tactic.”

Martin: What I really like about your view on this conversation is, you’re basically saying, open source or not open source, you have to build a solution that solves a real problem, take advantage of a trend. It’s like a traditional software company and software sales, and then open source provides uplift, and etc. And often, open source has been viewed very differently, like, it’s got something magic that will save you. And so what you would see, is you’d see these incumbents that are desperate, and, like, in an act of desperation, the swan song is to release something open source, because they think that, somehow, that’s gonna magically save them.

James: Community development is really important, but if you just develop a little community and it doesn’t change CIO, CTO — anyone that’s writing <inaudible> checks priorities, like, by definition, you didn’t build a software company that had an opportunity to be the big new strategic advisor and…

Martin: That’s a critical piece

James: …new architecture.

Sonal: Right.

Martin: Another thing that I like about your view on all this — it seems like it’s a very, kind of, planned, top-down view. Which, often, discussions around open source, to me, are very organic, which is, you know, you get a community, and that community will grow. And like, this is kind of very, kind of, like, organic progressive thing, and then you have some…

Sonal: Well, this goes to my question about product vision, like, having someone to direct it top-down. But what I’m hearing is, the next Steve Jobs will be the Steve Jobs of an open source company, because he’s a great CEO, great product guy, great business. It doesn’t have to be open source or closed source. To your point, it’s building a great software company.

Martin: Exactly right, which is like, listen, there’s an industry trend, you’re gonna build real value-add in that trend, and you’re gonna go from the top-down, make it a direct salesforce. And this is really about kind of top-down, like, strategy and planning, and not just, kind of, progressive community…

James: A great thing that’s happened is because buying open source has become a normative behavior.

Martin: Yeah, exactly.

James: Which means that you don’t have to be shy about doing something audacious and asking for money.

Martin: Push that even further, do you think open source has become a requirement, or is becoming a requirement?

James: It’s close to a requirement, I think, for major new bets. Like, if you look at our biggest buyers, they don’t wanna get into what Oracle’s doing to them right now. Like, every year, they’re using less Oracle, and every year, they’re paying as much. That’s a very negative feedback loop.

Sonal: Wait, explain why.

James: ELA structuring, enterprise license agreements.

Sonal: And what is that? Like, just kind of talk us through it.

James: So what buyers are trying to escape from, and one reason they want open source as a protective measure, even if they don’t use it, is the right to keep using their software without paying more and more for it every year. Companies like Oracle have been out there charging more and more for the same usage over the last few years. So, they basically trap you into that and say, “Well, we can increase your unit cost even if you’re decreasing your unit consumption.”

Advice for entrepreneurs

Martin: Yep, that’s great. To get back this original question, which is, a lot of the constituents that listen to this podcast are entrepreneurs or aspiring entrepreneurs or existing entrepreneurs. So let’s say one of these entrepreneurs wants to create a company, Acme, Inc., and they have…

Sonal: Acme, Inc., great naming there.

Martin: Acme, Inc., that’s right, because I’m creative. So they’re creating Acme, Inc., they’ve got a software project they think is gonna change the world, and they have a decision to do open source or closed source. Is there a difference, or is it the exact same thing?

James: I think we got incredible lift off of open source. For instance, IBM showed up and standardized, then, what we were doing, HP, SAP, other people. We got huge lift off of open source. Now, that doesn’t mean that you have to accept the usual low-end tarball and support contract business model. So I would probably not start a closed-source software company today, unless I had a very niche, novel idea that I felt no one else would try to do. Because quickly, there’ll be an open source alternative to what you’re trying to do if you do not open source it.

Sonal: Okay. Last question then, you started off describing how you’ve been through, like, two, sort of, failures before you kind of struck the right model for building a great software company, not necessarily a great open source company. What advice would you give to entrepreneurs today who are trying to do the same thing or the next thing? 

James: I think the tradeoff that you’re gonna make as an entrepreneur is to come out into an existing trend with a lot of people having already validated that. That was easy to get into that trend. It wasn’t a big risk. So, I think the challenge for entrepreneurs is trading off, like, a vision of change that will happen over the next couple of years that you’re gonna grow into, versus just joining an existing parade around a standard open-source community.

Martin: Yeah. And I would add to that, things that you could tangentially apply to previous architectures don’t necessarily carry forward, right? I mean, like, often, we’re like, “Oh, this worked for VMs, therefore, it works for containers.” And so I do think that…

Sonal: Oh, interesting.

Martin: …you really need to be piped in the nervous system of the evolution of the market, and you really need to be bold about, kind of, going after new solutions that are part of the evolving landscape.

Sonal: I mean, it’s a first principles way of thinking.

Martin: Exactly.

Sonal: It’s sort of like saying, “Don’t derive it from what happened before. Like, think about it from scratch.”

James: Martin said one of the most brilliant things, which I was sitting at home, listening to the podcast, cheering on, which he said, there’s so much metadata and microservice now that you can change the way you think about networks. That is the fundamental shift, that if you catch something like that, then you have a game-changer.

Sonal: That’s wonderful. Thank you for joining the “a16z Podcast,” James.

Martin: That was great.

• 

  Sonal Chokshi is the editor in chief as well as podcast network showrunner. Prior to joining a16z 2014 to build the editorial operation, Sonal was a senior editor at WIRED, and before that in content at Xerox PARC.

  Follow Twitter
• 

  James Watters
• 

  Martin Casado is a general partner at a16z where he invests in enterprise companies. Prior, he was cofounder and CTO of Nicira (acquired by VMware) and is the creator of the software defined networking movement.

  Follow Twitter

Your brand, says head of a16z marketing and Outcast Agency co-founder Margit Wennmachers, is what people say about you when you’re not in the room. And it’s going to happen, whether you choose to have an active part in it or not. But what does this mean at an individual, not just company/product level?

In this episode of the a16z Podcast, Wennmachers and Outcast CEO Alex Constantinople — both longtime veterans of public relations and building executive profiles — de-mystify what having and building a personal brand takes. It’s not only about “thought leadership”, either… a personal brand can also provide a filter for choosing what to do (and what not to do), as well as define your aspirations for where you want to go next. Even if you cringe at the idea of putting yourself in the spotlight.

This conversation, moderated by a16z partner Hanne Tidnam, was recorded as part of the BreakLine Tech program for military veterans, an immersive education program for veterans transitioning into new careers (including a week of talks and courses hosted at Andreessen Horowitz, some of which can be caught here).

Show Notes

• Defining what a personal brand is and how to get started [0:30]
• The importance of storytelling and tension [5:20] and why everyone needs a personal brand [7:19]
• Examples of effective personal branding and the importance of authenticity [10:38]
• Logistics around creating a brand [13:55] and how to know when it’s working [17:12]
• Managing mistakes [19:29] and audience Q&A [23:45]

Transcript

Hanne: Hi, I’m Hanne, and welcome to the “a16z podcast.” The phrase “personal brand” is something of a cliché, but we all know we’re supposed to have one. So what does it really mean, and how do you go about actually creating one? In this episode, a16z’s Margit Wennmachers and Alex Constantinople, CEO of The OutCast Agency, both break it down into basics and also give us a sense of nuance on how best to think of a personal brand. This podcast was recorded as part of the BreakLine tech program for military veterans.

Defining “personal brand”

So, I just thought we would start with a really basic question, just to lay a little groundwork, which is, what do you think a personal brand actually is? How would you define it?

Margit: I think, in a nutshell, it’s basically what people think or say about you when you’re not in the room. That’s how you should think about your brand. What is your reputation? What is the association that you occupy in someone’s mind? And so, that’s in a nutshell what it is. If you think of — companies are easier than people. If you think of Apple, you probably think of design and elegant products. If you think of Virgin, you probably think of irreverent and fun. Like, those are the brand attributes that you think of, not even consciously, necessarily, and that really what defines a brand.

Alex: Yeah, I mean, the good news is, just being conscious about it actually will help you. So, I think, while it is a huge part of what others say about you, I do think it’s what you choose to put out into the world as much — and is actually more important.

Hanne: So let’s say you’re starting totally from scratch. You know who you are, you know what you’ve done, you know what your resume says, but how do you go about — step one to finding what your personal brand is?

Alex: What we usually do is we’ll have an executive come in and really just do a whiteboard session. And we really start with, “If I talk to your neighbor, if I talk to your parents, or your partners, or your best friends, or your co-workers, what would they say about you?” And we find that’s an easier entry point than if I say, “Give me adjectives,” it feels weird. Like, “I am the smartest, the prettiest, the most fabulous…” Like, it’s harder to get it out of people because it’s awkward, right, to be, like, “This is who I am.” But this is how we usually just really start.

And then the next question, the most important one probably, as you’re thinking for yourselves, is also, “What kind of leader am I? What is it that I want to put out into the world and have people see? How do I want to be?” And this can be aspirational, all of this, by the way. It might not be who you are today. You might feel like, “You know, I’ve gotten feedback.” I mean, my 360s that I’ve gotten from GE, past to now, are hilariously the same. It’s sad. I’m not happy about it, but there are some feedback points in there that I’m like, “I can’t get rid of that,” you know?

And so, I think being conscious about it, like — what are some of those things, the way I lead, that I want to be seen and then how do I get there — which is part B of this. And then really your expertise is a big part. How do you want to be seen out in the world? This is where more of you heard the phrase “thought leadership.” When you’re out more in the external world, what do you want to be known for? What are you really, really good at? What can you own as an expert? And then that can be subject matter. It can be super broad or very, very narrow and all of the above.

And then the last for personal brand, I think, is really everything about you, because I find you can’t leave your personal stuff at home. You can’t leave that you might love the outdoors and you’re more the adventurer. It helps round up the picture. You love to read. You love to be with your family. Like, that is you. And if you come to a job, I find, without your full self, you can’t have the most value. And so we don’t leave that off as soft stuff. It’s really important that you are authentically you.

Hanne: Another way of getting at this is thinking about story. When you’re saying all that, I’m thinking, like, “Well, that’s so much information.” I mean, how do you know what the story is that pulls it all together? What’s a good way of thinking about that?

Margit: I think, to Alex’s point earlier, that is where you have a fair amount of control, right? Like, what are the anecdotes that you want to share, right? Like, what’s the part in your childhood that shaped you that made you join forces, or that made you the leader that you are. Like, you can control all of those anecdotes. If you think of, like, a very carefully crafted brand — whatever you think of the person or not, this woman Sheryl Sandberg — you all know who she is, right? Well, if you hear her speak, or if you read her book, or if you see her on TV, there’s always a story about when they were kids, she was managing her siblings, right? She put that out there, right?

So, a good way to get at what your version could be is, if you read — take any of your favorite magazines and read a personal profile that someone has written about a business leader. It’s probably the most relevant example, or an athlete, or whatever — and, sort of, look at, “Okay, what would my version of that be? How would I fill out paragraphs 1, 3, 7, right?” And you see, once you dissect an article, it becomes not as black box voodoo-ish as it seems when you first think about brand, right? You go, like, “Oh, they have their family interests. They have their childhood experience. They have their expertise.” And you can deconstruct the story. And then if you take one of those articles, go like, “Okay, if I had to write a story about me, or if I wanted a story written about me, what would be in that story?” And that gives you a control over what it is, and it also helps you build the body of how you talk about yourself.

Hanne: So are there things, though, that you think universally make a good story? You know, that you look for when you’re helping people do this, characteristics that you say, that’s…

Margit: So, I think it depends. We haven’t even talked about — we’ve talked about what is your brand, how do you want to describe yourself. We haven’t talked about, like, how you put it out in the world, right? So that’s a whole…

Hanne: Right, mm-hmm, which we want to come to next.

Margit: Which we want to come to. But when you think of stories that other people tell about you, like a magazine article or something, they always want some tension. And that’s fine, as long as I think there’s a happy ending at the end, right? And tension can be anything from a tough childhood or a really tough mission, to the extent that you can talk about it, or, you know, what countries were you deployed in and whatnot, right? But, like, they all want some tension. They want the reader to go along and go like, “Okay, I want to read the next thing,” right? It can’t just be like, “Here’s my picture-perfect resume, and, like, yay.” Nobody wants to read that, right? Like, we don’t even want to read that.

Alex: I was going to say like, if there are lessons learned — people really like what can you bring to somebody else. And so, I love people who put themselves out there and are a little more vulnerable. And I know that is hard. So being able to say, “I tried this,” or, “This was something I did that didn’t work, but I learned x from this,” I think is a great way to think about that particular tension.

Margit: Just last night I was reading your profiles. I was like, “One, we shouldn’t be talking. You should be talking.” <Exactly.> But there are great stories in there, right? Like, one of you tried to land one of those planes, and it didn’t quite work out.

Alex: Raise your hand if you tried to land a plane that didn’t work.

Margit: That’s amazing.

Man 1: <inaudible>

Margit: You’re here, right? But I was reading that and, like — look, it stuck with me. So there are amazing stories. I thought these profiles were really, really interesting to read, and there’s a lot of good stuff in there, where I thought, like, “Oh, my God, they have a lot to work with.” And also, thank you for all the stuff that you’ve done.

Why everyone needs branding

Hanne: So, maybe this is totally obvious, but does everybody need a personal brand then?

Margit: Yes. Soapbox moment.

Hanne: Even if you’re interested in a job where, say, you don’t want to put your opinions out there that much.

Alex: I do this with college graduates. Anybody who needs to tell — to be — they’re going out into the world doing something, I feel, like, “Shit.”

Margit: If you have interactions with people, you need to think about this. And if you think of the startup world, here in Silicon Valley, most of them — they languish in obscurity. You do want to stand for something. You want to be remembered for something, you know, as much as there are a gazillion jobs out here, right? Like, everybody needs to go, like, “Okay, I want this person, because they struck me as such, such, and such and such.” And a brand doesn’t have — it doesn’t mean fame. I think people confuse brand with fame. If you have a powerful brand with the right 20 people, that may be it.

Alex: Yeah, that’s exactly right.

Margit: It doesn’t have to be fame. It’s not like — you don’t have to be on CNN or whatever. Like, no one’s saying that, but you just want to have a deliberate way of thinking about, “Okay, how do I want people to think of me?”

Alex: I wish there was another name for it. I think, especially here in the Valley, personal brand, I found it was — I spent the majority of my time in New York and D.C. before I came out here, and personal brand, it was no big deal. Here, it’s sort of like, “Oh, I don’t want one of those. Like, that sounds too much for me.” The personal brand can turn people off because they — certainly if they came up through the technology and the engineering side, it’s very uncomfortable. It’s like, “I’ll go out there. I’ll only do my Fortune story or whatever…”

Margit: They feel like they’re lying.

Alex: “…if it’s good for the company, but I’m really not interested.” And they’re kind of missing the point, because, to Margit’s point, it’s not necessarily the next step — which [it] can be for people — then what’s the communications plan against your personal brand. And that can be, “Okay, start talking at these places and giving these kinds of speeches, and let’s work toward, you know, this kind of profile and this kind of publication that will help your business grow.” But I think for purposes of just any executive — and I’ve literally done them for professors, for, you know, scientists, people who you wouldn’t think this would be — but it really can work with anybody because it’s just a way to frame your activities.

What we also say to a lot of executives, no matter what level — it really also helps you as a filter for your time. If the activities you’re doing — and you’re going to get asked to do a lot of things, right. If it doesn’t necessarily fit within what you’ve laid out for yourself, then I think it’s easier to say no. It’s like, “Right now, I’m focused on this.” You know, obviously, everyone has a favor for a friend that they’ll do, but I think it will help you focus and save your time. So that’s another really practical reason that we try to put this down. Now, it doesn’t mean it won’t evolve over time. You can look at it again in five years’ time, in two years’ time, in a year, and say, “Okay, now that I’ve been doing this…”

Margit: Achieved that.

Alex: Yeah. Is this good? Does this feel right? Because I think we do want to push ourselves all to be aspirational.

Margit: The other side of the coin is, like, brand happens to you, whether you want to or not. Like, people will describe you in their heads. So, would you rather have some say in what that is, or do you just want it to kind of let it happen, right? So it just happens. I mean, think of business executives that you admire, or hate, or whatever. Like, you have opinions about them, right? And so would you rather shape how people perceive you, and have it be true to yourself and what you want it to be, or just have it happen, right? So just take control, like you always do with everything.

Hanne: Are there some examples of people — I like that you distinguished between fame and brand — that you think, maybe, are not on the famous side, but did such a good job telling their story and establishing a brand?

Alex: You mentioned Virgin already. I think Richard Branson.

Together: Yeah.

Alex: I do think he’s done quite a good job. The companies he’s built absolutely are from him. You don’t feel a disconnect. He doesn’t say things that then don’t show up in his companies.

Margit: By the way, it’s the only business, I think, that has a brand that’s consistent across very different businesses.

Alex: Yeah, yeah. He’s a horizontal growth, so it’s like, you know, mobile to hotels to airlines to…

Margit: Trends. Everything.

Alex: Yeah, it’s crazy, but it’s — that thread through works, and then the way he jumps out of planes and, you know, doesn’t have insurance or whatever the hell, like, seems to work for — you know, it absolutely works. It absolutely works.

Margit: Another person I think who’s done consistently a good job of their brand is Warren Buffett. He’s just very authentic. He’s got that folksy style, but he’s also smart. You know, he shows up consistently, and I think he’s done a really lovely job of managing his brand. And I’m not even sure he’s consciously doing it. Maybe he’s just lucky and gifted. Sometimes people are more gifted than others. Another person who actually doesn’t even tell her story, but I think has done a good job, is Angela Merkel over in Germany. I think she’s just like no-nonsense, right? It’s not a flashy brand. It’s not a “let me use my feminine charm” brand at all. It’s just, like, you know, walks the line.

Hanne: She is who she is.

Margit: She is who she is, and she’s just like boom. She keeps marching, right? So I think she’s done a good job. And then the example of someone whose brand has changed a lot for the better will be Bill Gates. If you — you know, some of you are too young but, like, he was just, particularly in Silicon Valley but I think widely, hated because of their hardcore business behavior. And now he is one of the most admired — and rightly so, one of the most admired human beings. Now, it’s easy when you have that much money to throw at the problem, but still, a lot of people have quite a bit of money and don’t bother to try and improve the world. So I think he’s done a really good…

Alex: That’s a good one.

Hanne: That makes me think of — you mentioned authenticity and, like, the role that authenticity plays. I mean, how do you avoid feeling overproduced or over…

Margit: I think it starts with, if you’re trying to portray something that you truly are not. So let’s just say you are hardcore competitive. Then don’t try and make your brand be, like, “I’m a little puppy dog,” right? It’s just, like, not going to work. Just own who you are, right? And I’m sure there’s an okay version of who you are, and own that, right? So that’s step one, authenticity.

And then Facebook, and Twitter, and Snapchat, and whatever else, they just demand authenticity, because it’s so easily detectable if it’s someone else doing the writing, or if these photos are too curated. From a content point of view, make it who you are. Like, everybody has features and bugs, in Silicon Valley parlance — like, “I have a lot of bugs, but you’ve got to find the place where the features are valued,” right, and you’re going to be successful in those jobs and not in others, right?

Branding logistics

Hanne: Okay, so let’s talk about logistics a little bit, and platforms. You’ve made your list of adjectives. You know, you’ve figured out you, sort of — how do you actually go about getting it out there? And, like, are all platforms — do you have to be on all platforms all the time and…

Margit: Well, I mean, that’s an entire book of a conversation, but to start with, let’s just assume you’ve done not just the adjectives but also, like, “Here’s my story,” right? Here’s the biography that is not, sort of, your official resume that you send out in the world. I would start with — if it’s something you love — if you’ve done a lot of speaking as part of your work, and if it’s something you love, like, go to town. Try to get the TED Talk. But don’t try to get the TED Talk if it’s not something you really love, because the worst thing that you can do is just, sort of, do a very high-profile thing and then just fail at it miserably. A, it doesn’t feel good. And then, B, it just don’t you any favors. So I would start, if it’s something that’s totally natural to you, I would start with something really small and comfortable. I don’t know, it could be your alumni newsletter. It could be very, very small.

And then, the other thing I would say is, not every medium is for everyone. So I’ll use Marc Andreessen as an example. If I do Q&A, he’s brilliant. He’s just very good at the repartee, the question and answers, being quick on your feet, getting to the heart of the matter. He talks fast, and the whole thing works, right? Just find what you are. If you are good at speaking, speak. If you’re good at writing, write. Now, if you’re good at speaking, you still need to write, because you want to make sure that what you say is, like, really deliberate and whatnot.

But, like, everybody is different. There are things like LinkedIn and Medium, right, where you can share things like what Alex was saying, like, lessons learned or tips or, you know, like those kinds of things. Then obviously there’s press, which is the least controllable, because whatever you say, it goes through their filter and, like, they end up what gets used, and how, and all of that. So I’m sure you know the pitfalls. But it also is, in some ways, the most credible, because it’s not just you doing your own talking but a third-party, and they have their readership, and whatnot. But there are all kinds of options.

Alex: I will bring it back to — in case none of that is where you are, right, it’s actually — within a job, it’s what activities are you doing and what are those — are you doing the kind of work that you want to be doing? Are there projects that you want to be on? I think there’s also ways to use this for your advantage within a company, or within your environment, and maybe it is more community advocate as well on the side, and then what are you doing to do that, right? You know, do you want to sign up for something? Do you want to participate in a non-profit? Like, whatever those other things are, that also can be included. So I think there’s a quieter way, also, to think about the execution of a personal brand exercise that can be — how do you show up wherever you are?

Margit: Just to add another thing. It can also be, like, maybe you want to create your own personal network. Let’s just say you’re here, you have a job. It could be just, like, you corral a bunch of people and you have dinners. It shapes — as Alex is saying so eloquently, it shapes your activities and also what you say, and what you focus on, and what you want to impart.

Hanne: How do you know when it’s working? I mean, is it followers? Is it, like, getting places published? When do you know, like, “I’m telling the right story”?

Margit: I think it’s —companies spend millions of dollars doing brand studies, and they’ll do things like sentiment analysis, and Twitter followers, and all that kind of stuff. I think you know when it’s working, and I think you would know when it’s not. And it sounds like a pat answer, and maybe Alex can help me refine it, but are you working in the right job? Is that fulfilling? Do you feel like you’re connecting well with people? Are you spending your time on activities that you enjoy? Do you feel like your expertise is valued? To me, it’s like, are you working on something that you think is important, that’s larger than just yourself? And do you feel like you play a meaningful role in it? If you keep running into trouble, or if you keep not interacting well with people, then, yes, then I think it’s time to revisit it and go like, “Okay, what’s not working here?”

Alex: I actually wrote down three things off of mine, that I wanted to use as a temperature check, which I keep looking at. So, you could have your own version of this but mine was, “Am I growing and developing?” So, actually, one of the reasons I took this job is that at first I was like, “No thanks,” and it was just, like, pre-“Lean In” territory. And I was like, “I can’t.” I just had my third kid. It was a surprise. Like, oh, my God, and then running a company? I don’t think so. I’ve never been trained for that. And then I was like, “No, this is — this matches growth and development. I’m going to push myself. I’m going to throw up probably every day, but that’s okay.” I cried a lot and, like, did “St. Elmo’s Fire” with the curtains a lot. Totally true for the first year, but I’m over it now.

Growth and development was one. Adding value was a big one because that is — and I have two versions of that, which is — am I able to do what I do best at the job I’m in? Am I bringing everything? Are they accepting what I’m giving, basically? And that was also the personal part, and I actually think I’m successful because it’s all of me in that whole brand platform page.

And the last one is just the fun, is my thing. You may have another one, but for me, I think, at almost 48, [I’ve] just been like, you know what? I am not working with people even on your crappiest day — I was going to say shittiest, but I’m trying to work on — on your crappiest day, that you can’t have a little bit of a laugh, or be like, “What the F is happening?” you know or whatever. And you just have to have that. So that’s my thing. So, you will have your own things, but I think that’s another way of thinking about it in the frame that you asked.

How to handle mistakes

Hanne: So what if you mess up, on a less happy note? What if you put something out there, and then you’re like, “Whoops, that totally doesn’t feel like me,” or you get a bad reaction? What then?

Alex: You have a famous phrase, “Never waste a crisis.”

Margit: Yes, never waste a crisis. It’s my way of coping. So, there’s a company version of this. If you mess up, like, how do you handle yourself, right? What are you doing? Because there are no secrets. We all know this, right? In theory, we all know this, and then we try to forget it when it applies to us, but there are no secrets — particularly not if you’ve tweeted something. It’s just, like, own it. Own it and move on. Just own it.

Alex: Interesting enough, with the PwC thing from the Oscars, right? And all the coverage was, they are taking it on the chin big time. The chairman actually came out and quoted about it. I always appreciate, and I’m sure you do as a regular consumer — think of brands that are messed up, whether it’s a food brand and something happened, or, you know, just saying, “We did this. We’re sorry. Here’s what we’re going to do to fix it.”

Margit: And then actually do it.

Alex: Yeah, I mean, you probably already tell your kids that. Like, just own it and say you made the mistake. I mean, you’re only going to get as much trouble if you make eight lies and make me hunt you down.

Margit: The thing about the human condition, we want to forgive. We just want to feel heard. We want to feel heard, and then we’re ready to forgive. But if you’re lying to us…

Alex: You cannot get that.

Margit: …then we get very needley and obsessed and whatnot.

Hanne: In your own experience in building your own, you know, personal brand, what do you felt like was the hardest, or what was the most challenging for you?

Alex: I would say the hardest and most rewarding was coming out here and not knowing anybody. I mean, my whole network was a completely different network, and I moved here, much like I think you guys are. And that was just hard, because it was, sort of, this blank slate of like, “Nobody knows me.” So it’s kind of awesome.

And what do I want to be? So, starting over and making that transition, I think, can be very challenging but incredibly rewarding and you just have to be patient. The best times, coming out of that, I was just extremely thoughtful, and I’ve never made a “mistake” in my career yet so far. I did a lot of due diligence. I really thought about, you know, what kind of company do I want to work for, what brand, you know, is it. Like, how will my story — I’m not a planner so I’m not, “My 5-year plan, and my 10-year plan, and I’m going to be this, and I’m definitely not going to run for president in 2034, and holy hell.” But, I do think I’ve been along the way — to, sort of, combat that scaredness about it — just trying to be really thoughtful, and not rushing a decision, or not rushing into it, and not looking at a whole company, or not looking at the people that I’m going to be working with, and the kind of work, and can I be successful.

Margit: So mine was — when I was running OutCast, we sort of had made a decision — it is going to be all about the clients, and we are not going to be out there and vocal. Maybe that was my excuse for not doing anything. But, like, my belief was, you don’t ever want to be in the news and have your client go and like, “What is she spending her time on while I’m paying?” Like, that just didn’t sit right, so I kept a very low profile. I basically did nothing. And then when I joined here, Marc, essentially, sort of, challenged me. He didn’t force me. He said, like, “I would highly encourage…” He’s very convincing. “I would highly encourage you to, like, up your brand profile a little bit.” And that was really weird. Like, it was so ironic, right? I’m sitting here. He’s going like, “You should work on your brand,” and here I am hiding in a corner, right? So he called me on it.

And it was really difficult at first. So, I did things that were comfortable. I did dinners. I did dinners with reporters. And, like, somebody wrote a story on me out of that, which we didn’t work on that. It just kind of happened. It sort of happened organically. And then I always have, like, my happy home place. Germany — the Germans, like, want to talk to me all the time, because there’s so few Germans in Silicon Valley, and there’s all this tech tourism happening now. So if I want, like, an easy win, I’ll just go talk to the Germans and it’s like, “All right, fine.” But, like, that’s what I was saying. Like, find where you’re comfortable, right, and work your way into it. And it doesn’t have to be pressed, as Alex was saying. Find your way where you’re comfortable and, kind of, worm your way into it.

Hanne: That’s a great note to end on, and we’ll take some questions. If anybody has, ask away.

Audience Q&A

Woman 1: Thank you very much for being with us this morning. Most of us were transitioning out of the military, right? And so we’re in the space of, somewhat recreating ourselves, trying to, you know, downplay — even though we’re proud of our achievements in the military, you’re trying to connect the dots where people see you being in an executive space, or being in the tech industry. So while we’re transitioning, is all the advice you gave the same? And then — or, also, maybe when you get more specific, of where specifically [there’s] a company or industry that you want to go into, how do you shift? How is that brand shifting happening? And can you do this by yourself, or is it something that you’d actually need to hire someone?

Alex: You can definitely do it by yourself. I think the interesting thing is, on the stories, it’s what translates. It’s — what [are] the activities in the work that you did. In your military experience, a lot of the leadership skills in general, without being very specific to what each company does and what you’ll need to do in that company — finding those bridges of the work that you did, and the kinds of teams that you ran and oversaw, project management. Like, take those very basic things that are core to any leader anywhere, and map those for people just with your experience.

Margit: Yeah, and I would say — I mean, you’re going to laugh, and rightfully so. But Silicon Valley thinks of itself as a place of disruption, which means there are uncertain environments that are wobbly. They can shift any second. And a company that’s hot now is not tomorrow and whatnot. And it’s full of people with engineering degrees, but not a lot of actual, sort of, real-world experience. So what do you guys do? You guys go into uncertain environments and make stuff work, basically out of nothing.

So, I think that’s highly, highly applicable, and so you just need to find out the specifics of how you’ve led, and explain those in plain English. But, like, we need so much of that, because a lot of the folks here, they are running large companies but, like, they’ve never run a thing. They’re out of a dorm into their new dorm, with kombucha and massage tables. It’s a little mind boggling. So, someone like you coming in there is like, “All right, people, here is how we’re going to go…” is a thing of beauty, and I think that should be highly, highly transferable, and desired.

Man 2: Thanks for both of your time. The idea of who you are is much more multifaceted than just, “This is my brand, this is what I want to be seen for.” Like, it can be situationally dependent. It can change on your life circumstances, and it can — you know, I may need to be a jerk in this situation, and that’s who I have to be, [and] in this situation I’m not. How do you encompass all of that authentically into one brand without having to, like, hide this side of yourself?

Margit: Well, look, the brand is not trying to prescribe every detailed behavior in every situation, but I think — you know, having to be a jerk in a situation, that’s just sort of adjusting your management style, right? But, I think if you have three or four —three, we like three — brand acquisitions, it gives you a well-rounded body of, like, the essence of who you are. It doesn’t describe every behavior. And it’s also not static. I mean, if you looked at me funny when I was a teenager, I would be blushing, and I rarely spoke. And I can speak now, even in a different language. Go look at that. So, it’s not static as I think we might have made it sound.

Man 3: I’m finding that it’s okay to be associated with a startup that fails. It’s actually positive for a lot of people, but it’s very negative, it seems, to be associated with a stolid, old-fashioned company who may be successful, but, if you go there in your career, you’re quickly known as one of “those guys.” Not good enough to make it at a, you know, high growth — is that a real concern, or is that something that we should ignore?

Alex: I so lived this. This is literally where — I came here. I got to — yeah, my first job since moving here was with “WIRED” magazine. So, that was, sort of, my first, kind of, couple years, which was great, and I got to learn the space. And then I got to OutCast, and it was like my GE-ness — because we worked on startups and [they] were like, “Ugh, ooh, how embarrassing for you, basically.” And I think that was part of my year of feeling horrible, like, all this stuff I learned. And then I realized, “You know what? All of the stuff I learned through osmosis, through being in boardrooms, or just my experience traveling around the world is actually bigger.”

So, I had to, sort of, move from feeling really bad about myself about it, right, and that it was an albatross. And I have to say, I over-rotated a little bit in the beginning. I tried to bring, like, too much project management, or too much process, to the company, I think, in the beginning, and then I found my way. You know, by people saying, like, “This seems too much,” right? So, I learned also a lot of, like, not necessarily my way was always the right way. So learning to be flexible like a startup, I think, was hugely valuable. But you will get that. A lot of startups don’t have that experience of how to run, you know, a big company, and that is actually what they all aspire to so it’s sort of ironic.

Margit: I think there’s the chatter, and then there’s, like — we have an executive talent team. When a startup gets a certain level of momentum, they actually do want someone who has sold to big customers before, or who has worked in a big security department before. Like, they do do that. There’s, like, the what’s cool and there’s like — Forbes does a list of, like, the 30 under 30, and the 20 under 20. And, like, nobody does the 60 under 60, right? But, like, you do — you know, I think in the real world, once companies get to a certain scale, they actually do want the experience, and they do want, sort of, the big company-ness.

Alex: But I do think you’ll pick up, when you’re interviewing or talking to these companies — you easily can pick that up, I think. There are some founders who aren’t very good at the “this is the way I did it at Microsoft,” and you could feel it very quickly, they’re not interested. And then it’s just, you know, “Fine, good to know,” or not your place or maybe that is your place because you don’t want to be like Microsoft.

Hanne: One more maybe?

Woman 2: Hi, thank you so much for your candor, by the way. It’s very refreshing.

Margit: We only have one version. It’s the brand working.

Woman 2: So, we’ve had a lot of feedback on translating military skills into civilian skill sets and things like that. And really what that boils down to is branding, in a way. And one of the things that I think is pretty universal throughout the military is the ability to be, you know, an athlete, and do a ton of things all at the same time. I think the problem with that, with our personal branding efforts, is how do you portray the fact that, “Hey, I have a lot of different skill sets,” without coming across as contradictory? I think my concern is just that, if we do brand ourselves as this athlete that can come in and do a lot of things, it’ll come across as we’re, sort of, a jack of all trades and a master of none. You did kind of touch on it, but how do we keep from being, I guess, pigeonholed into, like, the standard military, “Oh, you’re a military member, so you need to do this specific thing and kind of…” Does that make sense?

Alex: Would you say that you were wide and deep though?

Woman 2: Yeah.

Alex: That’s how I would phrase it, right, that you can go wide. Like, wherever you’re going to go, you’re going to be successful, because you know how to go deep, and then think of maybe, you know, two examples where you did that. Like, we say that all the time, even with what we do. Like, our portfolio companies we work with is — everything from Patagonia to Amazon to Airbnb. <inaudible> company, right? So, when they came to us, they were like, “Well, do you have life sciences?” and we were like, “No, but…”

Margit: You know that.

Alex: “…we know what to do.” We’ll learn really quickly on life sciences. Like, we get up to speed. We know so many industries, then we go deep. You know, we know there’s a way to get smart, and then we can go deep but we are not — and then we own it, by the way. We say, “We’re not a life sciences agency. If that’s what you want, we can make a recommendation for you, but I can’t pretend to be something I’m not.” And then usually they’re like, “Ooh,” or they’re like, “Thanks, we’ll be moving on,” and I’m like, “Okay, bye.” So that’s okay. Like, I would rather say that than be like, “Yes, we can do that for you.” And then you get in there and you’re like, “Shit, really there’s no way I can do that,” or public policy…

Margit: Only so much winging it.

Alex: All right, thank you.

• 

  Alex Constantinople
• 

  Margit Wennmachers is the head of marketing and content at a16z, where she also advises entrepreneurs on their communications and marketing strategies. Previously, Margit cofounded the The OutCast Agency.

  Follow Twitter
• 

  Hanne Winarsky

Evolution and technology have allowed our human species to manipulate the physical environment around us — reshaping fields into cities, redirecting rivers to irrigate farms, domesticating wild animals into captive food sources, conquering disease. But now, we’re turning that “innovative gaze” inwards: which means the main products of the 21st century will be bodies, brains, and minds. Or so argues Yuval Harari, author of the bestselling book Sapiens: A Brief History of Mankind and of the book Homo Deus: A Brief History of Tomorrow, in this episode of the a16z Podcast.

What happens when our body parts no longer have to be physically co-located? When Big Brother — whether government or corporation — not only knows everything about us, but can make better decisions for us than we could for ourselves? That’s ridiculous, you say. Sure… until you stop to think about how such decisions already, actually happen. Or realize that an AI-based doctor and teacher will have way more information than their human counterparts because of what can be captured, through biometric sensors, from inside (not just observed outside) us.

So what happens then when illusions collide with reality? As it is, religion itself is “a virtual reality game that provides people with meaning by imposing imaginary rules on an objective reality”. Is Data-ism the new religion? From education, automation, war, energy, and jobs to universal basic income, inequality, human longevity, and climate change, Harari (with a16z’s Sonal Chokshi and Kyle Russell) reflect on what’s possible, probable, pressing — and is mere decades, not centuries, away — when man becomes god… or merges with machines.

Show Notes

• How humanity is focused on changing itself rather than the external world [0:42]
• The illusion of the self and the problem of tribalism [6:32]
• How collecting personal data could lead to hyper-personalization [12:16], and even “religions” based on technology [20:53]
• The future of work and UBI [25:44], and what humans will be good at in the future [32:17]
• Political questions [36:27] and a look to the future [39:51]

Transcript

Sonal: Hi, everyone. Welcome to the “a16z Podcast.” I am Sonal, and we’re very honored today to have as our special guest Yuval Harari, who teaches at the Department of History in the University of Jerusalem and specializes in macrohistory and the relationship between history and biology. He’s the author of “Sapiens,” which is a mindbogglingly good book, and now has a new book just out, “Homo Deus.” Did I pronounce that properly?

Yuval: I use the Latin pronunciation, which is homo de-oos.

Sonal: De-oos. Okay.

Yuval: But you can say homo dee-uhs.

Kyle: I say the really bad, like, non-accent dey-uhs.

Yuval: Dey-uhs is great. Yeah.

Inward human evolution

Sonal: That, by the way, was Kyle’s voice, who is also joining us on this podcast. He’s on the deal and investing team and covers a lot of the technology like drones, AI, and a bunch of other stuff. So just to get things started, we talk a lot about innovation and technology, and I’ve always wondered what’s the simplest definition of technology and innovation. And reading your book, “Sapiens” in particular and then “Homo Deus,” the thing that really struck me is that technology is the greatest accelerator humankind — in fact, the evolution of all the species on earth — has ever seen, because it allowed us to essentially bypass evolutionary adaptations where we could become seafarers without having to grow gills like a fish, for example. And so that is an incredibly powerful idea, but that’s non-directional. Given that your new book and your work, essentially, the first phase was talking about organic history of our species, and your new book is shifting to a more inorganic version, I’d like to hear what drove that shift.

Yuval: Well, I think that so far, for thousands of years, humans have been focusing on changing the world outside us, and now we are shifting our focus to changing the world inside us. We have learned how to control forests, and rivers, and other animals, and whatever, but we had very little control over what’s happening inside us, over the body, over the brain, over the mind. We could stop the course of a river, but we could not stop the body from getting old. If a mosquito annoyed us, we could kill the mosquito. But if a thought annoys us, we don’t know what to do about it. Now, we are turning our innovative gaze inwards. I think the main product of the 21st century will be bodies, and brains, and minds. We are learning how to produce them. And as part of that, we may also for the time — not only in history. For the first time in evolution, the evolution of life, we may learn how to produce non-organic life forms.

Sonal: That’s amazing.

Yuval: So after four billion years of evolution of organic life forms, we are really on the verge of creating the first inorganic life forms. And if this happens, it’s the greatest revolution in the history of life since the very beginning of life.

Sonal: What do you mean by inorganic life forms? Because in your book, you draw a distinction between biological cyborg and nonorganic. Are we just gonna be, like, living in a network? Is that our identity, then? Is that who we are? Like, what do you see?

Yuval: It could be something that exists only in cyberspace. I mean, you hear a lot of talk about uploading consciousness into computers, or creating consciousness in computers. It could be life forms in the outside world, but which are not based on organic compounds. It can go any of these ways, but the essential thing is, it’s no longer limited by organic biochemistry.

Sonal: Evolutionary psychologists, biologists talk a lot about our hands and the formation of our hands as tools. One thing that’s happened to me, anecdotally, is as I use my mobile phone more and more, my hand muscles have literally atrophied to some extent. I know this because I started taking notes again instead of on my phone to be polite in meetings, and my handwriting is literally — I used to win awards for handwriting, and now it’s like chicken scratch.

Yuval: But it’s much more extreme, because for four billion years, all parts of an organism had to be literally in the same place for the organism to function.

Sonal: Oh, right. Like, physically — like, in a single entity.

Yuval: Physically connected. I mean, if you have an elephant, the legs of the elephant must be connected to the body of the elephant. If you detach the legs from the elephant, it dies or it can’t walk. Now, as inorganic life, there is absolutely no reason why all parts of the life form must be at the same place at the same time.

Sonal: That’s mind-blowing.

Yuval: It can be dispersed over space. This is something that for four billion years was unthinkable, and it’s just around the corner.

Sonal: We’re essentially already uploading ourselves into the cloud, online social networks, in the World Wide Web. That’s actually replacing writing as a major artifact. That’s our new collective history. One of the consequences of that is it changes the dynamics of what becomes real and not real, and it reminds me of this famous story from Ray Bradbury called “The Veldt,” which basically is this story where there’s a virtual world that these two kids sort of enter, and they end up killing. And you ask a similar question in the book. You give the anecdote of Jorge Borges’ short story “A Problem,” and the story of Don Quixote. It sort of is this blending of delusion and reality.

Yuval: The question is what happens when our illusions collide with reality. And with humans and human history, you see more and more that our fictions and illusions are more powerful, becoming more and more powerful.

Sonal: <inaudible> say fake news. This is a big debate that’s playing out right now in the United States.

Yuval: Well, you know, it’s fake news when we — with all this idea of the age of post-truth, I would like to know, when was the age of truth?

Sonal: That’s my question. I totally agree with you.

Yuval: Was it the 1980s? Was it the 1930s, the 19th century?

Sonal: It never existed, right?

Yuval: I mean, as far back in history as you go, what kept humans together in society is belief in shared illusions and shared fictions.

Sonal: Imagined realities or imagined orders.

Yuval: Yes, imagined realities, like when you swear the U.S. President to office, he swears on the copy of the Bible. And even when people testify in court, “I swear to tell the truth, the whole truth, and nothing but the truth,” they swear on the Bible, which is so full of fictions, and myth, and error. It’s like you can swear on Harry Potter just the same.

Sonal: Some people do.

Yuval: Some people do, that’s true. When, for thousands of years, human societies have been built on shared fictions and shared illusions, and there is nothing new about that, it’s just that with technology, actually, our fictions and illusions become more powerful than ever before.

Sonal: Invisible to, I think, one another.

The illusion of the self and tribalism

Kyle: One of the illusions that you talk about being broken down by the advancements in science and technology is the illusion that we’re all individuals. Free markets and capitalism is the idea that there’s, like, a bunch of products that appeal to you as an individual, and they try to put those individuals into buckets and market towards them. And, actually, it turns out that scientific breakthroughs show that, actually, there isn’t just this, kind of, one individual you that accumulates through all of your experiences. Your brain is just kind of spitting out a lot of things. Maybe it’s deterministic, maybe it’s random, maybe it’s probabilistic, but you don’t necessarily have control over that. And so if you don’t have control over the desires — that your brain is spitting out the random thoughts, how much of any of that is actually you? And so, what are the implications of that?

Yuval: I think what we are seeing is the potential breakup of the self, of the individual. The very word individual means, literally, something that cannot be divided.

Sonal: Indivisible.

Yuval: Indivisible. And it goes back to the idea that, yes, I have all kinds of external influences, and my neighbors, and my parents, and so forth. But deep down, there is a single indivisible self which is my authentic identity. And the way to make decisions in life is, just forget about all these external disturbances and try to listen to yourself. Try to connect to yourself. And the idea is, you just need to do whatever this inner voice tells you to do. But science now tells us that when you look inside, you don’t find any single authentic self. You find a cacophony of different conflicting voices, none of which is your true self. There is just no such thing. And even in the 20th century, the big fear for individualism was that the individual will be crushed from outside. Now, the threat comes from the opposite direction. The individual will break up from inside, and then the entire structure of individualism, and democracy, and the free market — it all collapses with the individual. It all collapses with the self.

Sonal: Or, just one alternative possibility, because this is actually what struck me most when reading “Sapiens,” and then reading “Homo Deus” afterward — is that the big theme of “Sapiens” was this great unification of humankind, and being able to collect people into empires, nation-states, outside of these, sort of, hunter-gatherer tribes. And now when I look at what’s happening because of this mass coordination online, you’re now seeing this return to tribalism in some ways, I would argue.

Kyle: Well, that’s, like, what the value of shared illusions are, whether it’s religion, or the idea that we’ve got this free market system but some safety net to keep it all functioning and keeping anyone from being exploited. The point of having that shared ideology or that shared illusion is, you get to pretend that we all care about the same thing, that we’re all coordinated towards the same goals.

Sonal: Right. Now, though, because of the internet, you can actually identify what the same thing is at a very micro-targeted niche level in a way that was unprecedented. No longer where you were born, to your point, physically located. It could be now — your political beliefs. It could be your belief about, you know, if you’re a fan of Harry Potter. Are you a Slytherin or a Gryffindor? Like, it could be any of those things, and people collect into new tribes. And I find this fascinating because you do see sort of this return to the past, not in a pastoral way, but you’re seeing this coming full circle.

Like, you know, the Industrial Revolution created adolescence. Are we gonna go back to a world where you don’t need adolescence again? You needed banking credit. Are we gonna go back to a world where, because of online algorithms and new information sources, you don’t need that version of a credit score. You can go back to this trusted personal manager who essentially knows what he needs to know in order to invest in you as a risk. So I always wonder in this context if this is another thing to think about, not just at an individual level, but sort of a return to tribalism, especially lately.

Yuval: The present stage of a new nationalism or tribalism — I think it’s just a phase. It’s a backlash against globalization. And the main problem — it doesn’t have any solutions to the deep problems of the 21st century. All the major problems of humankind in the 21st century are global in nature. It’s climate change and global warming, it’s global inequality, and, above all, it’s technological disruption. The implication of the rise of AI and bioengineering and so forth — you cannot solve any of these problems on the national level. The nation is simply the wrong framework for that. And, therefore, I don’t think that nationalism really has relevant answers to the problems we now face.

Sonal: I agree with you.

Yuval: So I don’t think that nationalism is our future. I think looking further to the future, what we will see with regard to the individual is that, at a certain point, external entities, whether it’s corporations or whether it’s governments — they will have enough data, especially biometric data, and enough computing power to be able to understand me better than I understand myself. Very soon, Facebook or the Chinese government will be able to do that. And once you have an external entity, an algorithm, that knows me better than I know myself, this is the real turning point. This is the point when individualism, as we’ve known it, doesn’t make any sense — when democracy and the free market become completely obsolete. And we need fundamentally different models for running the world and for understanding what’s happening.

Biometric data and personalization

Kyle: Right. For now, several hundred years, the market as a mechanism for saying what our opinions or our desires really are, has been the most efficient mechanism. We could best allocate production towards things that people find valuable because they’re voting with their dollars. But if you can accurately say, based on this person’s heart rate, what they’re paying attention to, how they react to particular inputs, you know, whether it’s an advertisement, or some new way of interacting with things based on new technologies like VR — you could know, like, the closest thing to the underlying motivation, desire — even better than the person themselves maybe would. But at the other side of it, there’s an example you give —  and this goes back to the topic of, like, free will and individualism — lab rats that have electrodes hooked up to the reward centers of their brain. Where you have them navigate a maze, or climb ladders, and go down little chutes by basically stimulating their reward center. And it basically influences that rat’s desire. It doesn’t feel like it’s being coerced into doing that activity. It’s like…

Yuval: Yeah, the rat doesn’t know.

Kyle: “Oh, wow. I’m really into the idea of climbing this ladder right now. This is awesome.”

Sonal: The rat race.

Kyle: So, what’s interesting is, markets, as efficient as they are, like — part of how they worked was this idea of marketing to instill desires. Car ads giving you this vision of being on the open road and free, and wind blowing in your hair, and then, at some point, the desire pops up at a time when you could act on it. You buy a car. Whereas the future state that you describe is, imagine you had a headset that was like a miniaturized fMRI that can detect exactly where the parts of your brain would need to be stimulated to make you really want to play the piano right now, so that you’ll be motivated intrinsically to learn it. You could basically sell the idea of being into this. And so, being able to read your desires — but also being able to shape your desires — what do you think the interaction of those two look like?

Yuval: We don’t know. I mean, the basic tendency is to think in 20th-century terms, that they’ll try to manipulate us. And this is certainly a danger but, intellectually, it’s the less interesting option — that, okay, they’ll use it to advertise in a different way, to shape our desires without even our knowing it, which they’ve been trying to do for decades. They’ll have better tools [for] shaping our desires. The deeper and more interesting question is, what if Big Brother can really do a better job than the individual in understanding what you want and what you need? Because many people discover during their life that they don’t really know what they want, and they often make terrible decisions in the most important decisions of their lives — what to study, where to work, whom to date, whom to marry. What happens if you have an external entity that makes these decisions for you better than you can?

It starts with very simple things, like choosing which book to buy. If the Amazon algorithm really picks books that you are very happy with, then you’ll gradually shift the authority to choose the books to Amazon. And this may happen with more and more fields in your life. And the really interesting question is not if they try to manipulate you. The really interesting question: what if it works?

Sonal: Oh, that’s such an interesting question.

Yuval: What does it mean to be a human being, when all the decisions in your life are taken by somebody else who really knows who you are? It’s like being a baby forever.

Sonal: It’s already working, on some level, because you might have a million other movies out there, but you really don’t care because you only care about what’s in the Netflix catalog because you’re looking for convenience of being able to binge-watch and get it on-demand in the moment. So, it’s already reshaping that cultural landscape. I mean, it’s already happening, [to] some extent.

Yuval: I think the big breakthrough will come with biometric data. So, for most of these algorithms, whether it’s Amazon, or Netflix, or whatever, they work mainly on the basis of data external to my body. They follow me around in space, see where I go, which places I visit. They see my likes and dislikes on Facebook, what do I buy, and so forth. But the real breakthrough will come when they start receiving more and more data from biometric sensors on or inside my body.

Sonal: Right, like quantified cells, wearables.

Yuval: Yeah. I read a book, and Amazon knows exactly what is the impact of every sentence I’m reading on my heartbeat, on my blood pressure, on my brain activity. This is really where you can see how an external system can make better decisions for you than you can make for yourself. 

Kyle: Yeah, today, these systems are basically reflecting ourselves back at us. If you look at products — because of cookies, when you go elsewhere on the web it’s like, “Oh, I see that thing again.” Like, it’s just being reflected back at me. Same thing with your Netflix queue. I gave certain star ratings to certain things. It’s reflecting that same pattern back at me with recommendations.

Something that’s interesting to me is the idea of mapping concepts in a future space using deep learning, and then basically projecting it in different forms. And so, the idea of tracking what your eyes are looking at, what’s keeping your attention, what makes your heart rate get up, what makes your eyes dilate while you’re reading a book — you can imagine, as you’re reading it, being formatted and communicated in different ways, because they know this different way will reach you better and you’ll be more receptive to it. And so it might not necessarily be what feels coercive to us — a system of plugging an electrode into your brain and saying, “Now you’re gonna care about reading history.” It’s gonna say, “Here’s the optimal way to present history to this specific individual.”

Yuval: This is especially being explored in new educational methods. An AI teacher that studies you while it is teaching, and adapting to your particular strengths and to your particular weaknesses. Also, breaking down all the traditional limitations and barriers of modern education. Modern education takes place in school, and you have this division. There is school and there is real life outside school. And, also, in school — now, if you have — consider [if] you have a single AI mentor that follows you around everywhere…

Sonal: Your whole life.

Yuval: …24 hours a day, connected to biometric sensors on your body, and there is no longer any division between school and life. There is no history teacher and mathematics teacher. You have the same teacher for both. And you don’t have to be part of a group, like, the 30 other kids in the class.

Kyle: Basically, an AI assistant where it’s constantly in Socratic debate with you.

Yuval: Yes.

Kyle: Kids are inclined already to say, like, “Okay, but why? Okay, but how? Okay, but why?” And they keep digging kind of deeper until you as a parent or teacher are just like, “Because it is, okay?” Whereas an AI system, assuming it’s mapped out, like, the entire cannon of human philosophy and knowledge, could basically just keep going. Even if it doesn’t go all the way to that extent, you could have a huge increase in productivity of, you know, education, just by providing those kinds of tools to kids.

Sonal: Mass personalization. I mean, I come from the world of developmental psychology and education, and the Holy Grail has always been this idea of mass personalization, to be able to customize. But I want to make two points. One, I agree with this idea. Vygotsky had this idea as a constructivist way of learning. You’re constructing, you’re learning your world, and that’s how you learn these concepts in a very fundamental way. And it’s really ironic, because educators have been trying to fake that in the school setting for years — by Montessori methods and all these other — Reggio Emilia — because of this false artificial divide between real life and school. The flip side, however, and I don’t think we can ignore this, is that there is a social element to why school matters — a socialization component that has arguably nothing to do with education — and where there is shared learning and collaboration and the interaction of students. And so, I wonder what this means for that.

Yuval: You can have it outside school as well.

Sonal: You’re saying there’s no distinction between school anymore. It doesn’t matter.

Yuval: It doesn’t have to be limited — that all your friends are the same age as you. There is no reason why the group with which you socialize in school, everybody has to be the same age.

Sonal: Well, that actually is another way that technology brings you back to the past, because if you think of “Little House on the Prairie,” the schoolhouse was essentially all the grades in a single school because of physical location. But you’re arguing that those boundaries, the idea of a schoolhouse, essentially melts away.

Kyle: That feels like an inevitable transition anyway, whether it’s corporations or education. It’s this idea of, “take in this large set of inputs, crank out some modified set of outputs that fulfills some need.”

A new technology-based ethics

Sonal: Well, the question that I have for you guys, and especially given “Sapiens” and the theme of “Homo Deus,” is what do humans have to believe in order to make this reality continue happening? Do they not have any agency in any of this? Because it sounds like we’re almost talking about, you know, these uploaded brains in a vat. Is there any sense of coordination, consciously? Is there a new religion? I used to watch “Star Wars” as a kid. I remember thinking to myself, because I grew up Hindu — and you learn a lot about all these Hindu gods and goddesses. I remember thinking, this reminds me a lot of hearing about the Mahabharata and all these other things happening. Anyway, I would argue that science fiction is like religion for a lot of people, but what do people have to believe in this new world? What is their religion? Is there one? I mean, you make the argument about — data is a new religion, but that sounds, to me, more of something that’s there versus something that people are choosing, like, creating new myths and gods around actively.

Yuval: I think we are seeing, and we will see more and more, the rise of kind of techno religions. Religions based on technology that make all the old promises of traditional religions — they promise justice, and happiness, and even immortality in paradise. But here on Earth, with the help of technology, there already has been one very important techno religion in history, which is socialism.

Sonal: Oh, I never thought of that that way.

Yuval: Which came in the 19th century with the Industrial Revolution. And what Marx and Lenin basically said — “We will create paradise on Earth with the help of technology,” steam engines and electricity, and so forth. When Lenin was once asked to define communism in a single sentence, the answer he gave was, communism is power to the workers’ councils, plus electrification of the whole country. You cannot establish a communist regime without industrialization. It’s based on the technology of the Industrial Revolution — electricity and steam engines, and so forth. And the idea is, we’ll use this technology to create paradise on Earth. It didn’t really work very well. So, now, I think we will see the second wave of techno religions. Now we have genetics, and now we have big data, and, above all, we have algorithms. They’re our salvation. Paradise will come from the algorithms.

Kyle: You talk about, in the book, the idea that the more you commit or sacrifice on behalf of your ideology or religion, the more you buy into it — because you have this sunk cost. And so the idea of, like, sacrificing a goat or a cow to a god made you buy more into, because I can’t have spent the last eight seasons sacrificing goats and have it been for nothing. So, looking forward then, we’re hitting some kind of productivity cap as normal humans — that autonomous machines and systems are going to beat us, so we have to sacrifice our own humanity to increase our own productivity and augment ourselves. You can also see the emergence of some kind of powerful ideology. Like, the religion of the 21st century onward is, we are the gods?

Yuval: This is actually an old idea. Humanism, which goes back to the 18th century, even 17th century, is saying humans are the gods. Humans are the source of all meaning and authority. Everything you expected previously from the gods — to give legitimacy to political systems to make decisions in ethics. Humanism comes and says the highest source of authority in politics is the voter. The highest source of authority in economics is the customer. The highest source of authority in ethics is your own feelings. Humans are the gods.

Now we are entering a post-humanist era. Authority is shifting away from humans. If, in the last 300 years, we saw authority descending from the clouds to Earth to humans, now authority is shifting back to the clouds — but not to God, but to the Google cloud, to the Microsoft cloud. The basic idea of this — if you want [a] new religion or new ideology, is again — if given enough data and enough computing power, an algorithm can understand me better than I understand myself, and make decisions for me. In the end, religion is about authority. The basic question of religion: where does authority come from? And the answer of the 21st century: authority doesn’t come from humans, authority comes from data and from data processing. There is also an underlying new ontology. What is the world? What is reality? In the end, reality is just a flow of data. Physics, biology, economics — it’s all just a flow of data.

Sonal: It’s all a type of algorithm.

Kyle: We are just computers interpreting some fraction of reality.

The future of work and UBI

Sonal: They’re all algorithms. That’s the connective tissue of everything, from biology, to computers, to everything. I have a quick question for you here. What does this mean for the future of the firm — work? I would love to hear your thoughts on the universal basic income debate that’s playing out around the world right now, because that’s essentially people opting out of the rat race, in some arguments.

Yuval: I think we need new economic models in place. For the moment, when AI and robots, and so forth, may push more and more humans out of the job market. And we might see the creation of a new class of people who are not just unemployed but unemployable. At present, the best idea so far that people managed to come up with is universal basic income. The problem there, is that we don’t really know what universal means, and we don’t really know what basic means.

Sonal: Right, and where the income comes from, but that’s another sidebar.

Yuval: No, let’s say you tax and use the proceeds to give people universal basic income. Now, then the question is what is universal? Would we see the day when the U.S. government taxes the profits of Google in the U.S. and uses it to pay people in Bangladesh or Mexico who lost their jobs? So this is the first issue of universal, because now the economy is global, and a disruption of the economy, say, by the rise of AI will require a global solution. And most people who think about universal basic income, they think in national terms. Universal, for them, means U.S. citizens. The other problem is, what is basic? Basic human needs keep changing all the time. We are beyond the point when basic needs meant food and shelter.

Kyle: And the problem is that humans are biased towards looking at examples that are based on who you know. It’s hard to see, kind of, that level of UBI pulling it off. It feels like people’s expectations would be much higher depending on where they are and what life they’ve already lived.

Yuval: The basic problem is that people’s expectations keep changing. Usually, they grow. As conditions improve, expectations increase. And, therefore, what you see is that even though the conditions over the last centuries of most humans have improved quite dramatically, people are not becoming more satisfied, because their expectations also increase. And this is going to continue in the 21st century.

Sonal: Yeah, I have a question here, because in “Sapiens,” you said something that I thought was very profound when I read it, which is that the agricultural revolution was actually one of the greatest frauds ever perpetrated on ourselves. And so if you think about this shift, from Agricultural Revolution to Industrial Revolution to now, essentially, Information Revolution — what’s the fraud that we’re perpetrating on ourselves now? Where does meaning come from? Because I think the thing that people often forget to address when they talk about the universal basic income and, you know, future of work debate is this idea of meaning. And does that even matter at the individual level?

Kyle: Restless people tend to pick up the pitchforks.

Sonal: Right, exactly. Exactly, because it also goes to your points — and this is a universal theme that we have to address on some level — of further entrenching inequalities. That’s an important thing to think about.

Yuval: There are two different problems. I mean, first, you have inequality. And once more and more people no longer work, they depend on, say, universal basic income, then they have no way of closing the gaps. They depend on charity, on whatever the government is able or willing to give them, and you just don’t see any way in which they can close the gap.

Sonal: That’s if they’re dependent on it, because it can also be something that’s supplementary to something else you do.

Yuval: I’m thinking in terms of what happens if, again, AI pushes more and more humans out of the job market, so they rely on universal basic income. And it provides whatever it provides, but if they want more, they just have no way of getting more. So this, kind of, entrenches inequality. And if you add to that biotechnology and bioengineering, you get for the first time in history the potential of translating economic inequality into biological inequality.

Sonal: Yes.

Yuval: If you look back at history, let’s say, the Hindu caste system — people imagined that the Brahmins are superior, they are smarter, they are more creative, they are more ethical. But, at least as far as scientists today are concerned, this wasn’t true. It was all imagination.

Sonal: Right. It was not true at all.

Yuval: It was not true. It wasn’t true that the son of the Brahmin or the son of the king was biologically more capable, smarter, more creative, whatever, than the son or daughter of a simple peasant. However, in the 21st century, it might be possible for the first time to translate economic inequality into real biological inequality. And once this starts, it becomes almost impossible to close the gap. So this is one problem of a rise in inequality. Another problem is the question of meaning — that even if you can provide people with food, and shelter, and medicine, and so forth, how will they find meaning in life? For many people, their work, their jobs provide them with meaning. “I’m doing something important in life.”

Sonal: A mission. “I believe in this.”

Yuval: Yeah. So, one of the answers, some experts say, is that people will just play games most of the day. They’ll spend more and more time in virtual realities that will provide them with more meaning and more excitement and emotional engagement than anything in the real reality outside.

Kyle: Everyone just lives in their perfectly-optimized-for-them Holodeck.

Yuval: Exactly.

Sonal: Because you’re freed from the constraints of the physical realities.

Yuval: Yeah, and you get your meaning from the game, from the virtual reality game. And in a way, you can say, oh, this is nothing new. It’s been happening for thousands of years. It’s simply been called religion. I mean, religion is a virtual reality game that provides people with meaning by imposing imaginary rules on an objective reality. You play this game [where] you have to collect points. If I eat non-kosher food, I lose points. And if by the time I die, I gathered enough points, then I go up to the next level.

Sonal: I mean, in Hinduism, karma is essentially this great game of collecting and subtracting points across multiple lifetimes.

Yuval: Exactly.

Kyle: So, really quickly — this goes back to the automation, kind of, question and potential future. If you look back at, kind of, the Industrial Revolution, where humans as mechanical actors — just imbuing something with value by acting on it with their hands, or bodies with agriculture — that became less important as using animals, and then machines, were able to do that same task much more efficiently. Now, humans are valuable because they are knowledgeable operators of that machine. As part of the Industrial Revolution, the shift to services led to this idea that we’re not just investing in capital, we’re investing in human capital. We’re making people smarter so that they’re better at their jobs. Now, with AI systems, suddenly, again, you can just kind of buy knowledge capital as this thing that can be dropped in. Okay. An argument I hear…

Sonal: AI as a service, even.

Kyle: Right, how humans remain valuable is, well, we’re still social animals. We still are better than any machine at interpreting how other people are thinking about this and, you know, assuaging fears, or whatever it is — where the power of empathy is what humans will bring to the table. An interesting point you make is, actually, how humans accomplish a task — a doctor giving bad news about a cancer diagnosis. They are looking at the physical way that a person is moving their facial muscles, how their tone changes, how their voice cracks as they feel a certain emotion.

And if you look, that’s actually just pattern recognition, which is exactly what deep learning is good at. And so, is that even an advantage humans are gonna have, or are computers gonna be much better at looking not just at those exact same features that humans can, but also, like, zooming in on the eyes and looking at dilated pupils, and guessing at heart rate by looking at someone’s wrist or chest? What are humans going to be good at? What should people be investing in for, you know, the future to come?

Sonal: Yeah, what happens when human capital becomes commodified?

Yuval: We don’t really have an answer. Yes, many people, when they reach that point, they say, “Okay. We’ll invest in social skills, in empathy, in recognizing emotions.” The emotions are like the last…

Sonal: Emotional intelligence.

Yuval: The last frontier. But the thing is that emotions are not some spiritual phenomenon that God gave homo sapiens to write poetry.

Sonal: Another electrochemical, just like everything else.

Yuval: Emotions are a biochemical phenomenon. There are biological patterns just like cancer. When your doctor wants to know how you feel, he or she basically recognizes patterns in two kinds of data, as you mentioned. It’s what you say, and, actually, the tone of your voice — even more important than the content of what you’re saying. And, secondly, your body language and your facial expression. When my doctor looks at me at the clinic, she doesn’t know what’s the level of my blood pressure at the moment. She doesn’t know which parts of my brain are activated right now. 

But an AI, potentially, will be able to know that in real-time using biometric sensors. It will have much better sources of data coming from within your body. So their ability to diagnose emotions will be better than the ability of most, if not all, humans. So what will humans do? We don’t know. Nobody really has an idea, a good idea, of how the job market would look like in 30 or 40 years. We’ll have some new jobs. Maybe not enough to compensate for all the losses, but there will be new jobs. Problem is, we don’t know what they are. Because the pace of change is accelerating, it’s very likely that you will have to reinvent yourself again and again and again during your lifetime if you want to stay in the game.

Sonal: Right, when you don’t have premature death anymore, and you live your full life, or you even have extended longevity through technology, you can reinvent yourself, like, 10 times until you’re 100.

Yuval: The basic idea for thousands of years was that human life is divided into two periods. In the first period of life, you mostly learn. You learn skills, you gain knowledge. And then, in the second part of your life, you mostly work, and you make use of what you learned earlier. This is now — it’s going to break down. By the time you’re 50, what you learned as a teenager is mostly irrelevant.

Sonal: It’s already true right now.

Questions for the future

Kyle: So, now, you know, again, thinking about autonomy — you know, we’re already seeing the shift towards smaller militaries with really advanced equipment and fighter jets. And we’re gonna see robots on the battlefield. As humans become less valuable economic actors, as they become less necessary to fight for power at, kind of, that scale, how does that factor into, you know, the extension or lack thereof of, you know, political agency?

Yuval: Most people today have absolutely no military value. In the 20th century, the most advanced armies relied on recruiting millions of common soldiers to fight in the trenches. Now they rely increasingly on small numbers of highly professional soldiers, super-warriors, all the special forces and so forth.

Sonal: Surgically targeted.

Yuval: And they rely increasingly on sophisticated and autonomous technology, like drones, and robots, and cyber warfare. So you just don’t need people militarily as before, which means not only that they are in danger of losing their political power, but also that the government will have a far smaller incentive investing in their health, and education, and welfare. Maybe the biggest project and achievement of most states in the 20th century was to build these massive systems of education, and health, and welfare.

Sonal: Safety nets.

Yuval: And you see this not only in democracies but also in totalitarian regimes. But if you don’t need them as soldiers or workers, then the incentive to build hospitals, and schools, and so forth diminishes. In a country like, I don’t know, Sweden, I think the traditions of the welfare state and the social democracy will be strong enough that the Swedish state will continue to invest in the education and health of most of the people there, even if there is no military or economic necessity. But if you think about large developing countries, it’s much, much more complicated. If the government doesn’t need tens of millions of Nigerians to serve as soldiers and workers, maybe it will not have enough incentive to invest in their health and education. And this is very, very bad news for most of the human race, which lives in places like Nigeria and not in places like Sweden.

Kyle: And so what’s the best course of action to follow if that’s the case? Is it, make sure that the most inclusive institutions possible are in place before that transition happens or…

Yuval: We don’t have enough time. I think that we are not talking in terms of centuries. We are talking in terms of decades, and once the transition takes place, especially in the civilian economy. In the military, it already happened. We are there. In the civilian economy, maybe we have 20 years, 30 years, 40 years. Nobody really knows. It’s a very short time. If we don’t have a workable model by the time the transition is in high gear, then it’s going to be both an extremely difficult situation for the majority of people in the world, and the social and political implications are going to destabilize the whole world, including the first world.

Sonal: You walked in your book, your new book, a lot about how there are three types of capital that — raw materials and energy, but people have ignored a third type, which is knowledge. And my question, from just an economic perspective, is how does this tie into how we think about growth? Especially given what you just talked about — this need to enlarge the pie in order to avoid war and violence.

Yuval: It’s often thought that there is a limit to the potential growth of the economy, because there is a limit to the amount of energy and raw material we have access to. But this is, I think, the wrong approach. We have a third kind of asset, which is knowledge. And the more knowledge you have, the more energy and raw materials you also have, because you discover new sources of energy and new sources of raw materials. I don’t think that we are going to bump into a limit in terms of, “Oh, there is not enough oil in the world. There is not enough coal in the world.” This is not the problem. The problem is probably going to come from the direction of climate change and ecological degradation, which is something very different. People tend to confuse the two problems — not enough raw materials and the problem of climate change — but they are very different.

Sonal: I actually wanted to probe about this, because in “Sapiens,” one of the things that you talked about was how we’ve had waves of climate change throughout the entire history of our planet. And I’m no climate change denier by any means, but I can’t help but ask the question, you know — whether we are the cause or it’s a cyclical effect — what it means for what the next cycle of change will be. Because the one thing that came through loud and clear was how every wave of climate change has brought about a corresponding change in human evolution.

Yuval: Well, there certainly have been many periods of climate change before, but it does seem that this time it’s different, that this time it is caused, to at least a certain degree, by human action and not by some natural forces, like plate tectonics or ice ages or things like that. And the potential impact for human civilization and for most other organisms on the planet is catastrophic. So, you know, it could be both natural causes and human causes at the same time. It doesn’t make it any better. It just makes it worse.

Sonal: The effects are the effects, right. In your book, you have this beautiful quote, which I thought was really straight articulation. “Modernity is a deal. All of us sign up to this deal on the day we were born, and it regulates our lives until the day we die. Very few of us can ever rescind or transcend this deal. It shapes our food, our jobs, and our dreams, and it decides where we dwell, whom we love, and how we pass away.” And I wanna know if you have any parting thoughts for people whose lives are being shifted by some of the technological change?

Yuval: Since the main theme has been technology and the future of technology and its impact on society and politics, I think that my closing thought is that technology is never deterministic. You can build very different kinds of political and social systems with the same kind of technology. You could use, you know, the technology of the Industrial Revolution — the trains, and electricity, and radio — you could use them to build a communist dictatorship, or a fascist regime, or a liberal democracy. The trains did not tell you what to do with them. In the same way, in the 21st century, we’ll have artificial intelligence and bioengineering, and so forth, but they don’t determine a single outcome. We can use it to build very different kinds of societies. We can’t just stop technological progress. It won’t happen.

Sonal: It’s inevitable.

Yuval: But we still have a lot of influence over the direction it is taking. So if there are some future scenarios that you don’t like, you can still do something about it.

Sonal: Yeah. Well, thank you so much for joining the “a16z Podcast.” If people have not already read “Sapiens,” they must read that, and especially the new book, “Homo Deus: A Brief History of Tomorrow.”

Kyle: Thanks for coming in. We really appreciate your time.

Yuval: Thank you.

• 

  Yuval Harari
• 

  Kyle Russell
• 

  Sonal Chokshi is the editor in chief as well as podcast network showrunner. Prior to joining a16z 2014 to build the editorial operation, Sonal was a senior editor at WIRED, and before that in content at Xerox PARC.

  Follow Twitter

You’ve heard a version of this story before: Steve Jobs calls some executive out of the blue to come work for him. Only this time the story turns out great … and the company wasn’t Apple. This episode of the a16z Podcast shares some of the journey that former CFO Lawrence Levy went on with Steve Jobs as they took Pixar — a company then on the verge of failure — to its IPO and subsequent greatest hits.

It’s sort of an adventure story but is really more of a quest for product-market fit. How did they figure out a model for such an old-but-new business (i.e., animation and entertainment)? How did they take an improbable plan and figure out how to make it work — both qualitatively and quantitatively? How did they then navigate and straddle the diverse worlds of Silicon Valley, Hollywood, and Wall Street? And finally, how did they price their IPO, which was also a symbol of Steve Jobs’ comeback story … a narrative that’s sometimes lost in the Apple story.

From the business of creativity to corporate culture, Levy — former CFO of Pixar, board member, and author of the new book To Pixar and Beyond: My Unlikely Journey with Steve Jobs to Make Entertainment History — shares his (and Jobs’ untold) story. But it isn’t just a story about finding the right model and numbers to build, explain, and measure the business; it’s also, partly, about how to get the measure of one’s humanity, too.

Show Notes

• How Lawrence Levy came to Pixar [0:48]
• Why animation is difficult [6:00], and how they decided on the company’s focus [7:51]
• What it was like to work with Steve Jobs [11:40]
• Getting investors to accept Pixar [15:00], IPO pricing [17:50], and lessons learned [20:02]
• Levy’s and Jobs’ long-term plans for Pixar [21:32]
• Personal and philosophical takeaways from the experience [25:00]

Transcript

Sonal: Hi, everyone. Welcome to the “a16z Podcast.” I’m Sonal. Today we have, as our special guest, Lawrence Levy, the former CFO of Pixar, who also took the company public. He’s on the board of directors of Pixar and was in the office of the president, and has written a book, just out, called “To Pixar and Beyond: My Unlikely Journey with Steve Jobs to Make Entertainment History”. In this conversation, we cover everything from his partnership with Steve Jobs, to how they figured out what kind of business Pixar should be, to the delicate balance between finance, strategy, and the business of creativity. We even cover details about how they priced their IPO, and end on a more personal note about bringing a more humanistic approach to business and corporate culture. But first, we begin with a story of how this lawyer came to Pixar. Let’s just start jumping right into your story, which essentially you describe as an adventure story.

The early days of Pixar

Lawrence: Yeah, I think my experience at Pixar could definitely be characterized as an adventure. It started in 1994. Imagine that I’m sitting in my office at the company I was at then, which was Electronics for Imaging. And I was sitting at my desk and the phone rings. I pick up the phone, and on the other end of the phone, I hear this voice that says, “Hi, this is Steve Jobs. I saw your picture in a magazine one day.” He said, “I thought we would work together someday. And I have a company that I would like to tell you about.” And I immediately thought that he was talking about NeXT Computer, because NeXT was very well known. It had gone through quite a number of bumps. But then he says, “It’s Pixar.”

Sonal: Had you even heard of it?

Lawrence: I had heard of Pixar, but I didn’t know anything about it.

Sonal: How jarring because you’re hearing from this computer founder who’s suddenly going to a cartoon company. Or how would people have described Pixar at that point?

Lawrence: Oh, I wouldn’t have known enough to even know it was a cartoon company. I didn’t even hardly know what to say. But I was intrigued. It was just Steve Jobs on the end of the phone. He’s telling me about something called Pixar.

Sonal: So at that time, Pixar was a flailing company. I mean, it wasn’t at all the success that we know today. It was very early days. What were they? Were they a movie company, an animation company, an entertainment company?

Lawrence: If you did research at that time, which I did, you would have discovered they were really a graphics company. So, Pixar had really set out to change the world of high-end computer graphics. And they made an imaging computer. They’d made some short films that had, sort of, won accolades at some of the festivals and shows, but not even Pixar conceived of itself as an entertainment company in 1994.

Sonal: It wasn’t an entertainment company at the time. Clearly, it became one.

Lawrence: It was in a broad range of businesses. It made software, it did animated short films, it did commercials. It had this little project going on called “Toy Story.” And so my goal, my work, was to, sort of, figure out how to put all that together. And at the beginning, the last thing that I imagined was that it was gonna be an entertainment company. Why would you hire me to go in there and figure it out? Because I… 

Sonal: What was your background?

Lawrence: I was a lawyer and an executive. I was at the Wilson Sonsini law firm. I represented all these high-tech startups, I knew software, I knew hardware, I knew semiconductors, but I didn’t know any more about entertainment than, you know — that you go to a movie theater, buy your ticket, and buy some popcorn. So, it never occurred to me that I was going to run an entertainment company. And I actually thought it was going to be a company where we would pull together different businesses, and they would help — each of those businesses would — the risks of one would offset the other.

Sonal: It’s interesting you say that, because one of the lines that really stood out to me is someone telling you, think of it as a portfolio business. And I think that’s really interesting, because it actually touches upon all kinds of businesses. If you think about an R&D of a big Fortune 500 company, that’s having a portfolio framework for deciding which projects to invest in and not to invest in. If you think of VC, that’s a portfolio business. How does that portfolio mindset apply specifically to Pixar?

Lawrence: Well, at the time, I thought that it could be a business where its different parts would offset the risks of the other. And the reason was because the notion of becoming a standalone, independent animated feature film company was a crazy idea. 

Sonal: Why was it so crazy?

Lawrence: Because only one company in history had ever done it.

Sonal: Disney.

Lawrence: That was Disney. And they had only succeeded at it for a few years, from 1939, when they released “Snow White and the Seven Dwarfs.” They did “Dumbo” and “Bambi” and a few of these great films. But it was such a struggle to be a standalone, animated feature film company that Walt Disney — brilliantly — began to diversify. And so then you get the theme parks in 1954, in 1955, and they went into ABC television, and the Wonderful World of TV, and they opened Buena Vista distribution. They couldn’t sustain themselves as an animated feature film company. If you looked at the history, every other studio that had tried to do that had essentially failed. So, the notion that you could build an independent, standalone, animated feature film company was one that I basically fought every step of the way, because it seemed impossible.

Sonal: Well, there’s actually an interesting exchange between you guys. Steve told you, “Investors will love this. Pixar can be in a multimillion-dollar video market.” And you responded, “I agree, but let’s get the data first. Even then I’m not sure we can count on home video alone to take Pixar public.” So, tell me about that. How did home video play a role in all of that?

Lawrence: Well, the video market was changing the nature of the film business. That was the time when Blockbuster films and Blockbuster stores and — it was creating a whole new economy for film, which was great. And Disney was having incredible success at it. But the challenge for Pixar was that animated feature films take a long time to make. So, if you release a film every four years, that’s too much time. And if you have one film that’s a miss, which could easily happen, you could have an 8 or 12-year dry spell, and the home video market wouldn’t be any help at all.

Sonal: That’s kind of frightening, actually, to think about because you can now pivot a company three times and not even blink an eye, and here your product is this thing that has so much planning and development that goes into it. Just to break it down at the technological level, why is animation so hard?

Lawrence: An animated feature film has about 110,000 frames in it. And each one of those frames has an enormous amount of data. At that time, just to render a single frame, could take a day. So imagine how much computing power it would take to render an entire film. And then you had to have the software to actually manipulate the characters — a three-dimensional image of Woody or Buzz or something like that. How do you then manipulate that image to make it seem as though it’s alive, to give it emotion? Because emotion is very subtle movements of the eyes, very subtle movements of the mouth. Pixar had to invent the technology in order to do that.

Sonal: So, given this context of the difficulty of the animation itself, the fact that you can only have maybe one good movie every four years, and you’re putting a lot into one movie, how did you guys get to “Toy Story?”

Lawrence: Pixar was already making Toy Story when I arrived. They had committed to that in a 1991 contract that they went into with Disney. That was a little bit of a Hail Mary, in a way. Because otherwise, Pixar was perilously close to going out of business. Disney came along and said that they would, basically, front the costs for making that first animated feature film. But the price that Pixar paid for that was enormous. It was almost like selling your soul.

Sonal: Why?

Lawrence: Because the contract — and this was really my hardest moment. You know I was a lawyer? This shows you how little I knew then about the entertainment business. They had entered an agreement with Disney, but it was written in this arcane Hollywood code. When I got there, I realized that Pixar had essentially been tied up by Disney for what could have been, you know, 12 years, 15 years with a very, very tiny share of the profits.

Sonal: Oh, that’s awful.

Lawrence: And that was only if the films were ridiculous blockbusters. It just didn’t seem to have any chance. Its hands had been tied too much.

Developing the company’s focus

Sonal: So how did you guys come out of that? I mean, how did you go from this situation to the success that you and Steve and everyone that worked at Pixar built? 

Lawrence: Well, we went off on this quest, if you will.

Sonal: This is an adventure story.

Lawrence: It’s an adventure. So there are two parts to it. One is qualitative. What does it mean to build an entertainment company? And the other is quantitative. What do the numbers look like? At the time, we didn’t even have a spreadsheet of numbers that told us how these movies performed. I recount the story in the book of how we had to beg, borrow, and steal, just to understand how the numbers were gonna work. We’re starting to very slowly understand how these films make money, basically. And at the same time, Steve and I are shuttling back and forth to Hollywood. And we’re meeting these, you know, Hollywood executives. We met Edgar Bronfman Jr. and we met Mike Ovitz, Joe Roth. Basically, anybody who would open the door to us. We’d ask them all these questions about the industry, because we thought we can’t just do straight animated feature films. So we wanted to learn about all those other businesses.

Sonal: Was Steve on board with that diversification? Because I’ve always heard famous stories about him being such a focus guy. Did he agree that you had to diversify in order to make this a viable business?

Lawrence: I think Steve was on board with the process that we were going through. He hired me to assess and analyze this business, and basically partner with him to do that. So, we went on this quest. We learned all about the live-action film business, but the live-action film business is also a terrible business. It actually didn’t diversify the risk of animated feature film. We came to the notion that Pixar should be an animated feature film company, basically by default, which is that this is our only shot. We literally created the spreadsheet and it had on it, this is how many films we have to release and this is how they have to perform in order for Pixar to have a shot, and these are the conditions that have to happen in order for that to work.

Sonal: And that was a quantitative activity that you guys did.

Lawrence: That was quantitative, but associated with qualitative notions of — this is what that means. We have to get out of the software business. We have to get out of the animated commercials business. We have to triple the size of the company. We have to increase output. We have to do all these things. We have to renegotiate with Disney. And if we do all those things, and if our films are more successful than anybody will possibly believe they could be, we could make it. And so, it was literally a million to one shot that you could pull it off. And then I was like, “I have no idea how we can get this financed,” because the risks in this plan are absurd.

Sonal: Insane. That’s interesting. You described that Steve had hired you to partner with them. So you guys were partners?

Lawrence: We represented the business and strategic side of the company together. The creative side of the company was represented by John Lasseter and Ed Catmull.

Sonal: And Ed Catmull, of course. So, we have the creative side, and you have the business and strategic side.

Lawrence: Yeah.

Sonal: Those are essentially three legs of a stool. And they all have to be functioning in order to make the company succeed, and build this amazing thing. How did you guys make decisions, though? It’s not like you just had spreadsheets and these rubrics. Was there, like, a visionary, sort of, “We’re gonna go do this?” Was there an instinct? Was there someone who’s saying no and yes? How did you guys negotiate that?

Lawrence: With Steve, and with Ed as well, and John, it was basically a constant dialogue. Decisions came out of this dialogue. It was like we were in motion all the time. So there weren’t, like, moments where we’d sit down, and now we have to decide. It was more like it emerged from this continuous dialogue.

Sonal: It sounds like an incredibly creative friction.

Lawrence: It is. It is, because we never fought, but we didn’t always agree. There was this healthy debate, I would call it, one pushing the other, always pushing the other, you know, to sort of, figure it out.

Sonal: Never fought but didn’t always agree. That is the definition of healthy conflict.

Lawrence: Yeah, that’s how I would describe it. It was collaborative and respectful and great.

Working with Steve Jobs

Sonal: So, tell us some of your stories about partnering with Steve Jobs. It’s a really interesting moment right now, in recent history, given how few years it was that he passed away. But already three or four books have come out trying to repaint the picture, or maybe paint the picture. And you had a unique, front row — not just a seat, but you were an active participant.

Lawrence: One of the reasons I was motivated to write the book was, in the aftermath of Steve’s tragic death, all these things came out. And I started to see, “Well, what about the Pixar story?” I mean, it was a little bit of an afterthought. Some of it made it seem as though, you know, he started at Apple, and then he went away from Apple, and then he went back to Apple. And I’m like, “Well, wait a minute, that part when he went away from Apple — that was really important.”

Sonal: That was a really big deal.

Lawrence: Both as Pixar and in Steve’s life, it was a really big deal. Steve had had a series of failures leading up to Pixar. Before he left Apple, there was the Apple Lisa, and there was the Apple Macintosh. After he left Apple, there was the Pixar imaging computer. And then there was the NeXT Computer. So, those were four pieces of hardware that essentially failed in the marketplace. Pixar really was his comeback, as I recount in the story. But Steve and I hit it off from the get-go.

Sonal: Why do you think you guys hit it off? I mean, everyone’s heard that famous story in the Apple counterpart, where he’s saying, like, “Do you want to sell sugar water the rest of your life?” And this is, sort of, like, a similar thing where it could have gone that way, where there might not have been as much chemistry between you and him. What made you work?

Lawrence: That’s a really good question. Sometimes there’s just chemistry. And from the moment we met, there was just this level of mutual respect and trust, and it lasted for a long time. And it was both professional and personal. We just kind of got each other.

Sonal: That’s amazing.

Lawrence: And so the relationship started in this, sort of, you know, adventurous dialogue of this crazy thing called Pixar. Of course, I’m aware of all the other accounts of Steve. I’m writing about my experience.

Sonal: You had the opportunity to work with, arguably, one of our most influential innovators, who’s created things that have changed our world. It’s kind of an amazing opportunity. What were the moments that were most trying between you guys, because, obviously, it wasn’t all peaches and cream?

Lawrence: I think when Steve began to see that an IPO was possible for Pixar, he couldn’t get there fast enough.

Sonal: He couldn’t get there fast enough. Interesting.

Lawrence: He couldn’t get there fast enough, because that IPO was really the symbol of his comeback. He was rushing toward that faster than I was.

Sonal: Which is, by the way, really interesting, because in the current ecosystem, it’s been, sort of, the opposite, where a lot of founders are not ready to IPO yet. One of the things you said in the book is that, “Besides making films that would enjoy unprecedented box office success the world over, we simply had to, one, quadruple our share of the profits, two, raise at least $75 million to pay for our production costs, three, make films far more often than we knew how, and four, build Pixar into a worldwide brand.” Piece of cake. But really tough. That was your plan.

Lawrence: Yeah, that was the plan.

Sonal: So how did you guys go from there? Because the thing that I think is super interesting is that you said it’s not going to be easy for investors to get their heads around Pixar’s business. We have a lot of explaining to do. Tell me about that.

Investors and IPO

Lawrence: Well, I went to see my old friend and mentor Larry Sonsini, of Wilson Sonsini. And I presented all this to him. And he knew Steve really well, and he knew Pixar. And I said, “Investors are really gonna balk at this. We’ve got to disclose this — all this risk, up front.”

Sonal: Yeah, it’s a lot of risk.

Lawrence: So, I thought he was gonna say, “Forget about it. You have no chance. Don’t even think about it.” But he didn’t. He said, “You’re right. This is an incredible long shot. But if you disclose the risks upfront, completely, then you’ll see investors will make an open evaluation.” And that’s actually always been my approach to, you know, when I was in business, and being a CFO, which is — don’t hide behind the risks. You know, just put it out there. And don’t be afraid to put it out there. Because they’re going to find out anyway.

Sonal: And they’ll probably trust you more, actually, for being open about it.

Lawrence: And they’ll trust you more, so just be an open book. Because once you’re a public company, you’re gonna be in a fishbowl anyway. I said to Steve, “You know, we’re gonna have to disclose this risk, and these things that we have to accomplish that are so difficult, and why they’re so difficult, upfront.” Steve wasn’t against it, but he was kind of like, “Fine, fine. If that’s what we need to do, we’ll do it. But the investors won’t care about that, because they’ll see this incredible vision and this incredible possibility.” So, I had one foot on the brake while he was rushing toward it. We were heading into some very choppy investment waters, which we did.

Sonal: How did you navigate that, exactly? Because you have to also have a model that works for people to want to invest in you. One of the things I’ve heard is that — a big part is the story you tell on your way to the IPO. How did you guys tell that story?

Lawrence: The story was, in some ways, the easy part. By that time, we had figured out Pixar was going to become an entertainment company. So, we told the story of its creative capability, its production capability, its technical capabilities. We laid it all out — the talent and the capacity. That part of the story was the easy part.

Sonal: The easy part, okay.

Lawrence: It was easy to tell. You couldn’t help but walk into Pixar — even today. But in 1995, when you walked into Pixar…

Sonal: Before the new campus.

Lawrence: …before the new campus, and you start looking inside a little bit, I mean, it’s staggering. There wasn’t a person there that didn’t leave just, like, blown away. It’s that impressive.

Sonal: That sounds amazing.

Lawrence: It’s really impressive. So, that was the easy part. It was the numbers part. But we did have a model. And we said if these things happen…

Sonal: Those four things.

Lawrence: Right. We build a worldwide brand, we quadruple our share of the profits, we raise $75 million, we release films more often — if we do all those things, this company will make it, and it’s really up to the investors to assign a risk to that. You may think we have a 1% shot, or we have a 10% shot. That’s for you to decide. But if we hit those, then we will hit the ball out of the park. It was more than a prayer.

Sonal: Well, this is where the pricing of the IPO is a really fascinating narrative. Because you have a version of the story where it’s like, “Okay, if we do these things, these four things, we will hit these amazing numbers.” The investors are like, “Okay, yeah, that’s a great story, and great model and plan. And yay, we’ll give you capital to do that. But hey, we’re not gonna value very highly upfront that we believe in this.”

Lawrence: Well, I would say that the price of an IPO on the financial side of a company is one of the most important decisions that it will make. And this was also a point of contention between Steve and I, about how to price it. My thinking about it was that it’s important to leave something on the table. It’s much more important that early investors be happy, and feel like they made money, than they’d be disappointed and feel like they lost money. And so if you underprice a little bit and they feel like they’ve made money, then you get a lot of confidence in your stock going forward.

Steve felt that, you know, wherever we priced it, it would just skyrocket. And the irony — sitting here doing the a16z — is that the reason he thought that was because of Netscape. Netscape and Pixar were the two hottest IPOs of that year, in 1995, but Netscape went first. And they, of course, were the first company to ride this new wave called the internet. It was huge. I’m looking at Pixar saying that, you know, Pixar is amazing. But no one is sitting around talking about, you know, the animated film business. This is a 50-year-old business.

Sonal: And the best model, till that point, was 1939 founded company of Disney.

Lawrence: Yeah, Steve was saying that if they’re valued there, then we ought to be valued there. And I was like, “But it’s different. There’s no frenzy out there for animation.” We went back and forth, back and forth, but we worked it out.

Sonal: How did you guys work it out?

Lawrence: The pricing comes together through a whole combination of factors. You have investment bankers, and they’re having a lot to say about it. You’ve got your lawyers and the disclosures, and you have your board, and there’s a lot of elements. And so it begins to coalesce. We managed to get it into a bandwidth where I felt that we could make it and investors would be happy, and Steve felt that we had a shot at the kind of valuation that he wanted.

Sonal: I want to ask you about some lessons learned and high-level takeaways from that whole experience of going to your IPO, and things that may be different today.

Lawrence: You know, people see their IPO as like an end game, but IPO is the beginning of the game. It’s a change event.

Sonal: Did it make you better, because you’re able to execute on your model in the public eye?

Lawrence: I think if you’re disciplined, that it doesn’t necessarily have to make you better. But by the time we were in public, we knew what we were doing. We really did. The plan that we’ve talked about, that we put in place, lasted Pixar 10 years. It took 10 years to execute and we just sort of…

Sonal: Kept going.

Lawrence: …kept going, steadfastly. Went at it.

Sonal: How did you navigate the cultural divides, where you have the Silicon Valley-centered company, and team for that matter, technology, Hollywood, and the business of creativity — and then you have the East Coast, kind of, banking coterie? How did you navigate all those?

Lawrence: You describe Hollywood as the business of creativity. But in some ways, one of the things we learned is that sometimes creativity in Hollywood isn’t all it’s cracked up to be. The big studios have so many presses. Some of our thinking back then is, we have to protect Pixar from some of those kinds of pressures, in order for it to continue to innovate and do original films. This is a tremendous tribute to John Lasseter and Ed Catmull, who really created the creation culture of Pixar. And Steve and myself — I think our contribution, in a way, was to recognize the importance of preserving that.

The creative process

Sonal: I think that is actually a pretty tremendous credit to you as well, because when you’re responsible and accountable for the financial performance, there are a lot of short-term things you can do that short-shrift some of that long-term creativity. You get eager to see some results. And to keep your eye on the long term is a very difficult challenge. All companies face this today — this balance between being able to execute on a plan, but also adapt and innovate. What are some of the ways you’ve made that work?

Lawrence: Well, these were very big discussions. Steve and I would talk about this a lot. Because, going back to 1995, 1996, the creative team at Pixar, as brilliant as it was, was young and untested. “Toy Story” was the first film. Now these famous directors — Andrew Stanton and Pete Docter and Brad Bird, Lee Unkrich, the others from Pixar — they’d never made a film. And you’re looking at these young directors, and each film is gonna cost, say, an average of $140 million.

Sonal: And that’s really expensive, because they’re animation.

Lawrence: They’re animation. It’s really expensive. So you’re literally betting $140 million…

Sonal: On this untested talent.

Lawrence: …on this untested talent. So, the temptation to wanna go in there and make sure it’s going well is enormous, right? Because in a project of that size, slip-ups are very expensive. They come to you and say, “We made a mistake in the story,” right, you can’t fix that, like, the next day. That’s usually a $5 or $10 million mistake. So, the most important decision that we made at Pixar was basically to trust the creatives and not interfere with what they’re doing. For executives, that’s really hard for two reasons. One is that executives think they know. They think they’re creative, and they think they know. And some are, but it’s rare. And Steve and I realized that that wasn’t our domain. We could watch movies, we could critique movies, but we were amateurs and they were professionals. And so we had to trust them.

Sonal: But what gave you — I think it’s important to push on this — the ability to trust them? They were unproven. Was there some indication that you felt like, “We can trust them?” What were the signals? Because otherwise, anyone could say, “Hey, I want to trust the creatives.”

Lawrence: Well, it was very clear that John Lasseter was something extraordinary. And he had hand-groomed these other — sometimes we called it, like, the John Lasseter school of animation direction. But that said, even if you’re the best director in the world, you’re gonna have misses, right?

Sonal: Yeah.

Lawrence: He was that good, that you could place that bet. And I think you put your finger on the issue. Not everybody, in a startup company — you don’t have to have John Lasseter — but you have to be very discerning about the level of talent that you have. And I think that’s the challenge for executives. Not to think that they know, but to be able to make a real assessment about what the level of talent they have is.

Sonal: We see this every day here in our own business, because, essentially, the definition of venture capital is you’re betting on talent. The product may or may not end up where it started off, but you’re really betting on that person. And you have to make that assessment on a combination of factors.

Lawrence: Right. Once Steve said to me that he felt the decisions that we made at Pixar were not his, were not mine — or they were the product of this process. I think it’s very rare to find somebody to work with like that. And when you have it, one plus one makes way more than two. And that’s what our relationship was like. 

Personal takeaways

Sonal: Has it changed how you live your life today?

Lawrence: Well, I went off to do something completely different. I left corporate life behind, and I wanted to explore Eastern philosophy and meditation. I love my career, I love what I did, but I felt there was something one-dimensional about it. It’s very oriented towards success and performance and acquisition, which are great, and I have no issue with any of it. But we also pay a price for having that intense orientation towards performance at all costs.

Sonal: What’s the price?

Lawrence: One of those prices is, it creates a stress culture. I asked myself, why is it today that we hear so many stories about anxiety and agitation and mental health? And I think a part of this is because we have this sort of performance orientation without anything else to balance it. I have all these years at Pixar. And what I learned at Pixar is that it’s all about story. Then I go off and study Buddhist philosophy. And after all those years, I realized — it’s all about story.

Sonal: What do you mean?

Lawrence: Pixar, of course, it’s all about story because that’s what’s driving people to enjoy the film. In Buddhist philosophy, what it’s saying is that we’re living by a story. That we can’t always see it, but our life is driven by stories that are inside of us. And so…

Sonal: Actually, psychotherapists say the same thing.

Lawrence: And now neuroscientists are also saying the same thing. It’s fascinating. So, the performance story that we live in now — it’s not something inherent in the nature of the universe. It’s a story. It’s a culturally-generated paradigm. The Middle Way essentially says that we are mistaking our stories for reality.

Sonal: Why do you call it The Middle Way?

Lawrence: Because The Middle Way talks about two extremes. One extreme would be the extreme of believing that your story is real. And the other extreme is, because that story is so real to you, you cannot conceive of any other possibility — even if the story is hurting you. 

Sonal: Very binary.

Lawrence: Exactly. It’s very binary. And so to move away from your story, to move away from your performance orientation, is so threatening and, sort of, fearful that you can’t imagine life is even possible over there. So The Middle Way is about, how do you find that place in the middle? It’s not about giving up performance or giving up all of these things. It’s about harmonizing with all the different elements of life.

Sonal: Just one last question. On a personal level, there are so many interesting threads. How does this tie into companies at an organizational level, like, in terms of corporate culture?

Lawrence: I think it goes to the very heart of corporate culture. If you look at what’s driving corporate culture over — I’m gonna say 300 years — it started with the Dutch East India Company. By 1402, for 200 years, the Dutch East India Company ruled the world. And the mentality — the corporate mentality going back is basically acquisition at all costs. Over the decades, what we see is, basically, a battleground between corporations that are trying to acquire at all costs, and other forces around them that are trying to get them to pay for the costs. So, we have to have environmental laws, and anti-pollution laws, and anti-child labor laws. They’re just going for it. And that creates that mentality of success at all costs within corporations. And that’s the paradigm that we’re in. I believe that we could change that paradigm to a more humanistic paradigm that values the individuals that are in cooperation.

Sonal: I actually wonder if that story is changing right now, because the firm for the first time in 300 years is essentially being reinvented with all these new types of models, decentralized models, B Corp, social business. There’s a whole category of new things that are coming around now.

Lawrence: I think that’s right. And I think it’s fantastic.

Sonal: The theme to me — and this is sort of the theme of this whole podcast — is this human side to the business. It all ties together.

Lawrence: I’ll tell you this, the number one question that I get when I give talks is, “Do I have to be a jerk to succeed?” People ask me that all the time.

Sonal: That’s a question that comes up all the time. Do you have to be a brilliant jerk in order to be successful?

Lawrence: Exactly. I’ll tell you my answer. My answer is no. You don’t have to be a jerk to succeed, but it’s harder, because you can be a jerk and succeed.

Sonal: You’re, kind of, going against gravity.

Lawrence: And so, if you want to do it the other way, it’s up to you. It means that whatever you’re doing — you have to make hard decisions, you have to fire people, reprimand people. Whatever you do, you have to remember that you’re dealing with a human being. It’s a choice, and it’s one that we can all make.

Sonal: That’s great. Well, Lawrence, thank you for joining the “a16z Podcast.” And for everyone who wants to read more about these adventures, read the book. Thanks, Lawrence.

Lawrence: Thank you for having me.

• 

  Sonal Chokshi is the editor in chief as well as podcast network showrunner. Prior to joining a16z 2014 to build the editorial operation, Sonal was a senior editor at WIRED, and before that in content at Xerox PARC.

  Follow Twitter
• 

  Lawrence Levy

“Incremental change may be good theory, but in practice you have to have a big enough stick to hit everybody with to make everything move at once”. So shares Adrian Cockcroft, who helped lead Netflix’s migration from datacenter to the cloud — and from monolithic to microservices architecture — when their streaming business (the “stick”!) was exploding.

So how did they — and how can other companies — make such big, bet-the-company kind of moves, without getting mired in fanatical internal debates? Does organizational structure need to change, especially if moving from a more product-, than project-based, approach? What happens to security? And finally, what happens to the role of CIOs; what can/should they do?

Most interestingly: How will the entire industry be affected as companies not only adopt, but essentially offer, microservices or narrow cloud APIs? How do the trends of microservices, containers, devops, cloud, as-a-service/ on-demand, serverless — all moves towards more and more ephemerality — change the future of computing and even work? Cockcroft (who is now a technology fellow at Battery Ventures) joins this episode of the a16z Podcast, in conversation with Frank Chen and Martin Casado (and Sonal Chokshi) to discuss these shifts and more.

Show Notes

• Discussion of how Netflix moved to a microservices architecture [1:26]
• Security advantages of microservices [8:13], and the general trend toward this architecture in the marketplace [14:21]
• How development teams and businesses stand to benefit from this shift [18:34]

Transcript

Sonal: Hi, everyone. Welcome to the “a16z Podcast.” I’m Sonal. Today’s podcast episode is all about microservices. And I’ve been super eager to focus only on this topic on the podcast, since we mention it a lot in passing, and I’m really excited because we finally get to do that. Our special guest for this topic is Adrian Cockcroft, who helped lead Netflix’s migration to a large-scale, highly available public-cloud architecture a few years ago — making Netflix one of the originators and early adopters of microservices. And Adrian is widely credited for helping pioneer microservices at web-scale. 

Also joining in the conversation are a16z partners Martin Casado and Frank Chen, who will be moderating the discussion. And in this episode, we cover everything from what [are] microservices, to the evolution of the architecture, to how it changes the shape of organizations, to operations, to changing the role of CIOs. And finally — and this is actually what really excites me the most about this topic — is what new opportunities come up when you have these extremely ephemeral systems that are, you know, just like ghosts in the machine — from containers to servers on-demand, to serverless and what’s happening there, and some really interesting trends on that edge. The conversation begins, however, with the story of how Netflix got into microservices.

Moving to a microservices architecture

Frank: Take us back to the days when Netflix had decided they were gonna move to Amazon and commit to a microservices architecture. Let’s pick up the story there. So, what’s it like inside?

Adrian: We started off basically running away from a monolith. We had over 100 people every two weeks trying to get all the code they’d written in the last two weeks jammed into one codebase, get it through QA, and get that out into production. And that was just getting more and more painful, and we basically decided we had to break it into pieces. You wanted it to be the work of one developer, basically, controlling what they had deployed independently of everybody else. And at the same time, we weren’t looking at moving to cloud.

Frank: Did you make both big moves at once? In other words, monolith to microservices, and then private data center to Amazon?

Adrian: Everything together. And sometimes you find incremental change a good theory, but in practice, you have to have a big enough stick to hit everybody with to make everything move at once. And the big stick was, we didn’t have enough data center capacity to support streaming. We were running the DVD business in the data center, on a system that was growing at a respectable rate. But the streaming business was exploding at a much, much higher rate. And because of that, we knew we would have to either build lots of big data centers, or get onto something else. So, the bet was, “Okay, we need to go on cloud. Then what’s the right architecture for doing that? What’s the right organization for doing that?” The developer group is getting bigger and getting less productive, and we wanted to unlock the innovation. So, we were simultaneously trying to get better developer productivity, better time to value — which is one of the key things we’re trying to optimize for, generally. And then there was a whole bunch of other cloud transitions bundled in.

Frank: As you went from the monolithic application to microservices, what did that entail? What’s that mean? What is a microservices architecture?

Adrian: Well, originally, I called it fine-grained SOA, service-oriented architecture. And there’s a lot — some people get negative reactions to SOA, because they were out there trying to do it 10, 15 years ago.

Frank: That’s right. So it’s all the same ideas over and over again with new dressing.

Adrian: Yeah. It’s a question like, “Why now, and why didn’t it work then?” And if you look at it, what we were doing was — on relatively slow CPUs compared to what we have today, on relatively slow networks, we were processing big fat lumps of XML and parsing it around. And we were really only able to break the application into a few large chunks, because the overhead of all of the message parsing was too high. If you come to today, you know, you can break it into maybe 100th of the size and 100 times as many chunks, because the overhead of the communication is now very low. We’ve got binary protocols. We’re not trying to, sort of, make everything conform to the big SOAP XML messaging schemes. So it became possible to build a fine-grained SOA architecture, and that ended up being called microservices by, I think, Fred George, who was the first to use the word. But it got written up by Martin Fowler, and then everyone said, “Okay, we’ll go with that.”

Frank: Yeah. So, big-bang moves. This was a bet-the-company set of technology decisions. Looking back at it, what are some of the lessons learned?

Adrian: I think one of the ways to approach this is to basically create, kind of, a pathfinder or a pioneer team. There was a lot of controversy inside. So half of the company thought this was stupid, and a few of us thought we could make it work and other people a bit more gung-ho. So, we got the people that thought they could make it work into a room and had a one-day project, where we all built a thing in the cloud to see if it would work — built out of the kind of technologies we’d need to use to build this. That team then, sort of, knocked down a bunch of the straw man arguments that everyone else was holding up against us. You know, a lot of the time, it is just straw man arguments, but you have to actually go and build something to actually find out what are the real arguments. Then you discover things you didn’t even know, which are hard. You run into the real blockers, as opposed to the imaginary ones. So, I think the trick is to get a small team, go very deep, discover what you can, and run a whole bunch of these little projects where you’re trying to learn as much as possible with the smallest possible input.

Frank: You had this cultural aha, which is, “Let’s get the people who are gung-ho about this, and let’s let them go deep, knock down the straw man arguments.” Sort of zoom up to the 30,000-foot view and sort of describe the organization at Netflix before and after. What did it look like before and after, from a skillset point of view, from an organizational design point of view?

Adrian: This is actually one of the big things that makes a difference. Some organizations are set up already to do microservice-based architectures, and others have to go through a reorg. Netflix emerged naturally out of the way we were structured at the time. We were already structured as small cells that own things, a lot of responsibility. Each team had a very clear idea of what it was building and how it related to the other teams. But it was assembled as a monolith, at the end of the day. So, breaking it apart was a fairly natural thing for us to do. What you see with traditional enterprise siloed organizations is they’re actually having to do a reorg, and set up teams that are responsible for services, and it’s somewhat unnatural for the way they’re currently set up. But I’m seeing an increasing number of people go through that transition. And sometimes you can see it as replacing project-based work with product-based work. So, every team becomes basically a product team for their microservice, and you have the product management aspects and the operational aspects within that team.

Frank: And did you find that the people who are used to working on the monolith could be retrained, or did you have to have a new crew come in?

Adrian: The culture at Netflix is interesting. Most of us had been around before. A lot of us had worked on SOA. You know, we’re gray-haired people that had been — there’s a few people that worked at Xerox PARC in the 1980s, and you could go and have arguments within their object-oriented programming. We had some younger people, but it was a lot of very experienced people taking all the stuff they’d learned and synthesizing it together. It was a very collaborative experience. And we came up with things that made sense based on this series of transitions we were going through. The other transition was from a single centralized database. We had this enormous Oracle machine, with a really complicated schema, to a distributed NoSQL database, in the end, based on lots of different Cassandra clusters. And that was the third transition, and that was probably the hardest transition — was getting all of the SQL code and transactional stuff out of the system. It’s actually breaking apart the databases, probably the hardest thing to do — and then splitting chunks of code off is also difficult if you’re trying to pick apart a monolith. And it turns out, if you don’t break apart your database backend, and you just create lots of services that talk to it, you’ve actually created what’s called a distributed monolith, which has all the same fragility of the monolith, and you can’t update things independently, because you’re tied by the database.

Security advantages

Frank: You can’t just take the Oracle database and break it up into little pieces. You have to think about it differently. Now, the same thing is true for the rest of the architecture as you migrate to microservices.

Martin: Yeah. So, I think what excites me about microservices, in general — it moves all of infrastructure up to an application layer. So, if you think about what you normally do in infrastructure, you’ve got these basic abstractions. Like compute and network and storage, which are pretty low level and they’re semantic free, right, you don’t have structured data. One of the huge advantages of going up to a microservice architecture is you can do infrastructure insertion. Things like, for example, security — things for, like, you know, even debugging — basic operations, and management. And you can do it in a way that has the deep context and semantics of the application.

The point here is that not only are you going away from the monolith, which is really important, and I think it’s great, but also, like, you’ve got more semantics than you’ve ever had before. I mean, this is actually meaningful stuff when you’re dealing with not IP headers, for example, not blocks but actual, like, structured data. And I think that we can actually reimagine a lot of these tools in ways that we’ve never thought of them before, because we’ve never had the ability to have this type of semantics in these toolchains. We’re seeing this burgeoning area of microservices where you almost have, like, a function per company coming up, and now, I believe that all of the old stuff that we had in the internet, whether it’s naming, or service discovery, or routing, or whatever, we’ve got an opportunity to bring this up in, kind of, a much deeper, richer level, which is really cool.

Frank: Right. So, we were going to the marketplace or the bazaar away from the cathedral, which is, any individual function can be provided by either an internal or external provider. It could be a cloud service. But then, the challenge is, now it’s up to every organization to coordinate, right. And so what are some lessons that you guys have learned along the way of picking best-of-breed and then making sure they work with each other, getting the version control to work?

Adrian: When you’ve got a monolithic app, everything is in there. If it gets broken into, you have all access. Its connection to the database lets it basically say anything to the database. When you break things into microservices, you’ve got the ability to have some parts of your system be low-security risk and other parts be high-security risk. You can innovate really, really quickly in areas of, sort of, personalization and user experience. And then you maybe have a much more tightly controlled thing for, say, the signup flow and where you’re storing personal information.

Frank: So, the great news is you have a lot more agility. The price that you pay is you’re doing a lot more coordination. With a monolith, it’s easy. You put all your eggs in one basket, and then, from a security point of view, for instance, you basically just pile a bunch of appliances in front of it. Easy, right? Because it was a monolith. You knew exactly where it was. Now that the perimeter is distributed across many machines, you have to be a lot more mindful of where the attack surface has gone and which security service you need to put in front of that part of the microservices architecture.

Adrian: So, you cannot have the privilege escalation of “because there is a little bit of PCI compliance needed in one tiny corner of this monolith, the entire monolith is now subject to PCI compliance and SOC 2 compliance,” and all these things. And by splitting it up into pieces, you can have most of your app be extremely agile and very innovative, and then have the bits that need to be safe be extremely safe. And then if you look at the attack surface, you’re basically keeping a very tight control over what can do what. And if you connect them to the databases, you’ve got very single-purpose connections into the database that are doing one thing, and you can start to control at the access level there as well.

What used to be policy-controlled by the operations people — what they felt was a safe sandbox for the developers — is now really being driven from the other end down. So this idea of developer-driven infrastructure is something that is turning things around. And a lot of what I’m seeing is that big banks, and people like that — they have their existing policy frameworks and rules, and they’re trying to apply it in the new world, and it looks the same, so they’re happy because they’re compliant. But they don’t actually have the real policy separation that they think they have, because it’s all totally reprogrammable, and it’s like you have the illusion that you’re still conforming to the policy.

A lot of these things were Ops-controlled. So the Ops would control the data center, and then the networks in the data center, and now it’s all developer-defined and software constructs which are controlled by your, you know, cloud APIs. If you’re updating it 10 times a day, there isn’t time to have 10 meetings a day with operations to do the handoff. So, what we’ve been seeing is, people just running it themselves. The only person that knows the exact state of the system is the developer that just updated it. That sounds scary until you realize that each of them is controlling a very small piece of the system, and the aggregate behavior of the system turns out to be really robust and reliable — partly because if you put a developer on call, they write really reliable code, and they don’t release code on Friday afternoons, because they want a quiet weekend. You know, they learn a bunch of practices about what it’s like to be on call and how not to break things.

Frank: So we went from an in-person change review board, infrequently, right, to vet the changes to continuous change and, “Hey, let’s coordinate over Slack.”

Adrian: Pretty much. Yeah, you have to tell people what you’re doing, but you don’t have to typically ask for permission and go and have meetings and things like that. This is part of unlocking the innovation. And the people that are most interested in these are large teams of people trying to build complex products, typically enterprises, and they are worried about getting disrupted by the latest Bay Area startup or whatever. There’s an existential threat here, that if you’re doing quarterly releases and your competitor is doing daily releases and continuous delivery, you’re gonna fall so far behind in the user experience that you’re just gonna suffer, right. So, that’s the big driver that is making people say, “Well, how do you get there?” There’s a whole bunch of things tied together. You’re bringing in cloud, DevOps is a whole other area, and microservices as an architecture — all these things tied together — and some cultural change as well in the organization of the company. The companies that are doing well at that are really starting to accelerate off into the distance.

Market trends toward microservices

Martin: It’s also worth teasing apart two trends. And one of these trends is, you know, a single company, instead of building a monolithic product, wants to build a microservices product, and gets all the efficiencies of doing that as far as the development process and the OEM process, everything else. But there’s kind of a broader industry trend where companies’ products are basically microservices, right? There’s companies out there that, like, basically, the only way to access the product is through a fairly narrow API. I mean, you know, there’s so many of these now that there are other startups that will just basically stitch them together, and they could build full applications without writing much code. So, I think that, in addition to a single company getting a lot of advantages, I think the entire industry is gonna get a lot of advantages and see a lot of innovation as a result.

Frank: Yeah. If you had said five years ago that there would be multiple independent public companies that all they do was offer an API, you would have been laughed out of the room, right? And now, look at us. Twilio, and Tribe, and on and on.

Martin: I like to do the mental exercise of, kind of, where this is all going, and I still love Chris Dixon’s quote of, you know, “Every Unix command becomes a company.” It’s like grep becomes Google, whatever. Like, I think, you know, we may be having an analog here, which is every function becomes a company, right? It’s, like, even more granular than a command line tool. Every single function, or a logical function, becomes an independent company. And I do think there are implications on things like ownership and dependability, and stuff like that, that we haven’t grappled [with] yet as an industry. It’s a very exciting direction.

Adrian: Yeah. You’re able to build something now that pulls in things from APIs and pulls in some containers, and you just have your little piece of code in the middle that stitches it together and build a completely new service from that. So, it’s just much easier to get things built. It’s more efficient for the big companies, but it has democratized all the way down to pretty much anybody with a laptop can go build something interesting. And if you go back 5 or 10 years, you’re doing things that would be just totally impossible to try and get together at that point. There’s much more room for innovation. 

It also makes it harder to compete, in some ways, because now it’s hard to build, you know, a billion-dollar software company on top of these things, because they keep changing underneath you, and they’re cheap to build. So, you’ve got lots of disruption coming, and it’s actually, you know, GitHub, and open source is another big player in here that’s just making it much lower cost to get things done. So, what you’re seeing now is Twitter, and Facebook, and Netflix, and Google, and LinkedIn producing the stuff that you actually want to use, which has already been tested at volume, and then it’s actually much harder to build a proprietary software company because you’re competing with these big end users, and you’ve got this thing you’ve just built, and it’s flaky and don’t quite work right.

Martin: We’ve talked about this, but it seems like closed-source shippable software is on its way out or dead. And there’s a number of reasons for this. One of them is just — the enterprise buyer likes open-source software, but another one is it’s a real burden on the company to ship software, right. I mean, especially if that software is a distributed system, right. I mean, like, you don’t have skilled operators often, every environment is different, right. So, you’ve got these heterogeneous deployment environments. You end up with this, like, the mother of all cache consistency problems, where you’ve got a bunch of versions out there, a lot of products you’ve got to maintain a bunch of versions, etc. It’s hard.

Frank: The QA matrix from hell, right? Oracle’s version multiplied by the flavors of Unix multiplied by whatever Windows versions you’re supporting, right?

Martin: Yeah, that’s right, that’s right.

Frank: Your poor QA manager.

Martin: Yeah, that’s right. And then distributed systems, generally, I mean, a real trick if you’re running your own operation is you have skilled administrators that know how to manage a cluster. And that, like, there are very, very few companies, and I think maybe one that’s actually managed to ship a distributed system that was manageable with a non-skilled operator. It’s a very, very difficult problem. And so a great thing about — if you offer something as a service is, like, okay, you don’t have any of these problems. And so, like, basically, your post-sales operation budget is way lower. It’s much easier to start a company now, but at the same time, there are questions about, “Okay, so what are the sizes these companies are gonna end up being? I mean, how big is the market for a single function?” I think it’s still to be seen, like, how big these companies are gonna become.

Frank: Yeah. Big challenge from an investor’s point of view, which is, if the essential argument is, “there will be no more cathedrals, it’s all bazaars from here on out,” it’s a little harder to make money, right, because the biggest…

Adrian: You’re investing in a food truck, and that’s as big as it’s gonna get.

Advantages for development teams

Frank: So, put yourselves in the shoes of the enterprise CIO. The pace of change is accelerating, right. The ink just dried on her team getting VMware certified. And now, we’re on to containers, and then people are talking about serverless and functions as a service with, sort of, Lambda architecture. So, talk a little bit about what’s coming, and then the ability of an average organization to sort of absorb these changes.

Adrian: Containers came along, really over the last two years, and it’s one of the fastest takeovers of enterprise computing we’ve ever seen. It’s quite remarkable how quickly they were able to colonize the enterprise space. It solved a real problem.

Frank: What role did containers play in moving away from the monoliths to the microservices architecture?

Adrian: What happens with the containers — all that stuff is packaged into a bundle which has all the right versions of everything inside it, and you can download it and run it. It also abstracts you away from the particular version of what you’re running on. There’s now containers for Windows as well. But originally, this was a Linux-based concept. You have the same container format if you want to run in-house, or on a public cloud. It doesn’t really matter. That container can run on VMware, or KVM on OpenStack, or on Amazon, or Google, or Azure, or wherever, right. You’ve just abstracted yourself up one level. It gives you that kind of portability. If you think — about machines used to sit at the same IP address for years. People would know a machine — they would actually know the IP address of by heart if they wanted to do something to it, right?

Martin: Well, I remember that. Yeah.

Adrian: And then you had VMs came along, and now the VMs are more transient, and you know, this thing would come and go, maybe in, you know, order of weeks or something, a biweekly update of your VM. And then, with containers, it’s perfectly reasonable to have a container that runs for less than a minute. You can create an entire test environment, set it up, run your tests, you know, automatically test it, strip the thing down again, and the size of the things have gotten much smaller. If you just take it to its logical conclusion, we’d basically fire up effectively a container to run a single request and have it sit around for about half a second and then have it go away again. And that’s really the underlying technology behind AWS Lambda. It’s a server on-demand that just isn’t there most of the time. And this is the bleeding edge right now. We have to figure out how to extract these, sort of, ghostly flickering images that are sort of coming into existence for short periods of time. How do you track what’s going on? You end up figuring out how to end-to-end tracing as the only way you can monitor things, rather than being a special case like it is now.

So, there’s a bunch of interesting problems here, but what’s really been happening is just this trend to more and more ephemerality. And these extremely ephemeral systems — and then the charging. Used to charge by three years’ worth of machine, and then it became, well, you can rent a VM by the hour. And then containers, you know, that’s lighter weight, and now you’re paying by the hundred milliseconds, right? It’s perfectly reasonable to run for half a second, which means that the setup time to create that half-second worth of machine needs to be radically less than half a second. And the time taken to bill for it needs to be less than half a second. If you remember the story of SMS, the SMS record for, you know, 140 characters — the billing record is much bigger than that. It’s more like a kilobyte. So, if you actually take a telco and rip out all of the billing step for their SMS things, you know, it will cost 1/10 of the amount to run if they didn’t bill for it. So, you got this effect that the overhead of doing the thing is actually vastly more than the thing you’re trying to do. So, actually, it’s a really interesting challenge — is how to create monitoring and billing and scheduling systems that work so quickly that you can afford to bill things in timed increments.

Frank: The portfolio company 21 is, sort of, right in the thick of this, right, which is how do you stand up an ad hoc agreement between an API and an API and, like, have the billing all work. And you know, Bitcoin might play a part in that.

Martin: Also, to your question, going back to the CIO — I mean, it seems to me, in general, with disruptive technologies, it’s like — the disruption happens first and then all the day-2 Ops happen second. I mean, whatever that is. And, I think, in this case, you know, the disruption is around delaminating the app and breaking it apart. I do think that CIOs should not despair and Ops team should not despair, because what happens very quickly in the vacuum being left from kind of, you know, this sprint on these new technologies is whole, you know, ecosystems and whole industries arise around them to provide visibility, to provide security, to provide Ops, and we’re seeing that now. And so, I mean, I think that it’s quite possible to decouple the disruption — which is this velocity around development — and then, you know, the basic operations. And that tooling is definitely going to happen as well. Understanding that ecosystem, understanding the players is very important if you wanna stay on top of this kind of big change.

Frank: Leaning forward into the change, assuming the tooling will meet you halfway, right.

Martin: Exactly right.

Frank: And then you get the benefit — the big benefit from the CIO’s point of view, in my opinion, is that you don’t have this loop where the business user asks for something. It took you 15 months to build it, only to discover that’s not what the business user really wanted, because the requirements were poorly specified. In these days, right, no problem. I’ve got a change for you, we’ll put it live this afternoon. Right? So, the rapid experimentation that happens in startup plan can now migrate into the big organizations, and you don’t have to get your requirements perfectly specified at the beginning of a waterfall process anymore. Let’s run the experiments.

Adrian: It’s actually even better than that. What the CIOs are providing now is a set of APIs for the development team that is part of the business to automatically provision whatever they want, with certain policy constraints around it for what they can and can’t do. But fundamentally, you’re providing APIs. Operations has moved from being a ticket driven organization to be an API. They are now no longer a cost center. That is a very profound move, and I’m seeing a lot of these CIOs buying into that. They want to be part of the product. They want to be — how do you support the business? And you provide APIs so that they can just get business done at a rate that you’re not slowing them down.

Martin: We’re actually seeing the creation of a new buying center in the industry of vertical platform engineering, of vertical DevOps team, whatever. This is, like, budget allocated. It’s actually viewed as a profit center. It’s product aligned, but it’s core infrastructure and operations. And these are very technical buyers, so it’s not the traditional enterprise go-to-market.

Adrian: This is also moving across industries. We’ve seen, obviously, media and entertainment and, to some extent, retail were early movers — mostly because of the threat of Amazon themselves causing retailers to step up to, sort of, reengineering. We’re now seeing FinTech, you know, or Wall Street is really paying attention. Some people are way down the road, some people are just starting. Manufacturing, that whole industry is just starting to think about this. There’s definitely a, sort of, industry-by-industry, sort of, domino effect as people are figuring this out.

Frank: So, we’re a decade on or so into this revolution, right. Many strands. What excites you now?

Martin: For me, what’s really exciting about this, and I’ve said this before — is if we just have the ability to reimagine all of infrastructure, you can now reimagine tooling, and reimagine security, and reimagine operations and management. We get to reimagine it with more semantics and context than we’ve ever had, you know. So, what does it mean to have a firewall in a world where everything is microservices? What does it mean to have operation management, and debugging — things that were traditional boxes, that were stuck on perimeters, now also become functions? And actually managing your infrastructure is almost like looking at a debugger, a context debugger. It’s like you have a symbol table with you. It’s, like, this whole thing is in one large IDE, and you can do that for your operations. I think it’s gonna push the state of the art on how we even think about Ops in entirely new areas. I’m really excited about that change.

Adrian: I think that the whole serverless area is the bleeding edge right now. The monitoring tools industry is, right now, bring disrupted pretty heavily by serverless. There’s only one or two tools that have really come into existence in the last year or two that have — effectively, a way of processing stuff that is this ephemeral and dynamic. So, there’s some interesting products coming out. It’s just a better way of living. If you’ve a developer and you’re working in the waterfall, siloed organization, it’s kind of soul-destroying for a lot of people, right?

Frank: Indeed.

Adrian: And when you get ownership of a product, you know, on distributed teams, you get each distributed team their own product ownership, and they get to define the interface and manage it and run it — yeah, you might be on call, but you’re in much more control of your destiny, and it’s much more rewarding, and it’s more productive. And the ability to get more stuff done as a developer is just rewarding anyway, right. It’s a better way of working for people.

Frank: Well, that’s great. Well, thank you, Adrian. Thank you, Martin. We’ll see a lot more unfold as the architecture shifts.

• 

  Adrian Cockcroft
• 

  Frank Chen is an operating partner at a16z where he oversees the Talent x Opportunity Initiative. Prior to TxO, Frank ran the deal and research team at the firm.

  Follow Twitter
• 

  Martin Casado is a general partner at a16z where he invests in enterprise companies. Prior, he was cofounder and CTO of Nicira (acquired by VMware) and is the creator of the software defined networking movement.

  Follow Twitter

“Raise prices.” Regular listeners of our podcast have heard this advice more than once. But why is this so key and yet so hard for many technical founders? And how should startups go about raising prices — or more specifically, creating value — for their products?

In this episode of the a16z Podcast, former sales VP Mark Cranney (and head of a16z’s EBC and go-to-market practice for startups) and former startup founder (and general partner focused on all things infrastructure) Martin Casado talk to managing partner Scott Kupor about pricing for startups … especially for category-creating businesses. It’s not all “pricing, pricing, pricing” though — there’s another important “p” in there too!

Show Notes

• Why it’s important to price aggressively from the outset [0:43]
• Using salespeople to gather information about the marketplace [10:07], and how to choose the right salespeople [16:50]
• Discussion around packaging products [20:57] and how to handle pricing errors [25:51]
• Scheduling pricing reviews and other sales strategies [29:16]

Transcript

Sonal: Hi, everyone. Welcome to the “a16z Podcast.” I’m Sonal. Regular listeners of the podcast have probably heard us say more than once that entrepreneurs should raise prices. Why? And the other question that comes up often is, how? On this episode of the “a16z Podcast,” we talk all about pricing, packaging, and more. Joining the conversation, our general partner, Martín Casado, who was formerly the co-founder and CTO of Nicira, which was later acquired by VMware. And most recently, he served as a general manager for one of their major business units. Also joining is Mark Cranney, who heads up our go-to-market practice and our EBC, or our Executive Briefing Center. And he was a former VP of sales. And moderating the podcast is managing partner, Scott Cooper.

Price aggressively

Scott: Hello, everybody. This is Scott Cooper. I’m here with Martin and Mark. And we are here to talk about pricing.

Martin: One of the reasons that this podcast, in general, is so important is, I think that pricing is one of the least intuitive things for technical founders, having been one. And it certainly was my bias, you know, coming out of, like, my Ph.D and doing a technical startup, that you almost always want to, like, take whatever your technology is and get into everybody’s hands — and then you kind of assume later on that, like, you can somehow monetize it. Ben Horowitz was on my board at the time. And I was explaining, I said, “Well, I think that we should kind of enter at a very low price, and that way we’ll have more people that use it.” And he looks at me with his very stern look and he says, “I want you to be very careful, because no single decision will impact the valuation of your company more than the decision you’re about to make on pricing.” And so, that kind of started my foray into pricing.

Scott: As long as you’ve got the mic, I mean, let’s start there. So, what’s wrong with that? So, why not be cheap? And then, why can’t you just raise prices later? What’s the problem?

Martin: So, there’s an interesting thing, especially if you’re dealing with pre-chasm-type markets. So, pre-chasm means you’re bringing a product to market where there isn’t a market yet, or the market is very immature, right? So, there may not be a budget. There may not be a buyer. They may not know how to think about it. And so, here’s the fallacy that many technical founders have, and I had as well. I think if you build technology, there’s intrinsic value to it. This is worth so much money because it’s intrinsically valuable. And that’s not generally what happens. What happens, just in human psychology, is actually, like — [people] will set the value on whatever they’re getting based on how they acquired it. And so, in early markets, nobody really knows how to value what you have. And so, it’s very important for you to establish the value in the market. Otherwise, you end up devaluing yourself right away, and you’ve just cannibalized your top-line revenue almost immediately.

Mark: The scenario you described to begin with is extremely common. And the reason it’s common, over and above what you’ve outlined is, in a lot of cases, that technical founder has not put themselves in the buyer’s shoes to understand what that prospective buyer needs to go through. Because like you described, they don’t know what the criteria is, and how they should look and/or evaluate your solution. They might not even know it’s a solution, because they don’t know there’s a problem. That’s kind of the job of sales and marketing — is to go put yourself in that buyer’s shoes, understand what their as-is environment is. 

Answer those questions along that sales process of, “Why should I even do anything, to even evaluate you or look at you?” And then “why you” and then “why now?” The “why now” piece is the piece where the pricing really starts to come into play, and that’s where you’ve actually gotten enough information out to understand what their return would be — what an ROI would be. And if I can’t go show what that value is, and not have that deep information about what that prospect’s costs are gonna be or what my impact of their business is gonna be, then, yeah, it’s easy to say, “I’ll give it away for free and we’ll monetize later.” So, that’s where that comes from, that natural inclination to, “I’ll figure this out later.”

Scott: Mark, so, you’re kind of describing, kind of, value-based pricing, right? So, does that mean in those early days, then, that you kind of iterate on the pricing with the customer, or are you suggesting that people do that analysis and then present that to customers, or is it really just iterative.

Mark: Well, I mean, a lot of it depends on what exactly your — you know, what is your value proposition. If there’s some kind of standards that are out there, and the customer is used to acquiring a solution, you know, and your solution is gonna be similar, there’s models you can bounce off as far as what’s going on in the as-is world. What Martin is talking about — what he was doing was so transformational, it was hard for a customer or a prospect to get their arms around. And Martin maybe didn’t have enough of, you know, the “as-is” — just intuitively, having been in these larger environments, to really, you know, early on understand what that would mean. And there’s a long development cycle before that value would even be delivered. Right? So, the value-based pricing, I think, is for sure true. But part of it’s just — you gotta go through a whole sales process before you can start to get comfortable and have pricing control.

Fast forward a little bit, you know, where you’ve gotten or you’re down the path of getting product-market fit, or you’re beyond that — you know, I see a lot of entrepreneurs that kind of get stuck in the mud on their pricing, or they’ve gone in too low, they’re not sure how to go start to raise pricing. A lot of that is more in the packaging. What is your roadmap? What sets of functionality has the prospect or the customer already realized value from? And then where are you going with the product? Is that something you wanna just charge the same price for, or do you need to start chunking that up, because they might be buying different ways?

The customer or the prospects asking for additional, you know, features and functionality, or scalability, or architecture <inaudible>, that shouldn’t be the same price. So, I need a different type of packaging and a different pricing model to go build on that. And then there’s the option of, like, “I wanna start a la carte,” versus, you know, in a package all at once. Some customers might not need the whole product when you’re first starting out, or they don’t need it in a user workgroup setting. But the enterprise, as they adopt, will. So, you’re holding some functionality back and waiting for them to want that, and then go establish that value to do it.

But I totally understand where Martin is coming from early on, because I see it over and over and over again. And it is, in a lot of cases, the value of the salesforce. The other thing I sensed in his initial answer was, you know — I don’t wanna go, you know, spend the time or the money. And a lot of that is putting the boots on the ground or, you know, inside in the marketing to go, you know, pay for that. But you’re paying for it either way. Right? You’re paying for it by giving everything away for free, because you’re not investing in sales and marketing — or you’re gonna have a channel partner do it, which you’re giving your margin away. Somebody is paying for it. And if it’s something that’s just so new and differentiated and not defined, and it’s gonna require, you know, the customer or somebody’s gotta go in there and kind of really tease out, “How are you doing things now? What is the pain?” They don’t know there’s a better way of doing it. That’s kind of our job as entrepreneurs, and/or as a sales and marketing organization, to be the translator between, you know, kind of the old world and the new world. And that’s where you start to understand, you know, how to piece this thing together from a value-based standpoint.

Martin: I think you can roughly, like, dissect the world into two pieces. You’ve got, you know, market category creation. Your constituency, whatever they are — they wake up in the morning and think about everything, but not your thing. Right? It’s not even something that, like, they consider. And there’s mature markets where you’re entering an existing market where pricing has already been set through, you know, a lot of transactions that have happened. And so, I think in the, kind of, post-chasm or mature market world, I think that there is existing pricing. There are comparables that you can work. But in the pre-chasm world, the thing doesn’t even exist. So, not only you’re describing that your thing exists, but you’re actually trying to attach a value to it. I’ve gotten so many times the question — it’s like, “Okay. So, how do you set that initial price?” You get a lot of these PMM types, they wanna go do this market research, and, like, all the stuff. But it’s very difficult to do research on something that doesn’t exist.

I mean, if you think about a lot of pre-chasm work from the technical side, you’re really being prescriptive, like, you’re not really asking the customer what they want. So, here’s my experience, you know, over a couple of products. The only way I’ve been able to establish pricing in a pre-chasm market is, you start pretty high and then you let the salespeople shake it out. You’ve got really, really good salespeople that go in there, have the dialogue, have the discussion, understand it. And it’s this really iterative process, where you’ve got the sales guys piped into the nervous system of the product development side, and then you get a sense for, kind of, what the market will bear. I don’t think you can have, you know, a bunch of MBAs out there doing research, because this is so new. And so, I don’t know, Mark, like, this is something you’ve done a lot of. I’d love to know your thoughts on that?

Mark: Well, I definitely agree with the start high. It’s way easier to go down than it is go up. When you talk about pre-chasm, the first thing you wanna do is really segment and target and get to those point-of-the-spear-type prospects. They’re gonna be quicker to understand that there’s probably a better way, and maybe have gone down that build-it-yourself-type path where they’ve got the recognition that there is another way. That’s where, you know, maybe kind of one of the first places is to start, from a segmentation and targeting. 

They’re gonna be those early type customers, as you’re going through that product-market fit, where you need to be able to understand what it could mean financially, and to also partner with them on what that pricing could be — because they’re the ones where you’re gonna get the most information and input, as far as what the value is gonna be. And it’s always gonna be involving — from a pricing standpoint, in some cases, we may be lowering the price of, like, the foundational piece of our technology, because the market is changing, right? Things are gonna get more commoditized, but we’re racing upstack and charging for new things that are more valuable. I mean, if you think about the whole lifecycle of pricing, you know, in a lot of tech companies, it’s not just where it starts, you know, it’s something you’re constantly — should be looking at.

Gathering intel from salespeople

Scott: So, just to put some meat on the bones, then. So we’re saying, “Look, start high,” you know, to Mark’s point, which is, “Look, it’s always easy to go down.” It is an iterative process. But within that context, there is some framework for how to evaluate the value-based pricing, right? So, in the Nicira case, presumably there are things that engineers can now do in terms of changes to the network that they couldn’t have done before. And that has demonstrable business value, or there’s either fewer resources that are required. There’s some kind of framework where we can begin to work with the customer to say, “Hey, the value that you would get in the organization, either in headcount cost reduction, or flexibility, or new product rollouts or other things, equates to some portion of that being captured through the software.”

Mark: Well, one thing, maybe there, just to keep in mind is, from a pricing standpoint, you know, if you think about a large enterprise — if you break the audiences down, you know, the pricing to, like, a user, like, say in Martin’s case, you know, like, the early people — he was probably talking to network engineers. That kind of intel is gonna be completely different than if you’re up at the top of an organization with the CXO that has a wider view and, you know, is gonna understand a bigger story. And they’re also gonna understand, you know, when you get into headcount reduction, or you get into, you know, different types of ROI-type modeling, you know, what you’re getting credit for around productivity from a financial standpoint or a pricing or a value standpoint is gonna be completely different with mid-level managers across multiple functions, all the way up to a CXO with higher-level initiatives. So, that’s something that maybe a first-time founder, that hasn’t had to go through that whole process — it’s not gonna be intuitive right off the bat.

Martin: Just to add to that. I think it’s just really seductive to think that like, “Oh, I’m an analytical person, and so I can do some basic research. And based on my research sitting in my, you know, offices in San Francisco, I know how to set the pricing, because I’ve got comparables and yada, yada, yada.”

Mark: I have my buddies, too, that’ll tell me.

Martin: That’s right. I could talk to my friends. They’ll tell me, like…

Mark: Not a bubble. I mean, it’s not a little bubble out here. I mean, everybody is gonna…

Martin: Everybody.

Mark: …like this, right?

Martin: And that’s actually kind of where I was initially, seriously. I am such a convert for a couple of reasons. One of them is, like, the value of a sales team is certainly to sell and bring in a number. But in my experience, when it comes to setting pricing and understanding what the market will bear, like, there’s nothing that can do it except for sales. This is just my experience. It’s not market research. It’s not marketing. It’s not the entrepreneur. I don’t believe, if you’re doing real category creation, you can just build an ROI tool. I don’t think it’s that simple. Depending on who you talk to in the organization, you’re gonna get, like, two different understandings of what the value actually is.

Mark: Some people aren’t gonna wanna be held to that ROI, right? They’re taking a risk. As you go up in the organization, they’re gonna want a bigger potential return for making this type of a bet, and the risk profiles are gonna be completely different as well.

Martin: Yeah. And I think something that entrepreneurs underestimate is large companies’ appetite to learn from startups. Very often in a hot area, say, AI or deep learning, you’ll have these, like, really smart entrepreneurs that have done their Ph.D, maybe they just peeled out of Google or something. Companies will pay to engage with them. They may pay 100k or 200k for a POC. And so, the entrepreneur is like, “I’ve got product-market fit. These guys are talking to me. The pricing is set.” But the reality is that it makes total sense for the company to do this and to learn from them. And so, I mean, one thing I really learned to appreciate about, you know, setting pricing aggressively early on, is you start to get real market feedback. You can’t delude yourself anymore. People aren’t buying it to learn about it, they aren’t doing this because you’re super charismatic. You get real signals. And the reality is early on in a company’s lifecycle, you really can only take on so many customers anyways. Setting pricing high really helps that.

Mark: Fast forward a little bit. Let’s assume we’re getting past POCs and our first 5, 10, 20 customers, but, you know, I see a lot of situations that we’re getting stuck in the mud with a lot of companies. And the reason they get stuck in the mud, particularly dealing with these bigger companies is — a lot of earlier stage companies, they haven’t thought through what the rest of that deal is gonna look like if it expanded throughout the entire environment. I see it time and time again and I constantly press. 

I say, “Look, in almost any proposal-type situation, you want it to have not only your initial pricing per unit set, but you need to go model and understand what would happen if that customer adopted throughout their entire enterprise.” Now, early on, it sounds like a pipe dream and, “Oh, my gosh, I can’t think that far ahead.” But you’ve got to go, literally — kind of, model these things up, because that’s what the customer is gonna be asking themselves. “All right. If I do this proof of concept for 100 nodes or 100 users, but I have 100,000 in my environment nodes or users or whatever the model is,” they’re gonna be thinking ahead, because they’ve been through this game before.

So, when you’re starting that quote process, to take a lot of friction out of the whole buying and selling process — to be able to, kind of, give them, you know, in the quotes, “Here’s what the small deal would look like to get started <inaudible>. Here’s what a medium-sized deal [would look like] with multi-BU. And then here’s what a large deal would look like.” I’ve technically validated it. I know I can scale. I know I’ve got the architecture or whatever. And I know I’ve got the security and things. But I also, I can actually back that up from a — you know, business case standpoint.

Martin: You’ve talking about pricing, which is great, but there’s organizational issues. Obviously, they’re incredibly important in selling.

Mark: There are different models. Right? I’m replacing something that’s been on-prem — now we’re offering SaaS. Two different accounting — I mean, we’re accounting for things different. In the old way, I capitalize it. In the new way, it’s coming out of operating. And so, that slows people down in a lot of cases, just from a financial chops standpoint, and dealing with the customer — and it changes the budget. The other thing that’s different is, I mean, that might have been centralized before with, you know, IT, but the whole world’s changing because it’s a line-of-business-type sell. People get stuck in the mud all the time,

Martin: I think it’s really worth underscoring the point Mark is making, which is, like, you work so hard in a startup, like, moving the ball an inch that often you don’t really consider, like, the macro success scenario. Because I would go in saying, “Oh, boy, wouldn’t it be great to get to the POC? Wouldn’t it be great to get to the initial sale?” But in reality, they’re trying to evaluate the full-on risk, which assumes that they like the technology and they’re gonna adopt it, which means they’re thinking all the way through it. And so, if you don’t walk in having thought through, “What would it mean in the full success scenario?” it’s much more difficult to have the conversation. And this is a trap, I think, entrepreneurs fall in all the time, not just in sales, but generally just trying to be incremental in the way that they build things.

Mark: That kind of goes into — it’s a competence and a confidence-type situation as well. And I might have technical competence around, you know, whatever my solution is, but I don’t have that customer knowledge competence. In a lot of cases, we’re trying to just throw a number out there, because we don’t have any data for that number. We don’t have any confidence in what it’s costing them, or what it would mean to them from a transformative standpoint. So we’re kind of guessing. And some of that might just be a lack of knowledge, and/or lack of willingness to go invest in, you know, the people and the processes that would, you know, be able to bring that to you. But you see it even with companies that do have sales and marketing organizations. And so, I’d be a little careful to say, “Hey, any sales and/or marketer is gonna be able to figure this thing out for me?” Because it does take somebody that’s a little more intuitive that it isn’t taking shortcuts, you know, somebody that wants to go understand and create the value.

Martin: I’ve actually found roughly two types of salespeople that roughly align with the maturity of the market. In a mature market, the customer is already educated. So, like, the type of salesperson that excels in that environment is very different than in the early days. I actually think that the actual competency of the salesperson depends on the size of the market. I mean, different salespeople are optimized for different types of markets.

Mark: I totally agree. One trap a lot of CEOs or founders might fall into here is, “I’m gonna go hire somebody that’s selling something similar to what I sell.” And the problem with that, like in your case — if I wanna go, you know, just hiring all network guys. Remember, I probably screamed at you and…

Martin: Oh, you did. I remember…

Mark: …the CEO back then…

Martin: I remember the conversation.

Mark: …who was — sometimes they are already native. And what you are doing and, in a lot of cases, what other technical founders are doing is so transformational that you’re not even gonna be able to sell that salesperson and/or those teams in a lot of cases, because they’re not even gonna believe it themselves.

Martin: Yeah. But not only that. I mean, this is something that you told me, which is, “Listen, you can’t take someone that’s selling into a mature market and put them in an immature market.” Here’s the big trap that — and I totally fell into this, and I think a lot of first-time founders fall into is — often the mature market salesperson is really compelling.

Mark: Yeah. And I probably might have gotten a little — well, I apologize. Five, six years later, I’ll apologize now.

Scott: Because you’ve changed so much between.

Mark: I’ve changed so much. Actually, I’ve gotten worse. The problem — you’ve really gotta go find somebody that really enjoys and has that intellectual curiosity, and that deep understanding of the customer and/or will tease that out. Somebody that gets excited about doing this exact thing — it’s typically not the one that’s been running in somebody’s playbook. It’s the one that really understands how to go create that playbook from scratch, and somebody that gets excited about this. With the founder and the technical vision, they can go map up to that, and translate that to, “Wow, that could be really transformational.” That person that can take you to those early potential targets and prospects, turn them into customers, learn, stop, put the recipe — you know, take the recipe, make it repeatable, turn that into a playbook, make that scalable. You can’t move it out of the lab, you know, into production unless you’re…

Martin: It’s infrastructure, of course. Yeah.

Mark: …you’re pretty sure, right? That’s a whole different profile than, you know, grabbing someone off the shelf that, you know, has been on a route-sell for their whole career.

Martin: So much of the early customer engagement should be, like, figuring out the product-market fit. So, you really want market feedback that you can use. And if you have somebody that, like, you know, every meeting is a good meeting and they’re using relationships, this and that, I don’t think you get real feedback from the market. But if you got, like, a good hunter that’s looking for the real opportunity and the large deal and will fight for it, I think the business gets that feedback. So, I think it’s so critical to hire the right type of salesperson early on.

Packaging products effectively

Scott: So, talk more about packaging. How do people think about new functionality? How do they think about upgrades to the existing functionality? How does all that play into it?

Mark: It depends on what kind of product we’re talking about. But is that customer early on gonna need everything that’s in the product the way it’s packaged? Or, in some cases — let’s say there’s three buckets of functionality, but in a lot of cases, depending on the user, maybe some buyers only need one. But as it expands, they’re gonna need two, and/or maybe eventually they’ll need all three. So, should we chunk that up to make it easier for them to get started, even though they don’t need the other functionality, or they don’t recognize that need yet? So, there’s — maybe we should have three prices for an a la carte type version, and then maybe we have a package if they get it all up front. But then the other thing is, let’s take the product roadmap and let’s look ahead. Let’s take the feedback we’re getting from the market, understanding what’s coming down the pike, or what we’re thinking about building.

And I think this needs to be a closed-loop process from the field, with product development. As you mature, you have product management, product marketing. All these groups need to be working together and, kind of, challenging each other as, “Is this something that should be added? Is this something the customer is gonna get value out of that we should be charging for?” Because we can put an ROI to it, and we can make it easier to buy early on, and they can grow into it later on with the packaging and/or different options. The other thing is that, you know, one solution in one vertical and/or use case, from a pricing standpoint or a value standpoint can be completely different in another. So, you’ve gotta kinda balance those types of things from a value-based standpoint. You’ve gotta have the knobs just ready to crank. I think a lot of companies wait too long and/or they drop the ball on it.

Martin: So, you’re doing your company, you’re creating your product, you know, you’ve set your pricing pretty aggressively, you’ve established a price in the market, you’re very happy, but you’ve probably necessarily priced yourself out of some constituencies, right? So, now you wanna go ahead and expand your footprint. You maybe want to go to, like, other areas or other verticals. And you wanna do this in a way where you maybe have tiered pricing, but the risk of tiered pricing is cannibalization. If you don’t know you’re gonna do this beforehand, you may not have the flexibility to actually pull out independent bits of value. And so, now you’ve got — on one hand, either you cannibalize yourself with, like, an over-feature-rich product for the lower price, or you don’t get sufficient market expansion because you’ve priced yourself out of it. And so, like, I think that — absolutely right. Early on, you should make sure that you’re thinking through how you can bifurcate this when you do your market expansion.

Mark: Yeah. Another good example on that is, you know, those early customers sometimes if you’ve really — if you’ve done a good job, you’ve got those early wins, and you gave them a sweetheart, lighthouse-type deal, you’ve really gotta pay attention to not just the pricing but how you’ve gone and contracted. Because later on, I’ve seen it happen over and over again, companies just kicking themselves in the tail because they really gave up the farm.

Scott: Because some of this, you’re right, is a product packaging issue. Have you literally sold everything that you will ever build in perpetuity to this customer upfront?

Mark: Right. Let’s go get those lighthouse customers, but let’s put a fence around the “the enterprise” thing and — a lot of cases, it’s just the startup is marching up to this big bad wolf, and come on in and, you know, I’ll huff and puff and blow your house down. And guess what? Yeah. These big companies know how to negotiate and contract, and they know every trick in the book. And we’ve seen some of these teams walk in with their junior varsity uniforms on, and then there’s a lot of injuries that happen. So, you really gotta be careful. So, how do you solve that? I mean, get some help and make sure you get the right advisors helping you with these early-type deals.

Martin: Yeah. I think it’s important people understand how deceptive this all is, because many large companies have outreach programs to startups. So, if you engage with a finance company, like a large bank, they’ll be like, “Listen, we work with startups all the time.” They have whole groups that will take these things and POC them. That, in my experience, never actually makes its way over to the procurement office. So, while they’re great at, you know, engaging with you, lightweight process, POC’ing, finding value, you meet the technical teams — you even have a deployment scoped out, and a professional procurement person in the room. And if you don’t know what you’re doing — I mean, I’ve been in multiple situations, we were very close to basically getting site licenses to, like, 100,000-person organizations for almost no pricing. And this is when, again, Mark comes in and kind of sets us straight. But it’s very, very important to be sure you know for every one of these large deals what you’re doing.

Mm: It’s pretty out there in the forest, but you will get eaten if you’re there after dark.

Correcting mistakes

Scott: This is all good advice. How do you fix problems or fix mistakes? What if we’ve gone out and we have priced too low, or what if we’ve, you know, discovered that we essentially haven’t ringfenced people and we’ve given them stuff? How do you approach that problem? How do you think about it? Are there things, at least, you know, in retrospect, companies can do to kind of right the ship in those scenarios?

Mark: Yeah. I mean, it’s really — I mean, it’s case-specific. A lot of it is in the language of the agreement that you’ve done. I mean, if you’re doing term licensing or subscription-type deals, you know, that’s different than a perpetual. That’s another piece on this contractual thing. The buyers are saying, “We’re gonna do it on our paper” and that type of thing. The startup doesn’t have any paper. They don’t even — I get these questions all the time. It’s like, “We’re trying to figure out how to deal with these guys.” And they want us to mark up their SLAs and their MLAs and, you know, master license agreements and stuff like that. And we don’t really know what to do. And so the reason you don’t know what to do is you haven’t put the work in to kind of figure out what yours is gonna be. So, the ideal situation, as quick as you can, is you gotta think this stuff through, and you need to have your language. And look, early on, particularly even later on when you’re a big company, you’re gonna swap paper and everybody’s gonna mark it up. But you should know what you need in, from a language standpoint, to protect yourself. As far as redoing the deal, it’s just customer or it’s case-specific.

One thing I’d like to add, though, along those lines, though. On the product development side, if you think about this early enough — and it doesn’t have to be right at the onset, but as you grow the product and functionality out — from a product standpoint or there’s things that can kind of turn on and off to help me modularize. You wanna think about how I can do that, particularly — and you get the bottoms-up models. There’s a lot of nuance in that, as well, and there’s always the resistance and/or in the DevOps-type environment on the open-source, as you’re moving up the stack to the premium and/or the enterprise-type pricing and functionality and support, that if you’re not constantly understanding where you’re at with the customers and/or in the competition a lot of cases, you’re probably put yourself at risk.

So, having, on the product side — be able to turn things on and off or be able to measure and monitor and be able to not only establish but measure that value over time. I see a lot of companies, even big mature companies — they’ve done all the front work. They go get the deal. Everybody agreed on, you know, what the value was gonna be. One of the most valuable things is to go back and validate that, and put that into a case study. And the best place to sell something is where you’ve already sold something. Right? So, they’re constantly working with the customer and with the engineering and product development, to kind of bring the best value to the customer — and then measure it and make sure everybody is on the same page. And then, you know, that helps the next 10, 20, 100 customers come on board too from a sureness standpoint.

Martin: Scott, to your question about, like, setting pricing in the market. I mean, the bad news about setting pricing in a market is, it’s really hard and it takes a long time to actually set the price in the market. The good news is, it takes a long time. So, if you enter at too low of a price point, like, it takes a long time for these things to solidify. And so, I do think that there are options to kind of raise price, you know, add differentiation based on functionality — especially early on in the product cycle. So, it’s not, kind of, like a misstep that you wanna do, but if you do it, I do think there’s a lot of time to correct it.

Mark: I agree. I totally agree. There’s gonna be a lot of trial and error in that process.

Pricing reviews and sales strategies

Scott: Yeah. I wanna touch on two other things quickly first. One is just — so we’ve been talking about, kind of, initial pricing and packaging and stuff like that. How do you think about, kind of, the ongoing process? So, how often should CEOs and the VP of sales and VP of product management be thinking about, “When do I revisit pricing?” Is it on a release basis? Is it new competitors enter the market? What are the things that people should be thinking about that give them some guidance as to, you know, how often or how frequently you think about these types of things?

Mark: I think it’s almost constant. Again, a lot of it depends on the stage. But if you’re not reviewing, you know, a win-loss type report on a monthly, at least a quarterly, you know — some kind of fairly frequent cadence depending on what’s going on in your business and in the market — and doing the post-mortems on both sides of that — I think you’re putting yourself at risk. And mapping it up with, “Where am I at with the product?” or, “Where am I at with my customers?” That whole roadmap discussion isn’t just an internal discussion. That’s something that you’re constantly doing with your early customers that may be partway down the journey to fully deploying your solution. 

You’re able to check that from a pricing and functionality standpoint. Then, what [does] the competitive dynamic and landscape look like? I might be getting, you know, pressure from below me — maybe it’s from the open-source world, or the homegrown, maybe it’s from lower-end competitors that are just doing small medium, and you’re starting top-down in the market, you know, with the bigger companies. It may be from incumbents that are reacting and starting to, you know, understand that you’ve done some damage to them, and they might be adjusting what they’re doing. So, a constant review, I think, of that is — and, really, a 360-type situation — is something that should be pretty frequent.

Scott: Another question I had was, you guys have also hinted this a couple of times, but the relation between the type of sales organization you can support and what kind of pricing you have out in the market. And I don’t know if there’s any heuristics we can give folks, but, you know, in order to be able to support, for example, a direct enterprise selling effort — which requires, you know, a lot of, you know, high touch — there are probably certain ways you have to think about pricing and what your average selling price is, and what you can get an account, versus obviously something that might be done through an inside sales organization. So, is that something that, kind of, CEOs and, you know, VPs of sales and VPs of products need to think about from the very beginning? Which is, how does pricing relate to the actual mechanism by which I’m gonna go to market? And how should people think about that?

Mark: Yeah. I think it’s pricing, it’s packaging, it’s my product-type strategy. If I’m gonna be a bottoms-up, I mean, it’s gonna start with, you know, some marketing and, you know, a freemium to premium type model, and I’m gonna work on that adoption, and then convert them from freemium, more to the premium. And later on, it’ll be enterprise, then that’s — because my transaction and my ASPs are gonna be extremely small, then that’s gonna drive your go-to-market. A lot of it is product-related. If it’s something that requires you to come in high and move left and right, and be agile in organization, as well as up and down — and because of the complexity is — that typically means there’s also gonna be a big outlay, either kept capital end or operating, and that’s probably gonna require, you know, an outside direct salesforce. The pricing better — and the size of the deals are gonna have to map up to support that to go fund that.

So, you can’t have a mismatch. In some cases, you’re gonna kind of have [it] all, right? You’re gonna have, you know, the ability to sell across all. And you might have to put in the layers of your go-to-market, you know, from a sales and a marketing standpoint to build that bottom-up. You know, maybe a mid-market-type or commercial-type teams and more of the enterprise major account type named account situations, and different types of marketing to drive different types of penetration across the board. So, it really hinges around the product. And it also can hinge around what your product development strategy is, and what a competitive landscape might look like. I mean, I can think of tons of examples.

Martin: Yeah. Often in technical organizations, sales basically dominates, like, costs because it’s a variable cost. We need more sales to bring in more money type thing, where R&D is often more fixed. And the cost of, like, an ISR is gonna be much less than a direct sales force, but only certain types of products or markets are amenable to an inside sales model, right? If it’s a very mature market and the customer is educated, it’s probably more amenable than if it’s something totally new. Also, if it doesn’t require a lot of integration, or isn’t super technical, or the product is made very simple to use, it’s more amenable to an ISR. An ISR is an inside sales rep, which is basically someone on the phone which will call, rather than, you know, has a briefcase…

Mark: Lower cost.

Martin: …hops in an airplane. It’s lower cost.

Mark: Inside. Yeah.

Martin: What we’ve seen in our portfolio companies — a number of them — is actually they’ll experiment with both. So, they’ll start with ISR and a couple of fields and they’ll actually play with the model to understand what works best, especially if they’re moving towards more non-traditional buyers for IT. And so, we’ve seen this in a number of companies. And they’re able to determine over time which model is the most cost-effective.

Mark: Yeah. I also see, in a lot of cases, you know, it starts one way and they wait too long to add the other layers in. I mean, it’s very common on the inside-type bottoms-up motion that, particularly as they build out the management, that sometimes they’ll miss moving up market because the VP might not have that skill set, you know, to go build out that next level. So, that’s something I think the first-time CEOs need to understand and be questioning themselves, and get help to question. “Am I slowing the whole company down because I’ve, you know, under hired?” Again, it depends on the stage of the company, but you should look at and calibrate, and look at both, you know, “What does an under versus an over-provision look like? And how much headroom am I gonna have?” because you could damage yourself by waiting too long.

Scott: So, we’re talking a lot about, kind of, first-time CEOs here. And let’s just assume the product is not taking for some reason, it’s not selling. Yeah, we’re missing our plan. As a CEO, how do you know whether you have a product problem, whether you have a problem with your sales team, whether you have a pricing problem, a packaging problem? How do you tease these things apart in a way that actually helps you think about ways to address issues like that?

Martin: The one bit of insider advice that I would give to first-time CEOs or entrepreneurs is, it’s really hard to find product-market fit. And it’s a saga that can last for years and, you know, you’re gonna doubt yourself, and you’re gonna doubt the product, and you’re going to doubt the market, and all sorts of different feedback. And especially if you’re doing, like, serious category creation, I mean, it takes a long time for markets to mature. Markets mature at their own pace. So, I don’t have a simple answer for what to look for, but I do know that you have to be patient and you have to be persistent and, you know, it takes a while. And I also know, having seen it, once you hit the inflection, it’s really obvious. Once you hit product-market fit, you start to get more engagements than the organization can handle, and you can’t scale enough.

Scott: Mark, Martin, thank you for the time. It’s time for us to wrap. So, just to kind of encapsulate what we talked about, you know — price early, price often, this is definitely an integral process. Get your sales guys in and the right sales guys, right, whether they’re the kind of hunter or gatherer folks in early to help you with this process. And then, you know, kind of, the important point that, you know, Mark always makes, which is, you gotta think about packaging, right? Don’t give away, basically, you know, the entire collection of stuff that you will build for the next 20 years to your first customer. So, think about how to segment it both in terms of users as well as features and functionality. So, lots of things for people to chew on here.

Martin: And be aggressive with pricing.

Scott: Be aggressive with pricing. That’s right. Thank you.

Mark: Thank you.

Martin: Thank you.

• 

  Mark Cranney is the COO of Skydio. Previously he served as COO at SignalFx, and prior to that was an operating partner at a16z where he oversaw the market development team.

  Follow Twitter
• 

  Martin Casado is a general partner at a16z where he invests in enterprise companies. Prior, he was cofounder and CTO of Nicira (acquired by VMware) and is the creator of the software defined networking movement.

  Follow Twitter
• 

  Scott Kupor is an Investing Partner at Andreessen Horowitz where he is also responsible for all operational aspects of running the firm.

  Follow Twitter

This podcast is all about emoji. But it’s really about how innovation really comes about — through the tension between standards vs. proprietary moves; the politics of time and place; and the economics of creativity, from making to funding … Beginning with a project on Kickstarter to crowd-translate Moby Dick entirely into emoji to getting dumplings into emoji form and ending with the Library of Congress and an “emoji-con”. So joining us for this conversation are former VP of Data at Kickstarter Fred Benenson (and the ???? behind ‘Emoji Dick’) and former New York Times reporter and current Unicode emoji subcommittee member Jennifer 8. Lee (one of the ???? behind the dumpling emoji). 

So yes, this podcast is all about emoji. But it’s also about where emoji fits in the taxonomy of social communication — from emoticons to stickers — and why this matters, from making emotions machine-readable to being able to add “limbic” visual expression to our world of text. If emoji is a (very limited) language, what tradeoffs do we make for fewer degrees of freedom and greater ambiguity? How exactly does one then translate emoji (let alone translate something into emoji)? How do emoji work, both technically underneath the hood and in the (committee meeting) room where it happens? And finally, what happens as emoji becomes a means of personalized expression?

This a16z Podcast is all about emoji. We only wish it could be in emoji!

Show Notes

• How emoji originated, how they’re standardized, and more [0:48]
• The difference between emoji and emoticons [11:03]
• Social and political considerations around which emoji to include [15:30]
• Using stickers and images to express emotion [21:40], as well as Bitmoji [24:20]
• The story behind creating “Emoji Dick,” a translation of “Moby Dick” using emoji [29:11]

Transcript

Sonal: Hi, everyone. Welcome to the “a16z Podcast.” I’m Sonal. Today’s episode is all about emoji. But it’s also about bigger questions and how innovations come about, from the tension between open standards and proprietary systems, to the economics of creativity. We begin with a tour of different emoji and how they came about, the politics of emoji, where emoji fit in the taxonomy of visual communication, and why this matters. And finally, we talk about the difficulties of translating emoji when it’s not really meant to be a language. Joining us for this conversation are Fred Benenson, an early employee at Kickstarter who built their data team. He’s also infamous for kickstarting a project to translate “Moby Dick” entirely into emoji. Also joining us is Jenny Lee, former New York Times reporter, who is a member of the Unicode Subcommittee on emoji and who recently led the effort to get the dumpling emoji, which is where we start the conversation.

Emoji basics

Jenny: I wasn’t a really big emoji user. In fact, the first time I ever heard of emoji was when Fred started his Kickstarter called “Emoji Dick.” And I was like, “What the fuck are emoji?”

Sonal: What is “Emoji Dick?”

Jenny: This was before they showed up on our iPhone with, like, perky little yellow faces. I was like, what? It’s, like — sounds [like] something very bizarre.

Sonal: I just started. I didn’t even actually — just to be blunt, I had a very hard time using emoji, because I didn’t quite understand how to even, frankly, use them. I don’t understand when people send it to me, if it’s not the obvious heart, you know, etc. But as I’ve been using it more, I’ve found myself, sort of, expressing myself now in, kind of, quirky ways. And I don’t know if people really get it or not, but I’m getting a kick out of it.

Jenny: But that’s the fun of the ambiguity.

Fred: I have a friend who showed an exchange between a friend of his who was dating a guy, and he would only send her emoji. And she was like, “I just can’t — I can’t handle this.” And he showed me the screenshots of their exchange and it was hilarious.

Sonal: You’re helping translate.

Fred: But yeah, and so, like, I was like, “Oh this is…”

Sonal: You’re like the Cyrano de Bergerac of, like, emoji.

Fred: Yeah. I was like, “This is what this means.”

Jenny: I can definitely see it being, like, sort of an irreconcilable difference between people in relationships.

Fred: Significant other.

Jenny: Fast forward many, many years, emoji have shown up on our iPhone. And I’m texting with my friend, Yiying Lou, who’s best known as the designer of the Twitter “fail whale.” So, we’re texting back and forth about, like, dumplings. And so I sent her a picture of the dumplings I’m making. And then she texts me back knife and fork, knife and fork, yum, yum, yum, yum, yum. And she goes, “Wait, Apple doesn’t have a dumpling emoji.” I was like, “How could that be?” I was, like, because there’s so many obscure Japanese food emojis, since emoji are from Japan. Like, you have, you know, everything ranging from ramen to curry rice, to tempura, to, like, you know, the rice thingies on a stick to even — there’s even, like  — triangle rice bottle, looks like it had a bikini wax.

Fred: There’s also the fish cake, which is the white one with the purple swirl.

Jenny: Yeah, yeah, the spiral. Totally, right?

Sonal: Oh my God.

Jenny: And I was, like, how could there not be dumplings, right? Because it’s such a universal food, right? Because there’s, like, pierogies in Poland, and momos, and gyoza, and empanadas. Like it’s just, like, a food from around the world.

Sonal: I mean, technically, samosa is a dumpling.

Jenny: Yeah, samosa, ravioli. And I was like, okay, emoji are universal, and then dumplings are universal. How could there not be a dumpling emoji? And just — in my mind, I was just, like, clearly, whatever system in place has failed.

Sonal: How do you solve a problem like the dumpling emoji?

Jenny: Yeah, and I found out that emoji are regulated by the Unicode Consortium, which is a nonprofit organization based in Mountain View, California. It now has 12 full voting members that pay $18,000 a year just to vote on issues, including, like, emoji and other kinds of, like, technical…

Sonal: Are those members in Mountain View or from around the world?

Jenny: So of those 12, 9 are U.S. multinational tech companies — Oracle, IBM, Google, Yahoo, Adobe, Facebook, Microsoft, and Symantec. Then of the other three full voting members, one is a German software company, SAP. Another is the Chinese telecom company Huawei. And the last is the government of Oman.

Sonal: That’s a really interesting crew.

Jenny: Isn’t it an interesting crew? And they have these quarterly meetings, and then I just show up. And they’re, you know, very welcoming. You know, they’re like, you know, “Thank you for coming. What brings you here? Tell us about yourself.” It felt like showing up at church — like a new church. You’re a new member. They all knew each other very well. They’re very excited that there’s, like, someone, you know, young and, like, diverse, who’s just, like, randomly showing up. And so I in that process learn how you get emoji passed, and how they’re regulated. And so, in January of 2016, we submitted a full proposal for dumplings, take-out box, chopsticks, and fortune cookies and got those all passed. So, those will be in Unicode 10, which means that — that’s announced in June of 2017. And so, they’ll actually hit your phones several months after that. I was like, wow, billions of keyboards will be impacted by this and…

Sonal: That’s amazing. Were there other proposals submitted at the time?

Fred: Oh, there are constantly proposals. There’s this whole process that people like Jenny — some of them make it through.

Jenny: It’s complicated, yeah. No, if…

Sonal: It’s a lot of work. It does introduce some good, useful bars actually for making sure quality gets through at some point.

Fred: Yeah. And to their credit, the Unicode Consortium has an amazing list of emoji criteria, where they say, “Okay, here’s what we’re looking for for emoji. It’s gotta have like, you know, kind of a unique meaning, in that it’s not covered by other stuff, but it also should have, like, you know, some ambiguity. So, it’s not just, like, literally one thing. It could be used in other contexts.”

Jenny: Also, there’s one of the more interesting rules, which is no celebrities, deities, or logos.

Fred: Whoa. The Easter Island head is kind of a violation of that one, but that’s got its own story. A couple of years ago, with a big update, the Easter Island head showed up in, like, the back of the travel section of emoji. And I was, like, what is that doing there? Who is traveling to Easter Island so often that they need to use the Easter Island emoji? And it kinda just stuck in my mind. And then I started using it in this, kind of, like, slightly culturally insensitive way to, like, reference some supernatural phenomenon that I didn’t understand, right? Like, if I was in a conversation with somebody and I was just, like, completely flummoxed, I’d just, like, send that one.

Sonal: Yeah, it’s like your version of Bermuda Triangle or something.

Fred: Yeah, yeah, I was just like, “Who knows? Stoneface.” Other people use it for, like, stoned, right? Like, there’s lots of combinations in there. The reason why it’s in there is that there’s a statue in downtown Tokyo. I think it’s a Shibuya station that is called Moai, which is a name of just, like — it’s a proper noun of that statue, which was made by an artist that was, like, a reference to [an] original Easter Island head. So, it turns out, Japanese teenagers use this waypoint to meet each other. And so, that’s how it ended up in Japanese cell phones, and that’s why it ended up in emoji. The artist used this inspiration of Easter Island. The interesting twist is that when you look at it on the iPhone, it doesn’t look anything like the statue in Tokyo. At some point, Apple was like, “We’re not gonna make it, like, this Tokyo one. We’re gonna do it [like] the original one.” Android, on the other hand, their Moai emoji looks like the Tokyo station one.

Sonal: So fascinating. I read a study — I actually included in our newsletter months ago of someone comparing how emojis look on different platforms and how it actually changes meaning, because…

Jenny: Totally.

Sonal: …you can actually think you’re sending one thing and you get something else.

Fred: That’s gonna happen in any system that has standardization. Like, you’re gonna try really hard to make sure people hue to the specification. But, you know, people do their own implementations and things change. In fact, the whole reason why emoji are in Unicode was because you would send your friend an emoji, and then their cell phone would actually just render the incorrect one. It could be so much worse. And the fact that there is a standard means that, like, you only get these, like, weird edge cases.

Jenny: There are still some interesting vestiges of, like, the different telcos between Apple and Google. One was Docomo and the other one was SOFTEL.

Fred: SOFTEL.

Jenny: SOFTEL. So, they’re basically — depending on who their partner was locally, they kind of inherited those generations of emojis. For example, on Apple, “women with bunny ears” is, like, two women dancing in kind of, like, a “let’s party” kind of way with their bunny ears. Whereas on Android, it’s just the headshot of a woman with bunny ears.

Fred: And it’s referencing this slightly misogynist part of Japanese culture of bunny woman, which is itself a reference to the Playboy bunny.

Jenny: Oh, right.

Fred: And so, they were cocktail waitresses working in nightclubs. That made its way into the Japanese set. And so when it came over to America, like, I think Apple must have been like, “Let’s make this a little more fun.”

Jenny: One of the easiest things actually to get emoji passed is showing that a vendor uses it. Another argument is for completion. This is actually why chopsticks got passed fairly easily, because we had, like, knife and fork, so you need…

Sonal: Oh, so you need completion of a set.

Jenny: …completion. So that…

Sonal: So, it’s actually you can tell a whole story, like, stringing together a bunch of…

Jenny: No, I just think that it’s, like, they’re engineers…

Sonal: Right. You can’t have ABCD and skip the D.

Jenny: Yeah, yeah. Actually, one of the weird issues is that there are red, yellow, green, purple, blue hearts…

Fred: Hearts. Yeah, yeah.

Jenny: …but not orange. So one of the big lobbying efforts has been to fill in the orange.

Sonal: So the case of the Apple bunny ears and the Japanese bunny women — that was a case where there was an intentional translation to, sort of, obscure the cultural reference.

Jenny: It’s more that…

Fred: There are just two separate ones, right.

Jenny: …they’re often — try to map technically the same emoji, but it’s, like, rendered and sort of interpreted differently. They like emoji that can have multiple meanings. You can also just have, like, emoji that have one meaning. But it really has to be a really good one if it’s gonna be one meaning. So for us, the Chinese take-out box, for example — one of the arguments that we made is that, one, it’s an iconic shape. It also symbolizes both an entire cuisine, which is Chinese food, and also a means of eating, which is delivery…

Fred: Takeout, right.

Jenny: …and takeout. Right. And so in that one symbol, you get a lot of, sort of, secondary meaning. And with fortune cookies, like, it’s technically a cookie, but it also means, like, mysterious, and the future, and the unknown, and like…

Sonal: So, like, sort of primary, secondary meaning. One of the criteria for an emoji to get passed is that it has to have a certain element of ambiguity to it.

Jenny: Well, I think, yeah…

Fred: I love this. I’ve been thinking about this so much. When I did “Emoji Dick,” it was more of an experiment around crowdsourcing an emoji itself. Like, I wasn’t, like, so much interested in making a formal case that emoji could be a language because it was still so early.

Jenny: Yeah, it was very early.

Fred: Could it get there maybe one day? Yeah. But Unicode makes a really good point. They’re like, “Emoji is not a language. It shouldn’t be a language. The value is that it’s ambiguous.” And I’ve really come around to that thinking, and this idea that the charm of sending an emoji is that it can be interpreted in a couple of different ways. And that’s actually why we value it. And I’ll go further and say that — a lot of people ask me why emoji have become so popular. And I think it’s tied to the fact that we now are just inundated with text. We live in a text culture, right? We communicate via text. Our careers are run over email. We read constantly. Everything we do is mediated through almost literal words. And so, emoji represents this kind of reaction to that. And the popularity of emoji, I think, is largely due to the fact that we need some other way of expressing ourselves over text.

Sonal: If the pipes are so mechanical, like, phones and machine, you no longer have the non-verbal aspects.

Fred: Absolutely.

Sonal: So, this is actually replacing sort of this human element of the glimmer in your eye or, like, the blush on your cheek.

Fred: Or even just…

Sonal: There’s an emoji that does that.

Fred: …you think about the amount of signal you get from somebody’s voice on an analog telephone. And when you strip that out and all you’re communicating is, like, LOL, you don’t actually know how sincere that laugh is, or that chuckle, or whatever that person’s trying to convey. And so emoji gives us a much bigger palette to convey this kind of, like, extra, like, limbic meaning that we wanna have in our communications, but we can’t because we’re texting all the time.

Sonal: So, to break down the taxonomy of figural representation not using literal text. Let’s talk about where emoji fits. We have emoticons, which are, like, a colon and a parenthesis, and that gives you a smiley face. Or, like, a semicolon and a parenthesis and that gives you a wink.

Jenny: Right. Using punctuation for existing…

Sonal: Using punctuation is an emoticon.

Jenny: Is often ASCII-ish.

Sonal: Right, because it’s got ASCII art as well.

Fred: And it goes way back. Some of the earliest references to emoticons go back to the 19th century as well, where people…

Jenny: Oh my God.

Fred: Yeah, yeah. People were using colons, and dashes, and parentheses to express, like, a wink. It goes way back. It’s important to add in hieroglyphs and iconography. Other humans have had this idea before. Like, the medium and the technology is kind of, like, incidental.

Sonal: I’m so glad you brought that up, because it’s so important to not get caught up in technology time. Well, technically, technology includes, like, sticks and stones, so that does go back in time. But in the context of this machine web that we live in, then we have emoticons as part of the taxonomy, and then we have emoji. But how would you guys define emoji?

Jenny: It’s Japanese. Drawing language.

Sonal: Emoji.

Jenny: I don’t know how to pronounce [it] in Japanese, but the Chinese — the “emo” is not for emoticon, or emotion or anything. It’s just totally a coincidence.

Sonal: Wow.

Fred: It’s hard not to just hue to the Unicode Standard and say it’s the set of icons defined in Unicode that represent objects, and nouns, and actions and…

Jenny: The way that I explain it to people is, an emoji is a character — an emoji is something you can put in the subject line of an email because it literally is text. So, in the same way that Unicode has, kind of, defined the standard to unify all the graphical representation of different languages throughout the world — and even non-languages, or like, you know, the Wingdings and all of that kind of stuff. Emoji actually slipped into that entire system. So, there is literally what they would call a codepoint assigned to each emoji — or, sorry, not every single one, because now they’re, like, compound emoji. But there are codepoints assigned to emoji, which basically says, “You know, when a computer sees this codepoint, they render it in a certain way.”

Fred: But it’s important to, kind of, wrap your head around what’s actually happening inside the computer, because the emoji is being sent as text. If your computer supports UTF-8, UTF-16 — that’s just like a standard way for your computer to handle text, whether it’s your phone or your laptop — then it’s being told, “Render this emoji.” But it’s actually up to your computer’s operating system, whether it’s OSX or iOS or Android or whatever, to go fish out a little image and put it on your screen. And so that image is actually controlled by the hardware manufacturer or the software manufacturer. You know, when it’s actually rendered on your screen, the operating system is choosing which image to show you. And those images are actually stored, you know, in the same way that other images are stored on your computer as little PNG files. And so, Apple, you know, puts those on your computer, and your computer chooses to render those, which is why you may get slightly different, you know…

Sonal: Different interpretation. Right. I’m glad you walked us a little bit — yeah.

Jenny: And it’s actually really interesting, because recently Facebook just introduced their own emoji and that, like, basically hijack Apple emoji. So, you can turn that on or off, but essentially, they’ll swap out all the ones on the Apple.

Fred: And Twitter has had their own set for a while and so…

Sonal: Why is that? Why do these manufacturers care? Yeah.

Fred: So, there are interesting copyright considerations here. My guess is a lot of those companies are doing it because A, they can afford to make their own set. B, they wanna avoid the legal liability of using Apple’s set.

Sonal: Apple, right.

Fred: And C, like, they think they might kind of have some, like, moment of, like, “Hey, did you see Twitter’s new emoji?” Right? And so, you know, these large companies are kind of…

Sonal: Innovating on emoji.

Fred: Yeah, yeah. Like, re-innovating and re-illustrating their emoji. And I think, you know — I think Microsoft actually just evolved to a new set, or was it Android? I think it might’ve been Google or Android. They just upgraded to make it seem a little bit more normal. Like, they had gone from, like…

Jenny: <crosstalk> the terrible blue and white…

Fred: Yeah, yeah. So…

Jenny: Or, there was, like, the blobby ones that were terrible.

Fred: Yeah, I think Google had blobby ones for a while. And now they’re doing somewhat normal ones.

Jenny: Scariest emoji ever — the Microsoft emoji are, like, blue and gray, and they look like monsters that hide underneath your bed.

Sonal: Why are they blue and gray? Why do they look like that?

Fred: I think it’s just an attempt to be, like, different from, like, the yellow skin tone.

Jenny: Well, also, you have to — part of the original emoji is, you wanted things that were skin tone neutral. So Apple and Google chose yellow, but Microsoft for some reason chose gray.

Sonal: Oh, gray because I was gonna say, for Hindu, like, blue is — actually not a bad thing to have your skin blue. It’s, like, a God.

Jenny: The other thing is, if you have your own set of emoji, you can actually start adding to that set without going through Unicode.

Sonal: Through the Unicode Consortium, right.

Social and political sensitivities

Jenny: So, like, a very good example is the gay family emoji, originally, where it’s not actually one emoji. Another one is, like, man, man, kid, kid. That is actually a compound emoji of four characters glued together using something called a “zero-width joiner,” which is basically like an invisible glue. So, if you are sending that emoji to someone else who doesn’t have the ability to render it out, it actually unravels itself into, like, a multiple character. Now, what you’re seeing is a lot of vendors making compound emojis. And, like, actually one of the places where this is being debated for use is the need for a professional female emoji, right? Because one of the big problems right now, on the existing set of women as represented by emoji is, like, there are only, like, really four roles for women to play, compared to men. You know, men, you can be a sleuth or you can be, you know, a policeman. You can be, sort of, a medical worker…

Fred: Construction worker.

Jenny: There are all kinds of things. You can even be Santa Claus. But as a woman, the four things you can be as a role are, basically, bride, princess, dancer, Playboy bunny.

Sonal: Oh my God.

Jenny: That’s it.

Sonal: It just goes to show you how the — I mean, of course, this is the politics of human life [playing] out in these systems. I mean, the perfect example I was thinking of is the rifle emoji, and the case of, I believe, Apple, Google, and Facebook. Charlie Warzel at Buzzfeed wrote a really detailed article investigating this, and about how they, sort of, helped suppress — as part of the Unicode Consortium — the rifle emoji.

Fred: Right. Emoji already has a gun in it, right? And it’s like, okay, so how many more versions of that do we need? And you’re right, it’s absolutely a political topic. I mean, that issue manifests itself in so many other places than emoji. The country flag stuff is super interesting, because that uses kind of what Jenny’s talking about with these compound emojis. Unicode didn’t actually wanna decide which flags were and weren’t an emoji. So what they did…

Sonal: Right. You’re legitimizing, then, political issues.

Fred: What they did was they built this kind of, like, meta country system, so that you would actually be pairing these country letter emojis together. So CNN would go together, and then it would be up to your phone to decide if you showed the Chinese flag. They pushed that decision-making — that, like, political decision-making of which flags to support — off to the handset manufacturer so…

Jenny: Microsoft actually does something weird there.

Sonal: What do they do?

Jenny: They don’t show a flag. They show a flag plus the two letters.

Fred: The two letters.

Jenny: Microsoft doesn’t render it, like, normally.

Fred: To the point about politics being kind of embedded in emoji, it’s not just because these are icons that represent the parts of our lives that we feel passionate about. It’s because there’s a finite palette. It’s not like language, where you can only — you know, you can kind of combine, say, whatever you want.

Sonal: It’s combinatorial. You can take multiple combinations and turn it into whatever you want.

Fred: Yeah. Language is, like, you get way more degrees of freedom to kind of express yourself. There’s a finite number of food items that are available to go in there. And when you think about the vast, like, multitudes of humanity, whether it’s, you know, people’s relationship status, their sexual orientation, or skin color, it’s like — emoji is never gonna be able to express that. And so, like, how do you contain this thing that’s, like, growing and kind of has to grow as more and more people use it, but also, by definition, has to be a finite list of icons?

Sonal: Well, how do they handle the skin tone issue? Because one of the things that I noticed is that on Apple — because I use an Android so I didn’t notice this — you can press down on a thumbs-up, for example, and then you can pick among 15 different shades to, like, pick a skin code shade that’s closest to you.

Jenny: Five and yellow.

Sonal: Oh, five.

Jenny: Yeah, it’s based on the…

Fred: Do you remember the name Fitzpatrick skin tone scale?

Jenny: Yeah, it’s actually used — it’s the same skin tone system that dermatologists use to categorize.

Sonal: This reminds a little bit of being a kid, when you had [a] Crayola box. I remember that the only shade you had — there was, like, a nude shade, or, like, a skin tone.

Fred: Yeah, and nude was always Caucasian.

Sonal: And I’d use sepia. I remember using sepia to represent my skin color.

Fred: I mean, there’s a great history about this in — this is gonna sound weird for me to say. But, like, women’s pantyhose, like, had this issue where nude was always considered Caucasian, and people were, like, “This is ridiculous.” It was one of the earliest blind spots of emoji I remember. It was like…

Sonal: Right. Well, I mean, if you have, like, only white men designing them. Do you remember when Slack — there was this guy who wrote a post about just the brown hand?

Jenny: Yeah, yeah.

Fred: Yeah.

Sonal: And I remember it was so meaningful, because it’s such a minor seemingly arbitrary thing but then it is true. Like, the first time I saw that I could find my skin color in a system, and to be able to use it, was kind of amazing and empowering. And I think there’s something significant about that.

Fred: I would totally agree. I don’t share your experience as the person on the other side. And so, it’s funny for me because I don’t…

Jenny: He’s a white male, for those of you who cannot see Fred.

Fred: Yeah, so for those of you listening, I’m a white guy. I don’t share that, like, sense of identification with the bright, white, like, <crosstalk> index.

Sonal: Right. You’re like, “That’s not necessarily me.” It’s just, like, a thing.

Fred: Yeah. And I’m like, it feels odd to opt into that, which speaks to my privilege as a white male where I just like…

Sonal: No, it’s not just that. If you’re not exposed to it, you’re not exposed to it. The bottom line is if you’re any person of color, you’re always aware of your color, especially if you’re in a context where everyone else is not the same color as you.

Fred: And so, when I texted my friends who are not white, and I’m like, should I be choosing that one? And I just choose the yellow skin tone.

Jenny: Yellow.

Fred: And that’s just like the — I feel way more comfortable with that.

Sonal: Yeah, yeah.

Jenny: So, my solution is, I often send four. Like, it’ll be, like, yellow, light, dark, and then, like, the beige one.

Sonal: Oh, that’s great.

Jenny: So it’s like — it’s like a Benetton ad in emoji world.

Sonal: Benetton emoji, that’s fabulous.

Fred: So now the, kind of, evolution is that we have yellow for, like, all the human face characters, and then you can choose skin tones for some of them. But it doesn’t get at, like, more nuanced issues about, like, cultural and racial identity having to do with facial structure or hairstyle.

Sonal: Oh, right, the features.

Fred: And these are…

Sonal: That’s a great point, actually, because one of the pet peeves I have is when I used to go to foreign countries and look at billboards, it always glorified that aquiline nose, the face structure — whereas there’s a totally different type of face structure in different areas.

Fred: Emoji probably won’t ever have that amount of, like, customization, and Unicode gets this. And they actually say, like, “We’re adding, like, 60 emoji a year. This is unsustainable. We feel like the future is inline images.” And that, kind of, breaks my heart as, like, kind of a, you know, nerd standardization guy, like, who really appreciates all the hard work that went into Unicode, and the idea that it is a standard. Because if you’re just sending inline images forever, then, like, you know, you have no idea what’s gonna be on the other side and if they can render the image.

Jenny: So stickers. So Kimoji, for example, Kim Kardashian’s “emoji…”

Sonal: It’s awesome.

Jenny: They’re not actually emoji. These are just stickers. They are images that you can text back and forth. But, you know, again, you know, standards — can you put it in the subject line in the email? And those you can’t.

Sonal: You can’t so therefore, they don’t qualify. So just to go back to the…

Jenny: They’re not technically emoji.

Sonal: Right. So then, going back to our hierarchy, we went from emoticon to emoji and now stickers you would define as a…

Jenny: Stickers. Stickers are basically inline images. I mean, stickers are just images that you can pick from a palette.

Fred: And I think you can — you know, in certain apps, you can, like, apply a sticker to an image that it, like, sits on top of it. But you’re then in this, kind of, like, proprietary ecosystem of — that’s okay. But, like, you think about the stuff that really works, and the stuff that really changes the future of the web and communication, and it’s all standardized.

Sonal: It’s all standard — and you’re saying this as a standardization person. Because my friend, Connie, who wrote a wonderful post on the topic of stickers, argues that emoji are very limited for what you need to do, because she feels that you have so much more expression and the ability to convey so much more with stickers than you do with emoji.

Jenny: Emoji doesn’t preclude the use of stickers. There are some sets of images that are universal enough that should be hard-wired into the operating systems, and basically can be cross-platformed that an iOS device can talk to — you know, Microsoft Windows can talk to, like, an Android device, can talk to your Mac laptop. Like, the fact that — at least you’re not gonna get little square boxes as long as your operating systems are fairly up-to-date.

Sonal: Well, that goes, then, to your point about why standardization is important, because you’re now giving up that you’re in this proprietary ecosystem like WeChat or Line, and you only have their sticker set. And you can’t always transfer all these stickers across…

Fred: And also, if you think about the accessibility issues around stickers, right? Like, people using screen readers— they’re not gonna be able to interpret an image. And, like, emoji actually have names. And so, in theory, there’s much better accessibility for emoji for somebody who’s visually impaired, so.

Jenny: Yeah. Like, for example, last year, Oxford English Dictionary chose “face with tears of joy.”

Fred: “Face with tears of joy,” yeah.

Jenny: Which I always thought looked very sad.

Fred: Yeah, it’s…

Jenny: You know, the thing with the eyes and it’s, like, bawling. But that’s actually “face [with] tears of joy.” And you know that because, you know, all these emoji have…

Sonal: They say the label — Oxford put that in there.

Fred: That was the word of the year.

Jenny: So, the word of the year was an emoji.

Fred: Part of the reason they chose that was that it ended up as number one on my friend’s site, called emojitracker.com.

Sonal: Oh, right. That’s right. The emoji tracker, which tracks all the use of emoji on Twitter.

Fred: And for a while, it was just, like — it was, like, the heart emoji or something, or just the smiling face emoji. So, I think it’s really interesting when the top emoji shuffle, because, you know, whenever you start texting with somebody who hasn’t used emoji before, they’re, like, choosing, like, the safest ones.

Bitmoji and expressing emotion in text

Sonal: Going back to this idea of some of the companies owning their own emoji, and some of the proprietary open tension between standardization, freedom of expression — what do you make of this notion that part of what we’re doing here is essentially also creating a more machine-readable web, in terms of emotional reading? Because, essentially, you’re now adding a whole new layer where you can codify people’s emotion, sentiment — in ways beyond just a black and white, like, don’t like.

Fred: I’ve been thinking about this so much, actually, and not in the context of emoji, but actually Facebook reactions.

Sonal: Yeah, me too. I used to assign and edit op-eds on this topic because I was very obsessed with it.

Fred: I think it’s a really interesting topic because if you look at traditional sentiment analysis in the data world, it’s kind of a joke. You have to have training data, you have to know good cases. And the…

Sonal: And just to interject for a moment, as someone who’s tested a million of those systems and can never find one that actually works for my needs, they are so binary. You don’t get anything useful, and you’re not getting insight.

Fred: One of the reasons there is that words have these degrees of freedom. They can be used sarcastically, and you would never know it based on the semantics. And so, traditional sentiment analysis is really broken, because you’re using these, kind of, like, stale, rigid semantic definitions. What’s really interesting about Facebook reactions is, you know, you think you’re saying, “I love this thing,” or, “I’m sad about this,” or, “I’m angry about this.” But what you’re actually doing, in conjunction with that, is giving Facebook really great labeled data for sentiment analysis.

Sonal: That’s right. Machine-readable data. That is a holy grail of emotional sentiment understanding. When I was at WIRED, I assigned a piece to a sociologist, Evan Selinger, because I wanted to coin this phrase — the mood graph — because we have an interest graph, social graph, you know, all kinds of other graphs that link all these nodes and ideas. And now, to have, like, a mood graph, to essentially be able to put your pulse on someone’s mood — something very finite, yet constantly changing. It’s just a fascinating thing to be able to codify this.

Jenny: The sentiment stuff generally correlates very strongly with [the] human face and body. So I think this is also why people agitate so much for emoji that look like themselves. Like the redheads, and people with beards, and people, you know, who are bald.

Sonal: Or anyone who has curly hair. People with curly hair relate to other people with curly hair.

Jenny: And so, I think people really love seeing themselves represented in emoji, which is why Bitmoji, which is highly, highly, highly customized stickers in sort of emoji spirit.,,

Sonal: Oh, my cousins and I Bitmoji on WhatsApp all the time. I think there’s something really symbolically important about Bitmoji, because you are putting yourself in it and conveying in this sticker form. The fact that Snapchat bought it I think is really telling.

Jenny: Oh, yeah, for $100 million. Is that right?

Sonal: Right. Especially given that they are changing this culture of how you express yourself through your facial expressions, with face swapping and filters. Connie and I made the argument that it’s sort of like a new — like selfies. It’s selfies as a form of stickers. So what we’re talking about, with the machine-readable, is a little [more] distinct than this, but it’s sort of an interesting idea all the same.

Fred: I also think it ties into this slightly dubious notion of the uncanny valley where if you wanna try to represent yourself and you wanna have, like, configurability around that, it needs to be, kind of, cartoonish for it to be believable. I think what we’re seeing with Snapchat filters — and I don’t know if you guys have played with SNOW yet. That’s like a…

Sonal: No, I haven’t.

Fred: It’s, like, take Snapchat filters and just multiply them by a thousand. It’s just, like, amazing amounts of diversity around the amount of stuff you can put on your face. It is this weird convergence on identity and emoji that’s kind of happening.

Sonal: I agree and, in fact — this is gonna sound, like, a little out of left field for a moment — but the whole notion around the Chewbacca mask lady, when — you know, that was the most popular Facebook live video ever. It got, like, unprecedented views, and it was simply a woman who was trying on her Chewbacca mask in the car. And she’s laughing and giggling about it. And then she puts her mask on and then she takes it off, and she laughs so uninhibitedly, it’s insane. And I make the argument that what was so empowering — because it took off for obvious reasons — is not the fact that she was laughing so uninhibitedly. It’s the fact that it took putting on and then taking off the mask for her to do that, which is not unlike what happens with communication through these filters, and being able to now express yourself through these cartoon-like ways in a real way.

Fred: I mean, honestly, it takes me back to, like, theater, and, like, Shakespeare in, like, seventh and eighth grade. I remember having these, like, really intense discussions about, like, what it is to put on a mask and what a mask represents about yourself.

Sonal: It’s a very Campbellian idea, right? The Joseph Campbell, like, mask, and the myth, and the man. You’re right. There’s a theater — I mean, that’s why people say improv is so interesting for any career field, but I think that there is an interesting moment now coming together with selfie stickers, emoji, Bitmojis altogether, where we do have this new emotional web coming together.

Fred: Right. And using emoji — the first time I thought about this — could be kind of, like, putting on a mask over your, you know, self to…

Sonal: Words.

Fred: Yeah, over your words to convey to yourself this, like, this extra this kind of additional layer, this emphasis of your emotion that you otherwise might not get.

Creating “Emoji Dick”

Sonal: Okay. So, going back to you writing an entire book in emoji, and yet you were saying that you’ve kind of evolved in your thinking, you know, that emoji is not necessarily a language but clearly, it is a visual language. And it is a tool for communication. It’s not complete. So, how did you translate that? I mean, what were some of the trade-offs and decisions you made? And, by the way, for the audience — that book was, like, 2009 or that was, like, many years ago.

Fred: Okay. Okay. I’ll…

Sonal: So what emojis base were you working off? Did you make them up? Like, what’d you do?

Fred: So, I’d gotten a text from my college roommate whose wife is Japanese. He sent me an emoji, and I was like, “What is that?” They told me you could download, like, basically a Japanese app and it would, like, awaken your iPhone to the emoji keyboard, like…

Jenny: Come alive, emoji.

Fred: It just spoke to me in the, like, you have to hack the iPhone to get the special keyboard of, like, Japanese icons. And I was like, “Oh my God, I want this so bad.” I was like, “This is amazing. I should write a book in emoji.” And I was like, “Oh, that’s a lot of work. I don’t know if I can write a whole book in emoji.” And then I was like, “Or maybe I can translate a book in emoji.” I was like, “Okay, what books would work?” And I was like, “Well, it has to be in the public domain,” because I worked a lot in, like, the copyright reform space. Nobody is gonna just, like, let me translate their book into emoji without a lot of effort. For a moment I thought about the Bible. And I was, like, that’s too obvious. What’s, like, totally, even more inappropriate?

Sonal: So “Moby Dick” came to mind.

Fred: Yeah, it came to mind as, like, this, like, impossible book to put into these symbolic characters. As soon as I thought I was like, “No, I can’t do that. That’s crazy.” And I was like, “That’s, like, too hard.”

Sonal: Honestly, it’s a little bit like — I just came back from seeing “Hamilton.” And so, it’s a little bit like the idea of putting a rap to, like, the founding fathers. That’s what I find so fascinating.

Fred: Yeah, I would say “Hamilton” was probably…

Sonal: It’s like a mashup of mediums, and time, and culture.

Fred: And it’s like one of those things where you tell it to somebody and they’re like, “You can’t do that. That’s crazy.” And then you’re like, “Well, the fact that you just said that made me wanna do it.” And so…

Jenny: And not only that. There are not one but two whale emoji. Were there at that time?

Fred: No, there was only the original…

Jenny: The cute one?

Fred: The cute one. The, kind of, <inaudible> style one.

Jenny: Cute one, aww. So, there was a whale emoji.

Sonal: What’s his name? Ahab is battling the cute whale. Aww.

Fred: Yeah.

Jenny: That’s the second one.

Fred: Yeah, I think it’s called sperm whale — didn’t come up until later. So, I was like, okay, wow. That would be really interesting to do all of “Moby Dick,” because it’s also, like, really long. I mean, it’s 10,000 sentences. And okay, well, if I don’t wanna do this, maybe I can hire somebody to do this. And I was, like, experimenting with Mechanical Turk at the same time. I think it was, like, one of the original Amazon Web Services. It was, like, it would later become, you know, part of that AWS umbrella.

Sonal: Yeah, I remember people using it for research and stuff.

Fred: Right. It’s still used for research. It’s still invaluable for that. But, you know, a couple of other people had done, like, an experiment here or there, like, using it, like, off-label. I had made a task at Mechanical Turk — just to ask Turk workers, “If you could ask anyone, like, to do anything on Mechanical Turk, what would you have them do?” And they came up with this long list of stuff. And I don’t think “translate a book into emoji” was one of them. But there was some creativity out there. I was, like, okay, I’m gonna try this thing where I’m gonna hire people to translate “Moby Dick” into emoji, some portion of it, and see if this works. So, I did the first chapter and the results came back, and they were hilarious. They were so good.

Sonal: They were good.

Fred: Yeah, they were great.

Sonal: How did you assess that? First of all, what do you mean you did the first chapter? Like, did they break it down word by word? How do you capture that in emoji? Is it like a sentence?

Fred: So, I decided I was gonna do it on a per sentence basis. And that actually turned out to be one of the challenging parts of the project was, like — splicing sentences is actually like, kind of, like, a classically hard and a natural language processing problem. And so I kinda, like, figured out a hack to, like, chop it up. And I wrote a lot of regular expressions to basically get the whole book into sentences.

Sonal: Wow. But you decided basically the sentence was the unit of analysis, not a phrase, not a word, a sentence.

Fred: You would have a sentence in the task and you’d say, “Pick any of these emoji.” And then I actually wrote my own little emoji picker, because these things didn’t exist at the time. I had gotten the emoji from a friend. He had reverse engineered the iPhone SDK and basically hacked out the PNG files from the software kit to basically have the raw emoji in image form. And so, I took that and just made, like, a little JavaScript, like, HTML thing and, you know, dumped that into Mechanical Turk. And it came back and I was like, hey, this works. And so, I think the sentence that’s kind of, like, on the cover of the book if you go to the website, it’s, like…

Sonal: The website being Emoji Dick.

Fred: emojidick.com. “Call me Ishmael” is the first sentence of “Moby Dick.” And the emoji that the Turk worker chose was, like, telephone, man with face, sailboat, whale emoji. It’s perfect.

Sonal: That’s amazing.

Fred: That was just like — but the rest of it was just, like, indecipherable emoji nonsense. And some of the people were just, like, all right, “Give me my five cents. I’m gonna click some random emoji.” And other people just, like, clicked every single emoji. So, the plan became — have people translate the same sentence multiple times. So, you get three different emoji translations for one sentence. And then have another set of tasks where people vote on the best, most appropriate translation. So, like, of the three, which one got the meaning across the best? And I was, like, oh, I was just, like, getting really excited about this. And I started doing the math on how much it was gonna cost. And it was, like, oh, it’s gonna be thousands and thousands of dollars. That summer, I met the Kickstarter guys. I started talking with Andy Baio. He was like, “You should put it on Kickstarter.” So that night I went home and put it on Kickstarter, launched it the next day, and ended up working for them and…

Sonal: And, by the way, how much money did the campaign make?

Fred: My goal was, like, $3,500. I ended up raising $3,700. So I worked on it for, you know, nights and weekends for another, like, eight or nine months. And then, you know, self-published it on lulu.com. You can still buy it. It gets printed on demand. And, you know…

Jenny: Do people still buy it?

Fred: I’ve sold, like, thousands of dollars of “Emoji Dick.” And I’d say hundreds of copies. And probably, like, 500 or 600 copies of it have sold since then, which is not a lot.

Jenny: I bet this podcast is gonna sell a bunch.

Fred: Yeah, well…

Jenny: You better share some of the proceeds with me.

Fred: Okay, so there are two copies. There’s the black and white copy, which is, like, the easy to print one, and that’s, like, $20 or $30. And then there’s the full color one, which, like, is obviously preferable, because emoji are so colorful. But when you’re printing on demand, 800 pages of color laser hardbound copy — it’s actually really expensive. So, that thing costs, like, $180.

Sonal: Right, because you’re not printing in bulk.

Fred: Exactly.

Sonal: Because you actually save money when you print in bulk, right.

Fred: So I have to sell that one for that much.

Sonal: Damn.

Fred: And, like, people still buy it. In 2013, The Library of Congress contacted me and, you know, they said, “We would like to acquire ‘Emoji Dick’ as our first emoji book.” I was like, “Are you sure?” They’re like, “Yeah, yeah, we’re sure.” And I was telling a friend — and David Gallagher, I think you must know from the Times. And he’s like, you know, everyone submits their stuff to the Library of Congress. It’s not that big of a deal. And I was like, “No, man, they asked for it, like, they’re acquiring it.”

Sonal: I think it’s a big deal because it was a curatorial point of view.

Fred: Totally.

Sonal: They’re saying, “This is a cultural moment. It’s not just a book that was published, and we need to figure out how to acquire it.”

Fred: I was like, “All right. I’ll spare a copy.” I signed it. I sent it to them. And then they sent me this little, like, you know, certificate in digital form. It was hilarious — and this is my favorite part — is that it’s somehow listed as a translation of “Moby Dick.” So when you look up “Emoji Dick,” it says all these libraries have it, because it’s really just saying that, like, they have a translation. They have the original “Moby Dick.” Now it’s got a life of its own, and people still discover it, and yeah.

Sonal: That’s amazing. I mean, you actually even curated an art show, didn’t you, based on this?

Fred: Yeah. Friends of mine put together a kind of emoji survey art show, and there [was] some really great stuff in there. Emojitracker was there. There was a programming language built out of emoji. There was a lot of other good stuff.

Jenny: I mean, emojis can have their URL. I mean, that’s another thing. They’re literally text, so you can have, like, emoji@…well, I don’t know, @gmail. But you can have emoji in your email address.

Sonal: Oh, you can?

Jenny: Totally.

Fred: You can also buy emoji domains.

Sonal: So you have an emoji book. You have emoji art shows…

Jenny: Emoji hackathons.

Sonal: Emoji hackathons.

Jenny: So, our big news this week is that in November, in San Francisco, we are going to throw the first-ever Emojicon, which is basically…

Sonal: What? Is that like Comic-Con?

Jenny: It’s like Comi-Con, but emoji, of emoji.

Fred: I really hope people show up dressed in emoji costumes.

Sonal: I was about to say, I’m gonna show up as — you guys are gonna — Yiying is gonna show up as dumpling emoji for sure.

Jenny: Or, like, poop emoji or, like, the ghost emoji. So it has many different elements to it. So one is definitely, sort of, this whole “emoji learn” aspect, where it’s, like, panels and talks. And there’s, sort of, emoji film festival, and there’s an emoji hackathon, and then there’s an emoji art show. And then, of course, the opening party emoji where, you know, our goal is to only have food that is also emoji.

Sonal: So, why a conference? I mean, of course, I see the cultural significance, but to bring people together around this idea of a first-ever Emojicon, like, what’s the significance of that?

Jenny: Part of it was, I thought it already existed. And to me, the fact it didn’t…

Sonal: I kind of did too, to be honest. When you just said that, I was like, what?

Jenny: Yeah. And then I was like, the fact it didn’t exist — and I kind of have this issue where of, like, if I think something needs to be — I will try to make it exist.

Sonal: You will make it exist, God damn it.

Jenny: Right. So, we did it with dumpling emoji. We did it with Emojicon. And so we actually have some really cool sponsors. We’re gonna have a lot of, kind of, emoji activists, kind of, out there.

Sonal: Emoji activists, I love that.

Jenny: And also, from our perspective — you know, there are a lot of policy decisions around emoji and, obviously, the world really cares about emoji. Whether or not it’s the rifle emoji, or the condom emoji, or, like, professional women emoji. Part of the goal of Emojicon is to open up that discussion, so it does not just held at the Unicode level so to…

Sonal: So, are Unicode members gonna be attending this conference?

Jenny: Oh, members of Unicode Emoji Subcommittee including, like, you know, the co-chairs. And we’ve timed it in November between the Unicode Conference itself and the Unicode Technical Committee meeting. And also, like, it’s right around election day.

Sonal: Well, you guys, thank you for joining the “a16z Podcast.”

Fred: Thanks for having us. This was so much fun.

Sonal: This was so much fun. We could keep going…

Jenny: So much fun, hours and hours on emoji non-stop.

Sonal: Yeah, I wish we could.

• 

  Fred Benenson
• 

  Jennifer 8. Lee
• 

  Sonal Chokshi is the editor in chief as well as podcast network showrunner. Prior to joining a16z 2014 to build the editorial operation, Sonal was a senior editor at WIRED, and before that in content at Xerox PARC.

  Follow Twitter