Garry Tan Decides Who Gets to Build the Future

Speaker 1: The Builders is a production of Kaleidoscope and iHeart, presented Speaker 1: by Ramp. I'm Walter Isaacson. Welcome to The Builders, my Speaker 1: series of conversations with the most consequential minds shaping our future. Today, Speaker 1: I'm speaking with Gary Tan. Y Combinator has become one Speaker 1: of the most powerful and important brands in Silicon Valley. Speaker 1: The companies that emerged from YC include Airbnb, Stripe, DoorDash, Speaker 1: and Dropbox. Together, they have transformed how we travel, spend money, work, Speaker 1: and communicate. Now, Gary Tan is deciding which builders come next. Speaker 1: Gary first entered Y Combinator as a founder. He became Speaker 1: an early investor in Coinbase and Instacart. And today, He Speaker 1: leads YC itself. And so in some ways, he's helping Speaker 1: shape those who shape the future. Here's my interview with Speaker 1: Gary Tan. Gary Tan, welcome to The Builders. Speaker 2: Thanks for having me. Speaker 1: As a leader of Y Combinator, you have to pick Speaker 1: founders who are going to make it. Do you look Speaker 1: for one particular thing or are there like 20 different Speaker 1: models of founders that can work? Speaker 3: Well, so the wild thing about Y Combinator is that Speaker 3: we get 80, maybe 90,000 applications a year now for Speaker 3: about 800 spots. So it is less than a 1% Speaker 3: acceptance rate. Maybe 8 to 10% of those people would Speaker 3: get an interview, sometimes less now. And the wild thing Speaker 3: is we actually spend only 10 minutes with them. for Speaker 3: that final stage. But the interesting thing is, you know, Speaker 3: Paul Graham, the founder of YC, figured out that you Speaker 3: didn't need half an hour. You don't need an hour. Speaker 3: You could probably actually do it in about five minutes, Speaker 3: but five minutes feels rude. So that's how YC ended Speaker 3: up at 10 minutes. Speaker 1: So what did Paul look for and what do you Speaker 1: look for in those 10 minutes? Speaker 3: I mean, it sounds almost too simple, but 10 minutes Speaker 3: is a wonderful time. constraint, kind of like how 140 Speaker 3: characters for Twitter way back in the day was a Speaker 3: wonderful constraint. If you can say something very simply that Speaker 3: hits us, like, you know, here's a user, here's a Speaker 3: real person who has a problem, and here's how the Speaker 3: technology that we build actually solves that. We just want Speaker 3: to know what it is and why it's awesome. Speaker 2: And then if we have a. Speaker 3: That founder is the person to go and build that thing. Speaker 3: And you can do all of that in 10 minutes, Speaker 3: and it's an up-down decision. Something like 10% to 15% Speaker 3: of all self-made billionaires as of present time, age 50 Speaker 3: and under, are YC alums. Speaker 1: Let me ask you a real quick question then. Suppose Speaker 1: Elon Musk walked in and you had 10 minutes with him. Speaker 1: What do you think the percentage chance is that you Speaker 1: would actually fund him? Speaker 2: Oh, 100%. I mean, in current form, 100%. No, no, no, no. Speaker 1: I'm saying you saw him. You didn't know he was successful. Speaker 1: He was a guy who'd just been, say, ousted from PayPal. Speaker 3: I mean, I had a managing partner meeting the other day, Speaker 3: and that's actually one of the example questions we ask Speaker 3: ourselves to figure out whether we're on the right track, right? Like, Speaker 3: by definition, if YC is operating properly... Like if an Speaker 3: Elon or a Bill Gates or a Steve Jobs or, Speaker 3: you know, a Drew Houston from Dropbox, a Brian Chesky Speaker 3: from Airbnb, Brian Armstrong from Coinbase, Tony Hsu from DoorDash. Speaker 3: Like if someone like that walks into our office, can Speaker 3: we say yes to them in 10 minutes? And if Speaker 3: that answer is no, actually, I'd have to rewrite everything. Speaker 3: I would have to retool everything, right? That is actually Speaker 3: almost the only question that really matters when it comes Speaker 3: to whether YC is doing well or not. Speaker 1: And if you had walked in years ago, would you Speaker 1: have accepted you? Speaker 2: I think so. Speaker 3: My startup, Posturus, it was dead simple blogs by email. Speaker 3: And I mean, I had a pretty banger demo. So Speaker 3: we had something that I think Paul and Jessica and Speaker 3: Trevor Blackwell, the founders of YC, really got immediately. They Speaker 3: didn't grill us at all. Like we came on with Speaker 3: a really good demo and they said, well, that makes sense. Speaker 2: And then we got funded. So sometimes it's that simple though. Speaker 1: And I was creating that company, running that company for Speaker 1: a while. How fundamental was that to your ability to Speaker 1: be at Y Combinator and to be making venture capital decisions? Speaker 3: Oh, I think it's quite important. I mean, There's a Speaker 3: Jay-Z line that I love. One of his songs, he goes, Speaker 3: everybody wants to tell you how to do it and Speaker 3: they never did it. And so, I mean, the entire Speaker 3: act of being an investor in technology is actually bewildering Speaker 3: and strange if you really think about it too long, Speaker 3: because what we're doing is selling money. Like tech has Speaker 3: now become such a fundamental force in the financial world Speaker 3: of society that that there's way more money that wants Speaker 3: to go into tech than good opportunities. Speaker 2: And that's really what YC is. Speaker 3: And if you look at most venture capital, it's crazy Speaker 3: that it's grown so much. The S & P 500 Speaker 3: is almost entirely driven by tech. And then the wild Speaker 3: thing that Paul Graham figured out and Jessica Livingston figured Speaker 3: out when they first started was, when they started in 2005, Speaker 3: the type of people who would get funding, it would Speaker 3: be your buddy at Stanford Business School, your golf buddy, Speaker 3: like maybe someone who your son went to school with Speaker 3: at like some prep school at Harvard Westlake or what, Speaker 3: you know, I mean, it was like access, right? Speaker 1: As it was for a few millennia. Speaker 2: Time immemorial, right? Speaker 1: But do you all think you broke the back of Speaker 1: that or is being a startup junkie or a startup Speaker 1: entrepreneur something that you still have to be pretty privileged Speaker 1: and pretty connected to do? Speaker 3: I mean, it would be a mistake to say that Speaker 3: we were actually a meritocracy, but I will say and Speaker 3: affirm loudly that we aspire to be. Speaker 2: We will never be. You know, I think that the Speaker 2: woke age of sort of like five. Speaker 3: Or 10 years ago, we sort of lost sight of that, Speaker 3: that like meritocracy could never be achieved, but it is Speaker 3: a virtue to attempt to achieve it. Speaker 1: Well, the reason I ask about that in terms of Speaker 1: a startup culture, which is what you very much nurture, Speaker 1: is that it used to be somebody coming out of college, Speaker 1: take a Steve case, for example, could go work at Speaker 1: a company like a Procter & Gamble and learn a business. Nowadays, Speaker 1: it's very hard to sort of get a steady job. Speaker 1: And the economy for young people tends to depend more Speaker 1: and more on startups. And yet it seems to me Speaker 1: that unless you got a safety net, unless you're somebody Speaker 1: who doesn't have to make sure your next meal can Speaker 1: be paid for, you can't be in that game, and Speaker 1: it leaves a lot of people out. Speaker 3: I mean, that's why YC gives $ 500, 000, for instance. When Speaker 3: I did YC, it was only $ 18, 000. Like today, you Speaker 3: hear about seed funds and Stanford kids raising $ 30 million Speaker 3: on $ 300 million post-money evaluation. It's like literally, like it's Speaker 3: just an idea and one person, and the company is Speaker 3: somehow worth $ 300 million. But when we were starting many Speaker 3: years ago, It was a very small amount of money. Speaker 3: You had to show that you had something that was great. Speaker 2: And so that's why. Speaker 3: YC didn't stay. And we easily could have just stayed Speaker 3: at $ 12, 000. Speaker 2: Or $ 15, 000. Speaker 3: But I think there is a recognition that if we Speaker 3: want to be a thing that allows people to participate, Speaker 3: you don't have to have rich parents. The half a Speaker 3: million dollars getting into YC, you could just have an Speaker 3: idea and you could come in and meet us for Speaker 3: 10 minutes. Speaker 1: You were one of the earliest employees at Palantir. You Speaker 1: got to see how Peter Thiel and Alex Karp in Speaker 1: particular really worked. What did you learn from their style? Speaker 3: I realize now, like, if I hadn't rage quit, I Speaker 3: would have been, you know, sort of a peer to Speaker 3: some of the, you know, real legends who really built Palantir. Speaker 3: You know, I would go on walks with Karp and Speaker 3: he would sort of, I mean, he was a philosopher. Speaker 3: You know, he was not a technologist, I would say, Speaker 3: but he was incredibly knowledgeable about the world. One time, Speaker 3: I was sent to DC to do some demos for customers. Speaker 3: And I remember being like a 25-year-old, and I was Speaker 3: in my ripped jeans and band t-shirt and getting sent Speaker 3: to stay at the Mandarin Oriental in DC. And we Speaker 3: would wonder, Alex, why are we staying at the Mandarin Oriental? Speaker 2: Look at us. We're just kids. Speaker 3: We don't have to do that. Speaker 2: Just set me up in the Motel 6. We're a startup. Speaker 3: Alex looked at me and my other early engineers for Speaker 3: the Palantir Finance product, and he said, guys, you're going Speaker 3: to go and talk to such-and-such general or such-and-such VP Speaker 3: of some Fortune 500, and you're going. Speaker 2: To show this incredible software that you've built. Speaker 3: And then as you go out, you're going to make Speaker 3: small talk, and they're going to ask, well, how long Speaker 3: are you around and where are you staying? And if Speaker 3: you say Motel 6, they're going to think that they're Speaker 3: buying Motel 6 software, and Palantir is... Mandarin Oriental software. Speaker 3: And I'm like, okay, now I get it. Speaker 2: And so, you know, yeah, I don't know. Speaker 3: I think embedded in that story is just a lot Speaker 3: of different things. And I learned a lot from Alex Speaker 3: from that. Speaker 1: Don't go anywhere. There's a lot more. right after this break. Speaker 1: I know you've been fighting this antitrust battle. How important Speaker 1: is that in order to keep innovation and startup culture? Speaker 3: I think, you know, True free market enterprise has never Speaker 3: been tried. I think that in tech in particular, we Speaker 3: have very short memories. I would argue that Google itself, Speaker 3: one of the big tech companies, was able to exist Speaker 3: and thrive and become a trillion-dollar company because of the Speaker 3: actions of the DOJ with the consent decree on Microsoft. Speaker 3: Microsoft was ready to take over the web, and it Speaker 3: used its platform advantages in Windows to default install and Speaker 3: take over the browser market. Speaker 1: And do you think Google is now doing the same thing? Speaker 2: I would argue yes. Speaker 3: The Delaware C Corp sort of actually insists on it. Speaker 3: Like the governance structure itself actually insists on you're supposed Speaker 3: to do your duty and your duty is to maximize Speaker 3: shareholder value. So people are following the letter of the Speaker 3: law and the spirit of like their Delaware C Corp Speaker 3: to close platforms. Like it's less and less percentage of Speaker 3: the screen real estate if you do a Google search Speaker 3: is even non-Google properties, right? Speaker 2: And so that's self-preferencing. Speaker 1: So a lot of your tech bro friends are very Speaker 1: much against antitrust enforcement. How do you have a startup Speaker 1: culture if you're going to allow big companies to self-preference? Speaker 3: Well, I mean, my view is that we shouldn't allow self-preferencing. Speaker 3: And then my super pro startup stance is we should Speaker 3: stop self-preferencing. We should require open markets. Like I think Speaker 3: the DMA in Europe, for instance, is a great example Speaker 3: of forcing big tech to open its own platforms and Speaker 3: not just data transparency and privacy, but also portability and Speaker 3: reverse engineering. These are sort of things that allow for Speaker 3: open markets and open access to startups. And if you Speaker 3: ask me, is big tech too powerful? I say absolutely. Speaker 3: And the answer and the antidote is little tech. I Speaker 3: think it's not that unlike robber baron times in that Speaker 3: if you own the railroads and there was a monopoly Speaker 3: on the railroad... you could set the price on literally Speaker 3: everything else. And so one thing would lead to another thing, Speaker 3: and you would have multiple monopolies across both commodities and Speaker 3: the railroad the commodities travel over. Speaker 1: Well, wait, let's look at AI. I'm going to push Speaker 1: back a bit. We get a whole lot of new companies, Anthropic, OpenAI, Speaker 1: they're all popping up. Somehow Apple wasn't able to self-preference Speaker 1: and embed its own system in its own phones. Even Speaker 1: Google is trying to do that where it almost forces Speaker 1: you to use Gemini when you're using Google search. And Speaker 1: yet these other companies are doing well. Speaker 2: Yeah, that's right. Speaker 3: I mean, I think what's funny about the AI companies Speaker 3: is like they're almost certainly, I mean, I think you Speaker 3: can fairly say that they are the new big tech. Speaker 3: And so I think I have concerns there as well, right? Speaker 3: Critics of some of the AI companies would say that Speaker 3: It's possible that as the AI wars progress, there might Speaker 3: not be open access to the latest and best frontier Speaker 3: models that those companies may gatekeep. Speaker 1: You think they should be open, that all the AI Speaker 1: models should be open? Speaker 3: I mean, that would be certainly the thing that supports Speaker 3: startups and the ability for new people to create new Speaker 3: companies and, frankly, new wealth and to give consumers choice. Speaker 1: So what is the role that Y Combinator wants to Speaker 1: play in AI? Speaker 3: I mean, what's funny is I think of YC almost Speaker 3: as like, you know, we're not a union, but we Speaker 3: have union-like characteristics. There's not really a group out there Speaker 3: that can speak full-throatedly about policies and, frankly, actions that Speaker 3: both governments and companies should take to make sure that Speaker 3: the golden goose can still exist. And, you know, whether Speaker 3: that's the billionaire tax in California or antitrust laws in D.C. Speaker 2: Or in Europe. Speaker 3: we're going to start talking about it and just having Speaker 3: a seat at the table. One of the things I Speaker 3: discovered as I started examining politics, most tech people are Speaker 3: not interested in politics, but politics sure is interested in us. Speaker 2: And so I decided. Speaker 3: Like three or four years ago, like, oh, well, I Speaker 3: should probably spend some time and reciprocate. Like, let me Speaker 3: understand what's going on. And what I discovered was pretty wild. Speaker 3: I mean, millions of dollars in AstroTurf money, going to Speaker 3: groups that would take around basically almost hired actors. They Speaker 3: would claim to speak for small software companies or, you know, Speaker 3: little tech. Speaker 2: And they were just like paid chills, actually. Speaker 3: And they would go to the halls of power, whether Speaker 3: it's Sacramento or D.C. And they would say, no, no, Speaker 3: like this thing that actually mainly helps big tech. We, Speaker 3: the little guy, really wants that. And I was like, oh, wow, Speaker 3: this is crazy. Like no real startups actually go out Speaker 3: there and talk to real people in D.C. And I Speaker 3: saw a system that was broken, and I wanted to Speaker 3: fix it. And so that, you know, we do very Speaker 3: little things. Like, we have a tiny budget. It's, you know, Speaker 3: probably one one-hundredth that of, like, even any given single Speaker 3: big tech firm spends on policy. We have one lone Speaker 3: full-time hire in D.C. in a man named Luther Lowe. Speaker 3: And we run it on a shoestring. But on the Speaker 3: other hand, like, we're people-powered. I see that like a Speaker 3: repeating pattern, like there are broken things in society. There Speaker 3: are systems out there that, you know, have money, like Speaker 3: just trace the money and like, you know, where's the Speaker 3: money coming from and who benefits, right? Speaker 2: And then the heartening thing. Speaker 3: You know, I think a lot of people talk about Speaker 3: social media in a very negative way. They think, you know, Speaker 3: it was Facebook that ruined democracy. And it's like, I Speaker 3: don't really, I don't think I agree. Like, I think Speaker 3: that It's the little guy not realizing he has a voice. Speaker 3: He and she do have a voice. And then, you know, Speaker 3: when I look at YC, like, our job one is Speaker 3: to find the next great founders. But I think right Speaker 3: under that, it's like, how do we make sure that Speaker 3: those people can have access to markets? How do we Speaker 3: help people who would not have a voice, would not Speaker 3: be able to go to D.C.? How do we give Speaker 3: them that megaphone? And then YC itself now has a megaphone. Speaker 3: You know, we have more than 2 million subscribers on YouTube. Speaker 3: We can sort of speak in aggregates. And I'm really Speaker 3: excited about that because I think that that's us trying Speaker 3: to speak truth to power. I don't want to deny Speaker 3: that YC is also a power in and of itself, Speaker 3: but the world is full of systems. These are all Speaker 3: systems that were created and constructed by, as Steve said it, Speaker 3: people no smarter or better than you or I. And Speaker 3: we can mold it and change it. And once you Speaker 3: realize that, you'll never be the same. Speaker 1: You were born in Canada, Chinese background. Tell me how Speaker 1: important you think immigration is to the United States. Speaker 3: I mean, I shout to everyone who will hear that Speaker 3: about 50% of all the startups that really make it Speaker 3: have at least one immigrant co-founder. And what we realize Speaker 3: is that's one of the things that makes America incredibly unique. Speaker 3: People from around the world want to come here. And Speaker 3: in tech, they want to come and build the future. Speaker 3: That's certainly my story. My mom's family came over from Burma. refugees, Speaker 3: and my dad came here from Singapore. Speaker 1: Well, you talk about your dad, and we biographers tend Speaker 1: to believe it's sometimes about dad, especially when somebody's successful. Speaker 1: You had a pretty rough time, though. I mean, your Speaker 1: father was at times unemployed, at times an alcoholic. Tell Speaker 1: me how you had to deal with all that. Speaker 3: I mean, it really made me who I am today. Speaker 3: I figured out that I can actually put myself back Speaker 3: into let's say 1992, I'm 11 years old. And basically, Speaker 3: I don't know when my dad would come home, like Speaker 3: maybe around five or six, but I wouldn't know how Speaker 3: many beers he would have already had. And I realized Speaker 3: there's sort of this scanner that happens where if you Speaker 3: don't know what father is coming home that day, then Speaker 3: you get very attuned to other people and their interstates. Speaker 3: And like, how to sort of react to that. And Speaker 3: then it's interesting because now I realize like that's a Speaker 3: scanner that I use all the time. Like I can Speaker 3: sort of sit down with someone and it doesn't happen Speaker 3: every time, but it happens relatively easily for me where Speaker 3: I can sit down and just kind of feel what's Speaker 3: going on for other people as someone who sort of Speaker 3: spends time with hundreds of startup founders every single year. Speaker 2: Like that's actually a pretty powerful superpower of. Speaker 1: Let me talk about what may be the next step Speaker 1: in AI, which is AI agents. And you've said that Speaker 1: you don't want to fund companies that look at AI Speaker 1: agents as a feature. You want them to look at Speaker 1: it as the core operating system. Explain how you think Speaker 1: that changes what we think about when we talk about AI. Speaker 2: Yeah, absolutely. Speaker 3: I mean... We're going to get a little sci-fi here, Speaker 3: but this is something that we're starting to see ourselves, Speaker 3: you know, rebuilding the way we operate at YC. I Speaker 3: think the number one thing that people get wrong is Speaker 3: they just think that AI is just chat. Speaker 2: And then it is so much more than that. Speaker 3: We now have an internal AI agent, for instance, named Straylight, Speaker 3: that kind of has a personality and then it has Speaker 3: a company brain. Like it actually has the total context Speaker 3: of everything. Founder we've ever backed, anyone who's ever emailed us, Speaker 3: like all the applications. The way I think about it Speaker 3: from an interesting tech perspective is like a human can Speaker 3: only keep track in their heads like seven plus or Speaker 3: minus three things. And what an AI agent is, is Speaker 3: already super intelligent. It can keep, you know, I was Speaker 3: explaining this to my 11 year old. What an AI Speaker 3: can do is it can keep a stack of 1000 pages, Speaker 3: think like three Harry Potter books. You know, it can Speaker 3: needle in a haystack, any page, any fact, anything that Speaker 3: happened in those three Harry Potter books. It can just know, synthesize, Speaker 3: work with. Like the working memory is absolutely gigantic. And, Speaker 3: you know, the thing is in another year, it won't Speaker 3: be three Harry Potter books in the brain of an Speaker 3: AI agent. It'll be five or it'll be 10 or Speaker 3: it'll be 20. And then the set of things that Speaker 3: it can operate on is even bigger. Speaker 1: Well, give me a better example and make this more Speaker 1: concrete for me, because data retrieval, even from three Harry Speaker 1: Potter books, That's kind of cool, but that's not, I mean, Speaker 1: that's just data retrieval. What makes something an agent? Speaker 3: I mean, I would point to what Peter Steinberger of Speaker 3: OpenClaw figured out, which is if you use cloud code Speaker 3: or codecs, I would say right now, they're so worried Speaker 3: about sort of repeatability and safety that they actually trade Speaker 3: off all of the beautiful parts of working with an Speaker 3: AI agent, which like basically can be so similar to Speaker 3: a human that like it sort of takes on human qualities. Speaker 1: So tell me what you ask your agent to do. Speaker 2: Oh gosh. Speaker 3: I mean, one of the things I did, I, you know, Speaker 3: it's read all of my email and all of my Dropbox. Speaker 3: It's read like all of my Apple notes, like, you know, Speaker 3: even my therapy notes going back 20 years. I mean, Speaker 3: for instance, like I used it for prep to be Speaker 3: able to talk to you, not to get too meta Speaker 3: about it, but in about 15 minutes right before this talk, Speaker 3: like it was able to you know, download and transcribe Speaker 3: your top 10 podcasts you've ever done. And then it Speaker 3: actually diarized them. So, you know, I had an LLM Speaker 3: go like line by line, timestamp by timestamp to pull Speaker 3: out like major themes. Speaker 2: And that was really helpful for me. Speaker 3: But to be able to do that for like almost Speaker 3: any thing that you do in your day-to-day life and Speaker 3: for it to do it like somewhat more autonomously. And Speaker 3: so it seems inevitable in about two to three years, Speaker 3: I would say, Everyone's going to do it, and it's Speaker 3: going to feel awesome. Think of it as a Jiminy Speaker 3: Cricket on your shoulder at all times. Almost anything you Speaker 3: want to do, it knows you, and it's there to help. Speaker 3: And it will actually sort of run ahead and say, Speaker 3: what about this? Or it'll help you see around corners. Speaker 3: And the reason why it's too expensive right now, I mean, Speaker 3: it's just literally too expensive. Speaker 2: In order to have this experience. Speaker 3: I spend about maybe $ 100, 000 to $ 200, 000 a year on tokens. Speaker 3: And so I realized that's a crazy privilege that not Speaker 3: everyone can do that. You know, a lot of like Speaker 3: CEOs that I know are doing it and they have wings. Speaker 1: Do you think that AI, especially the agents you've been describing, Speaker 1: will end up destroying jobs or will it end up Speaker 1: increasing enormously the number of jobs in society? Speaker 3: My hope, I mean, I don't know for certain and, Speaker 3: you know, I'm not clairvoyant, but I, you know, I Speaker 3: do believe that technology is over vilified. I think that Speaker 3: it does have problems. And AI safety is something that Speaker 3: we need to be concerned about. But the way to Speaker 3: fix that is not to stop the technology from even happening. Speaker 3: It is actually going after the abuses of people when Speaker 3: people actually use or misuse the tech. Speaker 1: But do you think it'll create jobs? There'll be more Speaker 1: jobs because of AI? Speaker 3: That's certainly my hope, right? In the short term, it Speaker 3: may not. And that's quite frustrating. So some of the Speaker 3: reason why I've been working on open source and trying Speaker 3: to teach people how to use AI, like my open Speaker 3: source project, GStack, is literally working backwards from if I Speaker 3: were trying to teach YC founders how to use cloud Speaker 3: code and get 400 times the amount of output that Speaker 3: they had before, that was like a project that they Speaker 3: could just drop into cloud code and get exactly the Speaker 3: stack that I use to get that. I think things Speaker 3: like that, are necessary in order to teach people how Speaker 3: to use the tools properly. And there will be people Speaker 3: who lose their jobs because AI will have, like, sort Speaker 3: of made that position redundant. And then, you know, I Speaker 3: think there's a big retraining issue here. But, you know, Speaker 3: I think something like G-Stack or lots of other efforts Speaker 3: out there could solve that, right? Speaker 2: If you have that. Speaker 3: Context, if you were doing job X, like, let's say Speaker 3: you're a medical biller, like, What I want is for Speaker 3: someone like that to maybe be able to pair up Speaker 3: with a 22-year-old. They have domain expertise and someone who really, Speaker 3: really knows how to make 400 times the amount of Speaker 3: software that you could before. Maybe that person who knows Speaker 3: that space inside out plus someone younger or even themselves, Speaker 3: like train yourself to do it. I want them to Speaker 3: be able to start a company in that space. And Speaker 3: maybe it's exactly what they did before. Maybe it's something Speaker 3: two clicks away, but either way, it requires an increase Speaker 3: in the amount of agency. What we want people to Speaker 3: do is go into their own history and their own lives. Speaker 3: And it's like, what do you uniquely know? What life Speaker 3: did you live? What obscure hobbies do you have? Like, Speaker 3: what have you seen that nobody else has seen? And Speaker 3: then pair that with this incredible ability, like not just Speaker 3: to code, but to think, to reason. And so I Speaker 3: think that's like raising the floor. And then on the Speaker 3: flip side, like the ceiling is sky high. Like I Speaker 3: think that the responsibility for CEOs and people who have Speaker 3: a lot of privilege is not to cost cut. Like Speaker 3: that is the easy way. And the thing is, like, Speaker 3: that can result in, like, a sort of dystopian moment where, Speaker 3: you know, capital accumulates more and more wealth, and we're Speaker 3: not sharing it. Speaker 1: We'll be right back after this quick break. You say Speaker 1: you push back against people who obviously vilify AI, but Speaker 1: those who are very pessimistic about it. When you're looking Speaker 1: at companies to back or fund or encourage, do you Speaker 1: consider an ethical component, whether what they're going to do Speaker 1: is going to be driven by being good for society? Absolutely. Speaker 3: I mean, we have so many examples at YC where Speaker 3: a team would come in and it was pretty clear Speaker 3: that it was going to make a lot of money. Speaker 3: The most extreme version of this I remember was, this Speaker 3: was not when I was back at YC, but I Speaker 3: have to call out my partner, Gustav Alströmer. He was Speaker 3: head of growth at Airbnb for many years, and then Speaker 3: he became a partner here at YC. And a company Speaker 3: that was focused on ADHD meds for kids came through, Speaker 3: and they applied, and it was a remarkable set of founders. Speaker 3: But the idea was tantamount to like, increasing the distribution Speaker 3: of ADHD meds to a society that, you know, we Speaker 3: couldn't tell if, like, that was a need, right? Like, Speaker 3: it was definitely going to make money. And in the end, Speaker 3: that company did get sanctioned by the FDA, and it Speaker 3: shut down. And I even believe that there were potentially Speaker 3: criminal charges about that. Speaker 2: I hadn't checked recently, but that was a bullet dodged, right? Speaker 3: And I'm very proud that YC is an institution where Speaker 3: that kind of thing can happen. Like, We not only Speaker 3: need to make money, really we have the responsibility to Speaker 3: fund the things and the futures that we want to Speaker 3: live in. Speaker 1: Give me an example of a couple things that would Speaker 1: fund the future we want to live in. Speaker 2: I mean, let's see. Speaker 3: I mean, the hard part is like it's almost anything Speaker 3: that we actually fund. Speaker 1: Well, I'll give you one, which is have social media... Speaker 1: that instead of dividing us and enrages us, helps us Speaker 1: be stronger as a society? Speaker 3: I think that's entirely possible. And actually, this idea of Speaker 3: personal agents is actually, I think, very directly a way Speaker 3: to do it. Going back to meta, right? Like, this Speaker 3: is a huge company that controls the algorithm. And we Speaker 3: all wish that they didn't at some level. And now Speaker 3: they might not have to. So you could imagine a Speaker 3: personal AI that you control, that you program, You don't Speaker 3: download the Facebook app or the Instagram app. That's just Speaker 3: an API. And you do have a social network, but Speaker 3: the social network is mediated through your own cognitive filter. Speaker 3: You figure out what you want to see, and then Speaker 3: you put that in there. And maybe that could even Speaker 3: be programmed by not just you, but the people who Speaker 3: are closest to you. I think that that's one example Speaker 3: of actually being able to put control of what goes Speaker 3: in your brain back into your own hands. And that's Speaker 3: something that, you know, before wasn't really possible. Like even Speaker 3: the idea of like, could a normal human being sit Speaker 3: there and program their own algorithm that is beneficial directly Speaker 3: for them? Well, A, they don't know how to program. Speaker 3: And B, like, what would that even look like? It's Speaker 3: so complicated. But with the advent of AI agents and Speaker 3: AI agents that can code, it'll know you, it'll know Speaker 3: what your goals are. Like right now, my year right now, Speaker 3: I just want to focus on in myself. I don't Speaker 3: want to see anything about national politics. I might only Speaker 3: care about local city politics. Speaker 1: Yeah, but we've had things like that in the past. Speaker 1: We've had personalized social media, and yet the algorithms have Speaker 1: a business model that works best when they enrage or Speaker 1: divide us. Speaker 3: Yeah, and the business model is hosted by someone else, Speaker 3: by a big tech company, and then that's their control Speaker 3: and that's their power. And so what personal AI is, Speaker 3: what I hope happens is sort of more like the Speaker 3: personal computer revolution, kind of what Steve Jobs did in 1984. Speaker 3: If you remember the original Mac commercial, you know, 1984 Speaker 3: doesn't have to be like 1984. Big Blue, you know, Speaker 3: IBM does not have to win the whole thing. And Speaker 3: I think next year will be one of the defining Speaker 3: years where there's the next browser war. And I think Speaker 3: it's going to be whose AI agent do you use? Speaker 3: And I think it'll go similar to operating systems. Like, Speaker 3: I think there will probably be a Mac. There will Speaker 3: probably be a PC. Speaker 1: So you're saying all these anthropics, open AI, Gemini, it'll Speaker 1: consolidate to just two or three? Wouldn't that be pretty horrible? Speaker 2: I mean, I think we're already there. Speaker 3: And, you know, I think last year, you know, our Speaker 3: friends inside some of the AI labs talk about the Speaker 3: five kings, which is actually a pretty astonishing vision where Speaker 3: a light cone of all future value, literally like all commercial, Speaker 3: anything that a business would do could potentially collapse into Speaker 3: five kings. Speaker 1: So what can you do to disrupt that? Speaker 2: We're trying every day, right? Speaker 3: So whether it's policy, whether it's directly being the funding Speaker 3: vehicle to find the next generation of talent, to help Speaker 3: them thrive, to make sure that the markets are open Speaker 3: for them, to make sure that there's open source, to, Speaker 3: you know, find, I mean, what, What I want is Speaker 3: there to be Linux, right, for personal AI. Speaker 1: Basically an open source operating system? Yeah. Speaker 3: And it just has to be easy enough to use Speaker 3: and cheap enough to use that everyone chooses it. Speaker 2: And I think that's possible. Speaker 1: Do you think we should have regulation then that pushes Speaker 1: in that direction? Or is that an impossible dream? Speaker 3: I wouldn't say anything is impossible, right? And I think Speaker 3: that that might actually be a preferable outcome to some Speaker 3: of the other things that Some of the politicians are Speaker 3: putting out there, you know, Bernie Sanders wants to sort Speaker 3: of seize assets. And speaking as an immigrant and children Speaker 3: of immigrants who came to America to flee asset seizure, that, Speaker 3: you know, we want prosperity. We want the ability for Speaker 3: people to start new businesses to proceed and continue. And Speaker 3: that's one of the defining traits of America. Speaker 2: It's like. Speaker 1: But so what would smart regulation look like? Speaker 3: I think it basically just looks like the Open Markets Act, Speaker 3: like being able to create an app that actually operates Speaker 3: on your iMessages, for instance. Speaker 2: Like, AI can't touch your iMessages. Speaker 3: And what Apple says is, like, we're the only ones Speaker 3: who could possibly be safe enough for you. And I Speaker 3: think that that's just not actually true. We shouldn't allow Speaker 3: them to say that, like, nobody else could possibly do this. Speaker 3: And that's just a smokescreen for, A, their own market Speaker 3: power and making money through their own modes, but also, B, like, Speaker 3: it actually just kills markets. And, you know, the next Speaker 3: Apple can't emerge, right? If Google, Apple, and all the Speaker 3: trillion-dollar companies don't open a lane and create markets, we Speaker 3: will live in a world where there aren't startups. That's Speaker 3: a bit of a nightmare scenario for me. I think Speaker 3: that this is the counterweight against tyranny and, frankly, mismanagement Speaker 3: in big tech. You need to have alternatives. And so Speaker 3: to the extent that the government can create regulation that Speaker 3: makes sure that these markets are open, at the extreme, Speaker 3: I would even say, you know, looking at how open Speaker 3: source models are progressing now, we are just handing Chinese Speaker 3: model companies a massive advantage. There's zero open source AI Speaker 3: companies in the United States. And it doesn't have to Speaker 3: be that way. Like there are probably regulations that could Speaker 3: allow distillation, which is the exact advantage that Chinese companies Speaker 3: have for American and European companies. And, you know, we Speaker 3: can have a vibrant ecosystem. Some of these things being Speaker 3: too big, I think that's real. Speaker 2: And that's correct. Speaker 3: And if you go all the way back to Brandeis Speaker 3: and trust busting, we live in a better world when Speaker 3: there isn't concentration in markets and that there isn't self-preferencing. Speaker 3: And I think that that is more true than ever Speaker 3: when it comes to tech and startups. Speaker 1: Well, in some ways, you're really going back when you Speaker 1: talk about Louis Brandeis and you talk about the railway companies. Speaker 1: You're really saying that we've been through this before. which Speaker 1: is the 1890s. An explosion of technology is even greater Speaker 1: than we have today. And then you create these great Speaker 1: robber barons that control it. And then you have a period, Speaker 1: whether it be Teddy Roosevelt or Louis Brandeis, that take Speaker 1: on Standard Oil or take on the railways and say, Speaker 1: you got to open up. You got to allow everybody Speaker 1: else to ride your rails. You think we could just say, hey, Speaker 1: let's learn from history and apply it now? Speaker 2: I think so. Speaker 3: I think we have to learn from history in two Speaker 3: different ways. I mean, I think when you look at Speaker 3: the Sanders and AOC end of things, or in San Francisco, Speaker 3: Saikat Chakrabarty, or even what Tom Steyer was saying in Speaker 3: his failed bid for governor in California, I think you're Speaker 3: starting to see a thing that is, to me, like Speaker 3: fundamentally un-American. Speaker 1: But wait, let me ask you, what are you more Speaker 1: worried about? That a populace on the left are going Speaker 1: to have some socialist system, or that we are going Speaker 1: to have a technology system that gets too controlled by Speaker 1: too few people, and it's not democratizing and it doesn't Speaker 1: incent startups? Speaker 3: I mean, honestly, both scenarios are terrible. And I would Speaker 3: stand against both of those scenarios. And I think that Speaker 3: there has to be a middle way. I think that Speaker 3: we should follow American history. And how do we create markets? Speaker 3: How do we not make it illegal to create prosperity Speaker 3: and wealth? Speaker 2: I don't begrudge Google. Speaker 3: I think there are people on the hard left, for instance, Speaker 3: that would say, like, Google should not exist. And we Speaker 3: want more Googles. We actually want more. OpenAI was a Speaker 3: spin-out of Y Combinator. Sam Altman sat in my seat Speaker 3: and said, you know, he got to see the front Speaker 3: row of all the builders, all the young people with Speaker 3: crazy ideas. And he and Ilya Suskovor and Greg Brockman said, Speaker 3: They sat in a room and created YC Research that Speaker 3: became OpenAI with Elon. And now we live on their timeline, right? Speaker 3: Now we live on an accelerated timeline where the age Speaker 3: of intelligence is much closer. And so I don't think Speaker 3: that's a bad thing. But, you know, AOC and a Speaker 3: great many Americans today unconscionably believe that all billionaires are ill-begotten. Speaker 3: And when people say that billionaires shouldn't exist, they're also Speaker 3: saying that society should ossify, that LLM should not exist. Speaker 3: It also is saying that things like Google that, you know, Speaker 3: organize the world's information that shouldn't exist. And to that, Speaker 3: I say, like, I, you know, I disagree, you know, like, Speaker 3: we should have more billionaires, we want more choice, we Speaker 3: want more prosperity, we want more jobs, there is a Speaker 3: middle way, it doesn't have to be big tech oligarchy. Speaker 3: But it also doesn't have to be mass asset seizure. Like, Speaker 3: I think there's a middle American way. And That's what Speaker 3: we very much will fight for and hope for. Speaker 1: Gary Tan, thank you so much for being part of Speaker 1: the builders. Speaker 2: Thank you so much for having me. Speaker 1: Silicon Valley is very good at finding people who think differently. Speaker 1: It rewards ambition, competitiveness, and ideas that can scale. But Speaker 1: the interview made me think about the fact that so Speaker 1: much of Y Combinator's energy is focused on AI. The Speaker 1: biggest opportunity isn't simply to build what the market will reward. Speaker 1: It's to build what humanity actually needs. And if AI Speaker 1: could focus more on helping humanity cure cancer, fight climate change, Speaker 1: rather than just making chatbots and threatening our jobs, there Speaker 1: would be a lot less backlash against what the industry Speaker 1: is doing. So I'm curious to watch how Gary and Speaker 1: Y Combinator will place their bets. Because when you get Speaker 1: to choose who builds the future, You have some responsibility Speaker 1: for the future they build. That's it for today's episode. Speaker 1: I'm Walter Isaacson. We'll see you next time on The Builders. Speaker 2: The Builders is a Kaleidoscope and IHAR production. Speaker 1: This series is hosted by Walter Isaacson. Speaker 2: It's executive produced for Kaleidoscope by Oswald Lishan. Speaker 1: David Weisbord, and Mungesh Atikudur. Speaker 2: Sheena Ozaki is the producer. Julia Nutter is supervising producer. Speaker 2: Eva Zorich is our consulting producer. Speaker 1: Fact-checking by Eamon Whalen. Research by Gloria Thiel. Speaker 2: Maria Paz Mendez-Hodes is head of video. Marketing by Christy Speaker 2: Mirabal and Elle Liu. Production support by Dara Luk-Potz. Speaker 1: Music for the series was created by Spencer Stevenson, a.k.a. Speaker 2: Botany, and it was mixed by Steve Bone. From iHeart, Speaker 2: the series is EP'd by Allie Perry and Katrina Norvell. Speaker 3: Special thanks to Bob Pittman, Conal Byrne, Will Pearson, John Sykes, Speaker 3: Nikki Etor, Carrie Lieberman, Binky Urban, Julia Prosser, Jeremy Blum, Speaker 3: Alison Goldberger, WYES, and ASL Studios. Speaker 1: For more from Walter Isaacson, be sure to check out Speaker 1: his newest paperback, The Innovators, how a group of hackers, geniuses,