The Exchanges

Q Is there a challenge function internally? Like how do you decide in, in terms of making the actual decision?

A Yeah. So there's, there's a, there are, if you kind of pull top end venture firms on this, you'll get actually every possible answer. Um, and so there, there are some firms that run on a unanimous vote model, um, where they, they literally need unanimity. Um, you know, there are other firms that have a, um, you know, they have a majority thing, um, or they need a certain number of people to get critical mass. There are some firms that run a top-down so-called investment committee where you, you know, young, the younger partners in the firm get to propose ideas, and then there's sort of a council of elders. You know, that either, you know, does the, does the Gladiator, Joaquin Phoenix thing, you know, thumbs up or thumbs down. We have optimized very strongly towards what we consider to be kind of the most aggressive and most, you know, anybody say contrarian model. So our model is what we call single trigger puller. There's no vote. There's no investment committee. There's no override. There's no, like, I don't, you know, I formally sign off on all, you know, Ben and I formally sign off on all the investments, but like, I don't, I don't like evaluate the investment proposal and then basically do thumbs up, thumbs down. Like that's not, like I don't do that at all. Um, instead what we do basically is we, we basically, we basically delegate budgets. Like, so the, the governance me…

AI assessment note: “So our model is what we call single trigger puller. There's no vote.”

Q How do you keep that bar rising, right, and not let success sort of turn into complacency?

A You need, you need some, you need some forcing function, right? You need a reason. This is why we always, and the smart, the smart entrepreneurs all know this, but this is why even entrepreneurs starting company number four or five, six, You know, this is why they should have, like, a board of directors, and they should have people on their board who are actually, like, very smart and strong-willed who will argue with them, right, as opposed to just them doing whatever it is they're going to do on their own. It's because, like, they, you need, no matter how good you, you need some force, most people need some forcing function where they're, they're sort of forced to fully think things through. They're forced to do all the work, and they're forced to be able to defend it, right? And if, at your point, like, if they don't do that, it, it, it, it turns into complacency and laziness, and at some point it turns into arrogance, and it, it just, it becomes very destructive, and then you stop being effective. And so there's going to be some forcing function. The, the, the, the Peter Thiel on the shoulder thing is, is a forcing function is how I use it. Um, it's like a forcing function on my own thought process. It's like, okay, if I were explaining this to Peter or to Elon or whatever, like, would they be like, oh, that's a good idea. Or they'd be like, Mark, like, Elon would say, that…

AI assessment note: “You need, you need some, you need some forcing function, right? You need a reason.”

Q What have you learned over the years about making exceptional decisions that most people miss?

A So, uh, quite honestly, I think, I think a lot of it is embracing the, uh, the, the probabilistic nature of the domain. You know, a lot of these questions, like venture teaches you the very direct versions of these lessons, because it, you know, what is it, the old Submariner slogan is, uh, stupid will be punished. When, when you pass on enough companies that turn out to be these giant, you know, kind of global, you know, kind of behemoths, you know, at some point you're just like, okay, like, obviously I'm not I'm not a super, like I have to have some level of epistemic humility because I'm not making all these decisions properly. And again, the nature of the domain is I'm not going to make all the decisions properly. Like what, what each of the top venture capital capitalists of all time has in common is they, they all, each of them missed almost all of the great deals of their generation. The bad news is you're going to like make all these mistakes. The good news is you don't have to score a hundred percent and not having to score a hundred percent basically means you can take chances. Right. And, and, and specifically the way that you instrumentalize that is it's basically, and this, we talk about this a lot inside the firm is success or failure is not at the unit of an individual investment. Right. Success or failure is at the level of a portfolio. Right. And so, um, and, …

AI assessment note: “I think a lot of it is embracing the, uh, the, the probabilistic nature”

Q We sort of talked about the, the probabilistic thinking in terms of the magnitude and test taking as it relates to the education system. To me, I mean, from the outside, looking in the education system seems a bit dysfunctional. Uh, you're a parent. What do you think is wrong with it? And how do you go about navigating it? And what can we do to fix it?

A So it's completely unfixable. So, so it's, it's, it's, it's like none of the specific point things people are worried about are fixable because the whole thing is built in a certain way. Uh, so it, it, it goes back to this, this, it goes back to this historical thing we started with, which is, it's, it's a, it's a mass education system, right? So it's a mass education system that was created in a time and place. Um, and it was created in a time and place, and you know, you probably know the history of this. It was created in time and place, you know, roughly a hundred years ago, a 110 years ago. And there were a set of, you know, theoreticians and youth names like, you know, John Dewey. Um, and they, you know, developed a, you know, a system. And, and at the time, by the way, at the time it was like a huge advance, right? Because it was like, we were going from like a nation of farmers Right, into this, into, you know, this sort of second industrial revolution and into the 20th century, you know, most kids didn't have much education in the 19th century and, you know, um, you know, group of farms or whatever, and, like, they, they knew, the theoreticians, the experts knew, like, we need a system to educate all the kids, um, and then, and then they built a system the way that you build systems, you know, the way that you built systems in 1910 or 1920. Um, which is you built a mas…

AI assessment note: “So it's completely unfixable. So, so it's, it's, it's, it's like none”

Q It's also interesting when you look from the outside in that there's sort of like no individual to blame, right? Like I feel for the teachers in the classroom. I feel for the principals. I feel for everybody involved in this system, but the end outcome is just terrible.

A Yeah, Charlie Munger has this great thing I really like, which he says, look, if, if you often in the world, and this is a great example, you often have a system, and the system has incentives, and the incentives result in terrible behavior. And then you have a choice, which is, do you blame the people who are exhibiting the terrible behavior, or do you blame the people who created the incentives? Right. And, and, and he, he believes, and I agree on this. I think a hundred percent you blame the people who create the incentives. Right. I think, I think you're right. Like, I think when you take human beings and you put them in a system where they have bad incentives and you get bad results, like at some level it's not their fault. Like, you know, you could say maybe they should resign a principle or something like that, but like most people need to live their lives and they need to like, you know, put food on the table for their families and so forth. And so like, I just don't have expectations of like mass revolt, um, uh, I guess people's self-interest. Um, and so, yeah, no, I, I totally agree with that. And so therefore I think you have to look at incentives. Um, and then you have to look at the people who are putting the incentives in place. Um, and you, and you have to judge them. And, and, and, and that's the, that's the thing with this thing is like the people who put these…

AI assessment note: “I think a hundred percent you blame the people who create the incentives.”

Q How do you assess the judgment of founders? If you're not assessing the idea, how do you assess the judgment?

A Yeah, so I'd say there's two categories of, two categories of founders, two ways to do this. So there's sort of, you know, there's sort of, there's the ones who've done it before, right? Um, and so the ones, you know, the ones where this is their second or third or fourth time starting a company, um, you know, with them, you have this just obviously enormous advantage, which is you can basically look at the work that they've done. You know, it's like anybody else who's done, you know, who now has a track record. You can look at the work that they've done. And again, on this, it's not that you're necessarily looking for the track record of like the level of Success, per se, right? Although, you know, as Napoleon, you'd rather have a lucky general than a smart general. So like, you know, somebody who's had a highly successful company once, right, has, you know, people, has an enormous benefit of the doubt when they walk in the door for obvious reasons. But, you know, but even people who haven't had a big success before, it's like, okay, you can look at kind of how they did the work, right? And so you can look at, like, you know, how, you know, how did they prosecute the opportunity, you know, the previous time, you know, and you can talk to the people, right, who they work with. It's like, okay, You know, how were they under pressure? How were they on, in terms of, like, inventiv…

AI assessment note: “you can look at kind of how they did the work”

Q Can you walk me through in detail how you make those bets? Like what is the big capital allocation decision-making process that you use at a 16 Z?

A So there's basically, there's three tiers. The way we operate is there's basically three tiers of that, um, sort of small, medium, large, and the small one is so-called C funding. Uh, the medium one is so-called venture funding, and then the large ones are growth funding. Um, you know, the, the seed bets, the small ones are, you know, two, you know, any, I don't know, anywhere two to five million dollars on average, um, which, you know, is a lot of money, absolutely speaking, but it's, it's small in the context of, you know, the, you know, it's small in the context of the global economy. It's small in the context of funding your businesses. It's small in the context of our funds. The, the venture bets are much more serious. The venture bets are 10 plus, uh, million. And then for the venture bets, they're more serious for us because that's when we get really involved in the companies. Like, that's when they form a board of directors. And we have, you know, somebody who goes to the board and we tend to really like work hard at the company. And so that, that's like a, you know, that's signing up from our standpoint to like a 10 year journey. Um, with the company, including, you know, an obligation to put more money in later rounds and so forth. So that, that's a, that's a more serious bet. That's sort of, sort of size 10, ten million and up. Um, and then the, the growth bets are f…

AI assessment note: “The way we operate is there's basically three tiers of that, sort of small, medium, large”

Q And what's sort of the most pessimistic scenario you can come up with, with how technology plays out? Do the incumbents start using it to squash free speech? Like, how do, how do you think about this?

A Yeah, I think you could paint, yeah, you could paint maybe two, two pessimistic scenarios. So pessimistic scenario number one would be the one you just alluded to. And it basically is like, okay, um, this technology had the potential, uh, to be, to be revolutionary and to kind of disrupt all these hierarchies and institutions. Um, but in fact, it turned out that it was, you know, fundamentally enabling technology for top-down control. There's 1984, which basically is the pessimistic view, which is like technology is used by the state, you know, kind of the full, the full kind of, you know, I don't know what the Hegelian like, you know, world state dominant, like totalitarian regime, Soviet Union kind of thing. Um, you know, technology is used to basically control the populace in sort of this very overt kind of top down way. Um, there's this other really wonderful book, um, called, uh, Orwell's Revenge written by this guy, Peter Huber. Um, and In Orwell's Revenge, he actually does a rewrite of 1984, um, and he does a rewrite of 1984 through the optimistic lens, you know, sort of your optimistic scenario, and, and the optimistic lens, the, the twist that he puts in the plot of 1984 is he basically just makes the telescreens, you know, the sort of technology in 1984, these telescreens interactive, sort of interactive TV thing, um, he basically, he makes the telescreens two-way, yo…

AI assessment note: “pessimistic scenario number one... fundamentally enabling technology for top-down control”

Q a long time, and this conversation does not disappoint. We explore why he studies history to see the future, his optimistic and pessimistic view of the future, how A-sixteen Z makes decisions, and specifically what Mark knows about decision-making that most people miss, our largely dysfunctional education system, assessing the judgment of founders, and so much more. It's time to listen and learn. What are you obsessed with right now?

A So there's a lot going on. So the, the full answer to that question would probably take, take the entire podcast, but, you know, I would say probably two big buckets of things, you know, my, my day job is, you know, being obsessed by all of the sort of changing, you know, technologies. Um, and then, you know, the, the, the sort of, uh, impact that all, all the, all the new technologies are having on, on the, um, you know, on, on, on different industries and on the world broadly. Um, and then the other thing I've just, you know, really, you've gotten just sort of extremely curious about, I would describe, uh, you know, like a lot of people, um, a lot of my assumptions about how, you know, the world worked and about how the U S worked and about how politics worked and about how society worked or kind of upended, you know, kind of sometime between 2008 and, you know, 2015. Um, you know, and then for sure heading into the last five years. Um, and, and, and so I've, I've become somewhat obsessed with kind of reading my way back and basically trying to figure out essentially what the hell is going on.

AI assessment note: “I would say probably two big buckets of things”

Q Of course. Like we have a biological tendency sort of to defend our territory, right? So it would be, it would be hard, especially with the organized unions and all of this other stuff that's involved with the education system to reform it.

A Well, it's like the tip off, the tip off. If they just, you know, step back and think about it for a second. Like you could be pro union or anti union. You have whatever people, you have whatever views in unions they want. Organized labor, you know, big debate for a long time. You have whatever views you want. But The idea of a unions, the idea of unions of public sector employees. Right? Like, so you, you now have people, you have a double layer of civil service protections and union protections. Right? And you're going to reform that system? Like, no, you're not. Like, they're going to take your money, laugh at you, and like, nothing is going to happen. And so, yeah, there, there, there is no, on this, there is no fix. There, there is only one thing to do. In my opinion, there's only one thing to do. Right? Which is to build a new system. Right? And the new system has to get built from scratch. It has to get built the hard way. Right. Um, it will probably not look right or sound or smell like, you know, these, it probably, you know, the new system will probably not be a giant parrot. It'll probably be something else.

AI assessment note: “And you're going to reform that system? Like, no, you're not.”

Q Can you walk me through how you sort of see some of these technological changes and how they'll impact sort of this social hierarchy or how people get into positions of power?

A So, so the really big one is, um, uh, is, uh, you know, basically hierarchy versus network. Um, right. Or you kind of say institutions versus network. Um, and so, and by the way, for people interested in this, um, uh, Neil Ferguson actually wrote, he wrote a really interesting book called The Square and the Tower that goes through this in some detail recently that's very, you know, kind of historically informed, another great source on this. Um, I should say I agree a lot with like the first two thirds of the book, and then I kind of take the polar opposite view of him on the last third of the book. Um, so it's funny. The last third of the book made me very mad. Um, and of course, the reason it made me mad is because he mounts sort of a rising defense of the old power structure, or sorry, rousing defense of the old power structure. Right. Whereas of course I'm, I'm on the side of the new power structure. So of course that would make me mad. It's kind of the reverse of that, right? I'm mad by the resistance. And so anyway, the, the big battle is sort of, you could call it kind of institutions versus, versus networks. You could call it kind of top down hierarchy versus sort of lateral, you know, kind of network building. Um, and so you start by saying top-down institutions. So top-down institution building is kind of how the West, you know, and, you know, it's a certain, I'm sure…

AI assessment note: “the really big one is, um, uh, is, uh, you know, basically hierarchy versus network.”

Q Is that inherent to even the winners today? Like Google is now an incumbent, right?

A So, so this is one of the funny things I talk a lot to our founders about. Um, and this is not a, you know, what's, what's so useful I think of this view is this is not a criticism of any individual person, right? Or any criticism of any individual company. This, this is why I try to get to the broader trends, right? Cause like the, the broader trends like are impersonal in the sense of, okay, if you, if you see this pattern over and over again, then it's not just you. Like when I, you know, when I tell somebody you're doing X, it's not like a criticism of you're doing X. It's more an observation of, you know, you're doing X just like everybody else who's been in your situation. Has done X, and you know, at the very least, be aware of it. If not, you know, decide to make a conscious choice not to be like that, you know, if you don't like that. Um, but one of the things I talk to our founders about all the time is, it's actually funny, a successful company has about five years, a successful startup has about five years until they become a new incumbent. Um, and, and they actually start to behave like an incumbent. That's rational. Like, of course, that's rational. Like, they've now built something worth defending, right? And they've, like, they've now established power in the world, right? They've established a position for themselves. Um, you know, they've built something of va…

AI assessment note: “a successful startup has about five years until they become a new incumbent”

Q Can you talk to me a little bit about those patterns? What, what comes to mind when you think of an example that you're, you're surprised by?

A I mean, you take almost anything, but I mean, I'll just give you the most basic ones. The most basic one is, is societal reaction to new technology. Right. Um, and so there are, you know, there are these new technologies of our era and, you know, we could go, you know, make, make, you know, whatever list we want. Um, you know, and it, you know, everything from, you know, social media and internet to, you know, many other things, automation, robots, uh, and so forth. And then there's, you know, there's sort of the societal reaction. There's sort of the, you know, the reaction in the press, there's the reaction from politicians. Um, you know, there's a reaction from different political movements, you know, political candidates right are now, you know, taking, you know, kind of major stands on these things. Um, you know, both in the U.S. and around the world. You hear it a lot, like, around the kit, like, what technology does to kids, right? It's sort of the go-to thing, which is, you know, that, you know, sort of this technology is going to ruin the new generation of kids, and then you just kind of, you know, ask yourself, well, is this the, is this the first generation of technology for which they have made those claims, right? Or, in fact, have those claims been made for, basically, prior generations of technology? And then, of course, the next question is like, okay, are, are,…

AI assessment note: “The most basic one is, is societal reaction to new technology.”

Q Can you give me both sides of the next 10 years? I'm interested in hearing your most optimistic scenario and your most pessimistic scenario.

A Look, optimism is just like, it's basically the Morrison book, right? It's basically you have, you have all these kids. So you have kind of the world as it is. You have the status quo. You have the power structures. You have the hierarchies and whatever. They work as well as they do and so forth. And then you basically have all these kids. Um, and, and this is kind of one of the things in my day job that's just always so interesting is like, cause these kids like walk in the door and they're just there and they walk in the door from all over the world, right? From every walk of life, from every possible background. Um, and they walk in with, you know, an idea and a skillset. And, you know, I would say the ideas are getting to be kind of increasingly, I think, profound, and the skill sets are getting increasingly advanced. And part of it is, you know, kids now are growing up on the internet in some fundamental way. And so, you know, we, we see kids now from all over the world who walk in and they're, they are, like, far more educated, right, and, and skilled already to be able to do, you know, a startup or, or, you know, create new software or do any of these things, um, you know, than I was at their age, um, because they've, they've learned on the internet. Like, they've gone online, they've read, you know, they've read everything, they've, they've watched everything, they've Y…

AI assessment note: “Look, optimism is just like, it's basically the Morrison book, right?”

Q What did a political policy platform look like that encourages innovation? What are the sort of like key variables that matter?

A I don't know if it's too, I don't know if it would work, um, because it's a little bit too abstract, but it, it, it is kind of at the heart. It's a little bit at the heart of populism. It's a little bit at the heart of actually a lot of the left-wing critique, um, actually a business also. So you, you see, you see it actually, there's a critique from both the left and the right. And I think the critique, um, The way the libertarians have the answer, but the way the libertarians would put it is there's a difference between being pro-business and pro-market. Right. Um, and so, and, and what they mean by that basically is, it's like, there's one concept of a business, right? Or an entity, an entity of any kind. Let's start with the business, but we can apply the same thing to other, other kinds of institutions down the road. Like, um, there's one form of businesses like actually forced to compete in a fully competitive market. Right. So sort of fully subject to market discipline. Right. And so like, if whatever, you know, and I don't even, I say like the car industry works like this or something, right. Which is, or I don't even, let's actually the car industry gets repeatedly bailed out. So they're probably not the right example. I don't know, soft drinks or something. It's like, you know, Coke and Pepsi and Dr. Pepper and so forth. Like they, you know, every time you, you know, …

AI assessment note: “there's a difference between being pro-business and pro-market.”

Q A lot of people get scar tissue. They're still sort of like fighting the last war. You've been around the first internet, uh, big wave in the late 19 nineties and the failures that happened there. How does that affect your thinking? Like web van, for example, failed, but the same thing today might be a resounding success. How do you switch your mind to, um, not getting burned?

A Yeah, so this is really, yeah, so this is really fun. This is one of these things, again, I've just realized it's actually really fundamental to human nature. Um, and by the way, why I think they're, why I think you have to ask questions about whether life extension technologies are actually going to be a good idea. I think they are, but I think there's a real question about that from a, from a broader kind of cultural societal standpoint, which is right. It's the most, it's the, to your point, it's the most natural thing in the world to try to learn. From experience, right? And so I, you know, we all go through this, right? You touch the hot stove, it burns, we don't touch the hot stove again, right? And that happens at, like, age three or whatever. And then, like, from, from then on, basically, like, one way of just viewing, like, our trajectory through life is it's just touching hot stove after hot stove after hot stove and learning don't do that, don't do that, don't do that. You know, it's, or it's like, you know, don't act out at work, you know, you know, you learn all these things, and then one of the things you learn is like, okay, like, you know, actually what a lot of people learn is, what a lot of people learn is they, they, they work at a startup, the startup doesn't work, you know, the one that they're at doesn't work, and so the lesson is don't work at startups, r…

AI assessment note: “it's the most natural thing in the world to try to learn. From experience”