Sam Altman on Loving Community, Hating Coworking, and the Hunt for Talent

Conversations with Tyler is produced by the Mercatus Center at George Mason University, bridging the gap between academic ideas and real-world problems. Learn more at mercatus.org. And for more conversations, including videos, transcripts, and upcoming dates, visit conversationswithtyler.com. Sam needs no introduction. If we read the tech press and listen to other people's podcasts—

Oh no.

What's the single most likely misconception about venture capital they're going to have?

That venture capitalists are, like, on the whole, like, smart visionaries who know exactly what's going to happen in the world. The trick about— the way you become really great at venture capital is to accept that the only way to figure out the future of the world is to identify incredibly talented, smart, creative, original thinkers, and back those people. And you have to sort of trust that that will work out over time. You can't just be the smartest person all the time.

So if I'm trying to place your view in contrast with others, your view on how to spot and identify and mobilize talent compared to Marc Andreessen, is there any difference? How would I place you two?

Well, I, I do have my own strong views about where the world is going to go, uh, and what's important to support. And I think it is important to have strong views about the world, and it's a shame that most people don't. But I have, at least at the stage of venture capital we operate at, I have become so convinced that the only thing that my firm has to do is find the smartest, most talented people in the world, discover those people and enable them and make bets on them as people, um, and trust that they will figure out over time, they will drift towards the right ideas and the right kind of businesses.

But if you're trying to spot that talent, you must have methods you use, or ways of thinking about talent, that's different from the other venture capitalists. Otherwise, you would be interchangeable with them. And how would you place yourself relative to them? What do you put more stress on?

I think our greatest differentiator is not how we identify talent, although I will answer that question, but the fact that we treat our own business, we run Y Combinator in the way that we tell our startups to run. As a successful startup, which almost no venture capital firm does. Almost every venture capital firm gives advice they never follow themselves. They don't build differentiated products. They are not network-affected businesses. They don't try to build a brand and a community, and they don't try to make something that gets better the bigger it gets and have the scale effects that anyone would tell you they want in a business. And, you know, when we at Y Combinator always say we want to get a lot bigger because this is a network effect, this is a network that matters, And most venture capital firms will say out of one side of their mouth, "Oh no, like, this, you know, smaller is better." I think because they don't want to work more. And then they'll tell all their businesses, like, "The network effect is the only thing that matters." So I think many people are as smart as we are, think about the world in similar ways, but I think we have internalized that we run our firm the same way we tell our startups to operate. And we try to— we view the most important thing that we do is to build a network and a network effect. I will answer them.

What's the most binding constraint on your ability to scale, if that's the difference? Greater interest in scaling your own activity?

I mean, I think we have scaled to an incredible degree already, and we will over time figure out how to get another 10x and then another 10x after that, and someday we will fund all the companies in the world, but all the good ones at least. But I think it's unclear where the scaling limits are. I come from a software background, and one of the things that I learned is You can always scale systems more than you think, and you can never predict ahead of time where the choke points are going to be. And so you profile the code, you figure out where you need to optimize, and you optimize that. And that's the continual evolution of Y Combinator.

Now, this is from Wikipedia, and I quote: In functional programming, the Y Combinator can be used to formally define recursive functions in a programming language that does not support recursion. So why did you name the company Y Combinator?

So I didn't name it that. My predecessor Paul Graham named it that, but I think it's a very brilliant name because, like, really what that is is a function that starts other functions, and we are a company that starts other companies. The thing that I think is the coolest about Y Combinator and that is different about us spiritually than most venture firms is, or at least some venture firms, is we like to start companies that if we didn't fund them wouldn't happen. Most venture capital firms, and me in my own personal investing when I'm not doing Y Combinator, are looking for— you want to be a really great stock picker. You want to find a company that'll be successful whether or not you do anything and get them to take your money. And the thing that I think is so much fun about Y Combinator is we make companies happen that otherwise wouldn't, and that's the name.

Let me play venture capital skeptic and you can talk me back into optimism.

I might not.

So let's say— I say tech has had a stream of big hits: personal computer, internet, cell phone, mobile. So you've had a lot of rapidly scalable innovations become possible in a short period of time. We're now in a slight lull. We're not sure what the next big thing is or when it will come. And without that next big thing, won't the current equilibrium require a higher rate of picking the right talent than venture capitalists are in fact able to do?

Oh, I will talk you out of that one happily. It is the most expensive investing mistake in the world to make, is to be a pessimist. And it's a common one. That's actually, I think, the most expensive mistake to make in life. It is true that we are in a lull right now, but it is absolutely categorically false that unless the world gets destroyed in the very short term, that we will not have a bigger technological wave than we've ever had before.

But why can't I be an optimist, but not an optimist about VC? So I think new ideas will come through established companies. They'll be funded by private equity. They'll happen in China. But the exact formula where you can afford to make so many mistakes because the hits are so big, to what extent does VC rely on that kind of rapid scalability that may not come back?

It does rely on that. And that's why the industry came about really truly at a time when software happened. And we had this— software is very unlike any other good because the marginal cost is zero. Networks are so powerful. It's always tempting to say that the big companies in any industry are never going to— you're never going to compete with them. They will be the big ones forever. It is also possible that the next set of companies are all going to— the next set of $100 billion startups are all going to be synthetic bio companies or nuclear fusion companies or whatever. But the big companies in any moment always look like behemoths that you will never be able to compete with. 10 years later, it always looks so obvious about how they got beaten, or 20 years later, whatever it is. So I do think, it does even feel to me, myself, that this current generation of tech hypercaps, like, I'm almost willing to say this time it's different, but I know in my heart that there will be technological change that will happen at a rate that the companies can't do or require a level of risk they can't take. And it has been a bad bet to bet against new companies, you know, for centuries. And I think that'll keep happening.

Your conceptual model of venture capital, what do you take to be its origins? Is it the early voyages of discovery, finding the New World? Is it whaling ships? Is it Is it Silicon Valley in the last 50 years? Is it Hollywood movies? Where does it start for you?

When I studied the Industrial Revolution as a kid, it was all about technology and that you had this great idea, you invent this great thing, and now you can make this new thing. The way I would have explained the Industrial Revolution as a kid was there was this handful of inventions and they were really important and the world changed in a short period of time. Studied it again as an adult with the benefit of hindsight of my career, it was very clear to me that the most important invention of the entire Industrial Revolution was the joint stock corporation. You had governments say, "You know what? We have this second order kind of mini-sovereignty, and although the world has historically been limited to family businesses and small groups of trust, now we're going to allow big groups of people to coordinate." We're going to figure out how to align their incentives for capital providers and risk-takers and labor and everything else, and we're going to let these big groups of people work towards a common goal and a common vision with all the elements you need to make that work, and we the government are going to enforce that and protect that and let that happen. And that, I think, was the fundamental root.

You once said, "Growth masks all problems." Are there exceptions to that?

Cancer? I mean, clearly, yes. I don't mean that so flippantly.

There is an article in the Jerusalem Post today, someone credible claiming that cancer has been cured. I don't know if you saw that.

I didn't see that, but I do— having talked to many biologists working in the field, I will say there is a surprising amount of optimism that we are within a decade or two of that being true. I— it's not an area where I feel anywhere near expert enough to comment on the validity of that statement, um, and I think it's always dangerous to just trust what other smart people say, who, especially when they have an incentive to talk their own book, but it does seem like a lot of people believe that. Growth is bad in plenty of times, but it does mask a lot of problems. So a statement that I wouldn't make is that growth is always an inherent good. Although I do think, I think you've said something like this too, that sustainable economic growth is almost always a moral good. And something that I think a lot of the current problems in the country can be traced to is the decline in that. And part of what motivates me to work on Y Combinator and OpenAI is that I think getting back to that, getting back to sustainable economic growth, getting back to a world where most people's lives get better every year and that we feel the sort of shared spirit of success is, is really important. But there are definitely And growth feels good, so it does mask a lot of problems, but there are definitely individual instances where you'd be better off with slower growth for whatever reason.

Why is being quick and decisive such an important personality trait in a founder?

Ooh, that is a great question. Um, I have thought a lot about this because the correlation is clear. That one of the most fun things about YC is that I think we have more data points on what successful founders and bad founders look like than any other organization has had in the history of the world. We have that all in our heads, and that's great. And so I can say with a high degree of confidence that this correlation is true. Being a fast mover and being decisive, it is very hard to be successful and not have those traits as a founder. Why that is, uh, I'm not perfectly clear on, but I think it is something about the only advantage that startups have, or the biggest advantage startups have over large companies is agility, speed, willing to make non-consensus concentrated bets, incredible focus. And that's really how you get to beat a big company.

How quickly should someone answer your email to count as quick and decisive?

You know, years ago I wrote a little program to look at this.

Yes.

Like how quickly our best founders, like our founders that run sort of billion-plus dollar companies, answer my emails versus our bad founders. And I don't remember the exact data, but it was mind-blowingly different. I mean, it was like a— it was a difference of like minutes versus days on average response time.

Now, I come from Northern Virginia, the Washington, D.C. area, and we have very few geniuses, and the few that we have tend to be crazy and sometimes destructive. We have what I would call—

they're very stable, though—

a lot of upper-middle-class intellectual talent, people who are pretty smart and good at something. Tyler Cowen: When it comes to spotting good upper-middle-class intellectual talent, do you think you have the same competitive edge as with spotting geniuses who will make rapidly scalable tech companies?

I think that's how I'd characterize myself. Upper-middle-class, pretty smart, not a super genius by any means. It turns out that I've met many people smarter than me, but I would say I've only ever met a handful of people that are obviously more curious than me. I don't think raw IQ is my biggest strength. Pretty good, to be clear. But I think the chances of me winning a Nobel Prize in physics are low. And I think physics is a bad field at this point, unfortunately. I think what we spot— you do have to be pretty smart to be a successful founder. That's not where I look for people to be true outliers.

So given that self-description, assuming I accept it, and I'm not sure I do. Do you think you're as good at spotting upper-middle-class intellectual talent as superstar founders? Let's say we put you in charge of—

I think, and this, there's a statement here that's just bad about the world, but I think if you look at most successful founders, they are pretty smart upper-middle-class people. They are very rarely the children of super successful people. They are very rarely born in real poverty. They are very rarely the absolute smartest, you know, people who otherwise would win a Fields Medal. They are never dumb, but upper-middle-class, pretty smart people that have grit and drive and creativity and vision and edge, uh, and a different way of thinking about the world. That is what I think I'm good at spotting, and that is what I think are good founders. And there's a whole bunch of reasons that that's a sad statement about the world, but There it is.

So someone else has to find the geniuses.

Again, I don't want to go for false modesty here. I think I'm a smart person. I think our founders we fund are smart people. And I think I would have maybe said 10 years ago that raw IQ is the thing that matters most for founders. And I've updated my view on that.

How useful is personality psychology in the classic 5-factor sense for judging potential founders?

I mean, I know I've heard the 5 factors. I could pick them out of a list, but I can't sit here and name them. The thing that I Paul Graham, who started Y Combinator, used to talk about people being relentlessly resourceful, and it became this mantra of ours. And I think it's one of the most important traits in a founder. I think, like, a vision that someone feels compelled to make happen in the world is probably the most important of a super successful founder. And then this idea of relentless resourcefulness, like, that thing is so important to me, that I'm going to figure out how to get it done, whatever it takes, whatever I need to learn, whoever I need to convince. That's really important. You need to be a great communicator, which I think is a very underappreciated skill because you have to convince so many people to join you on this quest. You become— your job as a founder ends up being like sort of chief evangelist for the company, and most people aren't good at that. It's not something that we teach in school. We don't teach vision either. Actually, none of these skills we teach, but those are the kind of things we look for.

And do you think of yourself in terms of personality as more of a CEO type or more of a founder type?

I mean, I've done both. I think of myself more as a founder type. The way that I would say it is I think most people who run organizations at a certain scale are either good managers or good leaders. Um, and I would put myself squarely in the good leader. Weaker manager category. And that, I think, correlates more often with founders.

People who perform extreme physical events, climbing or— yeah, strength or running or marathons, as personality types, how do they differ from founders, if at all?

Not so much. That's something that I've learned to look for.

And what do you think is the difference? Because they're not founders, right? And the founders are not out—

a surprising number of YC's best founders are also into some sort of extreme physical something.

And why do you see that correlation? What do you think?

Um, something about focus and determination and drive to win and perform at your highest level. I think one thing that is a really important thing to strive for is being internally driven. So being driven to compete with yourself, not with other people. If you compete with other people, you end up in this mimetic trap and You sort of play this tournament and if you win, you lose. But if you're competing with yourself and all you're trying to do is like for the own self-satisfaction and for also sort of the impact you have on the world and the duty you feel to do that, be the best possible version you can, that there is no limit to how far that can drive someone to perform. And I think that is something you see even though it looks like athletes are competing with each other. When you talk to a really great absolute top-of-the-field athlete. It's their own time they're going against.

I'd like to try a little exercise. I'll throw out a few names from popular culture and you imagine—

I am unlikely to know these, but we can try it.

Imagine they show up at Y Combinator and how would you assess them? Spider-Man, Peter Parker.

Like, I watched one Spider-Man movie, but I don't— like, all I can think of is him going like that through the buildings.

So you wouldn't be that impressed?

I might be, I have to, like, let's keep going.

Luke Skywalker.

Someone else I might know more about.

Luke Skywalker.

Oh, Luke Skywalker I'd be impressed by.

Why? Isn't he indecisive? What? He focuses on the short term, he doesn't have a big picture vision.

No, he like turns away, like without, you know, like that, that, that when he like leaves Yoda and says, you know, I'm gonna go do this thing, which his like guru is telling him, you, you need more training. And he goes off and he does the right thing. On instinct and then comes back and Yoda says, "No further training, you know, you've got it." Like, that's a sign of a great founder. Like, you kind of, you go off before the world thinks you're ready and you do it anyway.

Let's say Hyman Rickover shows up at Y Combinator. What do you think?

I think building the nuclear navy is among the most impressive operational engineering accomplishments of human history, so we would definitely fund him. Good.

James Bond.

Um, Paul Graham used to say that the prototype for a good founder was James Bond. Um, there's lots of reasons the analogy doesn't work, but this idea that you are sort of— a lot of things go wrong, you are unflappable, you never lose your cool, you know how to react, bad things are happening to you left and right, and you are a survivor, and you always figure it out with whatever resources are in front of you. Like that part is the spirit of a great founder.

What about Q, the guy in the lab who shows James Bond the gadgets? His explanations are so to the point. Would that impress you?

Yes, with a good co-founder.

Austin Powers?

I don't think so.

Harry Potter? Hmm.

That was not one I was expecting.

Yes.

Let's try Harry Potter.

A single teen mother, Aretha Franklin.

Look, on the— yes, on the spirit of this idea that grit and determination is the most important quality in a founder, um, and I think being a single teen mother and coming through to the other side of that successfully is probably one of the hardest things anyone can do in the world.

But say filters we have set up to identify her as a single teen mother before she's successful. What is it we should be looking for in Aretha Franklin?

Oh, so this gets to the question of how— like, the most common question I get about Y Combinator is, how can you make a decision in a 10-minute interview about who to fund? And where we do miss— where we might miss her is at the upper filters of our application process. We have far more qualified people that want to do YC each year than we can fund. But by the time we get someone in the room, by the time we can sit across the table and spend 10 minutes with somebody, as far as I know, we have never made a big mistake at that stage of the process. And we've looked at tens of thousands of— well, in person we've maybe looked at 10,000 companies. These personality traits of determination and communication and the ability to articulate a vision for the world and explain how you're going to get that done. I used to think that that was so hard to assess in 10 minutes. It was maybe impossible to try. And YC interviews actually used to be like an hour. I now think that most of the time we could get it right in 5 minutes. When you have enough data points, when you meet enough people and get to watch what they, they go on to do, because the one thing that's hard in a 10-minute interview and the most important thing about evaluating someone is their rate of improvement. So that's a little bit hard when you only get a single sample. But when you do this enough times and you get to learn what to look for, it is incredible how good you can get at that. And we will make some mistakes, but on the whole, I think one way that I'd like to see Y Combinator scale is right now, you know, we only have maybe, let's say, 25,000 companies per year apply. We only get to meet, say, 2,000 in person. I'm sure we were making a lot of mistakes there. One thing I'd love to do is to figure out how to meet all 25,000 in person for 10 minutes. And I believe our error rate would go down incredibly low.

Young Napoleon shows up. What do you think after 5 minutes?

Uh, how young? Which, like, like 18-year-old Napoleon or 21-year-old Napoleon? Um, I mean, from everything I've read, that would be a definite yes. In fact, the best book I read last year is called The Mind of Napoleon. Which is a book of quotes about his views on everything. It's like that thick just of Napoleon quotes. Obviously a deeply flawed human, but man, impressive.

What kinds of insights did he have?

The thing that stuck with me the most, and this is a trait we see among many of our good founders, was not his operational excellence. It was not his vision. It was not any of the things that he, I think, is well known for that are all necessary if you want to do what he did in such a short period of time. But it was his incredible understanding of human psychology. And that is something we see among many of our best founders. And there was one quote there, he was— and it sort of felt quite applicable to the US in 2019. The French motto is something like freedom, brotherhood, and liberty. And he said he had thought about that for a while. Liberty, equality, and brotherhood. And he realized that the French people actually didn't care at all, although they said it and that, that was what it was. They didn't care at all about liberty, and they cared even less about brotherhood, and it was really just about equality. But it was tied up with this weird French ideal of love of status and differentiation as well. And so he talked about how you build a system given that people say this one thing but feel this other. How you sort of build a system where you can kind of control the people. And I was like, wow, like, I'm glad he does not run the United States because that is a dude who understands something deep that I did not and clearly was able to use it for power.

What's the most interesting or surprising thing you've learned or changed your mind about in the last 3 months?

I mean, a lot of things, but one that I have been thinking a lot about in the last 3 weeks is how I— watching the US political system react from this sort of cultural war on the left, sort of this anti-otherness from a cultural perspective to this economic— sorry, on the right to this economic war on the left. And like, there's— I've studied this so many times in history, and I didn't think it was actually going to happen. And I think when you get those two forces together fighting each other, it's so bad. I've updated myself to thinking now that things can actually get much worse in the US much more quickly than I thought.

What does that process look like if it continues for 5, 10 years?

I mean, pick your example of— we could talk about some of the pre-World War II stuff, but I don't think it ends well.

Before we move on to OpenAI, I think you should fund Peter Parker.

Tell me about him. Maybe we should. I just don't know enough to say.

Of course he's Spider-Man, but the impressive thing is he has these little devices on his wrists that shoot out web, and those are not Spider-Man powers at all. That's a technological device that he implemented, and the first time he actually tried it, he couldn't have really known for sure that it worked. So I would give a big thumbs up to Peter Parker.

Yeah, look, that's a good— that's a good sign.

And he has superpowers on top of that.

In some sense, like, the superhero I would most want to fund would be Batman because he doesn't have superpowers and he's just a resourceful, driven dude to do the right thing. And I like that better than people who just sort of got lucky with superpowers. But if Peter Parker fits in that category, okay, we'll fund him.

And his superpower comes from a kind of nuclear fusion, right? Even better.

Is that true?

Radioactive spider goo.

Okay, deal.

Yeah, deal. Okay. Make him a partner. Okay, so OpenAI, you have both a funding and a leadership role in OpenAI. Given the odd sort of event it's directed at preventing, how do you measure whether your money and time there are well spent?

I think, first of all, we're not directed at preventing an event. We're directed at making a good event happen more so than we are at preventing a negative event. It is where I spend most of my time now. I think it is— I think building true superhuman artificial intelligence and figuring out how we as a collective humanity decide what the world is going to look like on the other side of that and making sure that it is safe, we retain autonomy and a governance system we like, and that the benefits of that are equitably distributed in the world is the most important thing to work on. I think that there are parts of it that are scary. There are parts of it that are potential downsides, but the upside of this is unbelievable. I believe that, you know, all— we talked about sustainable economic growth as a moral good. I believe at this point real economic sustainable growth comes from technology, and I think this technology is going to generate more of it than the sum total that has ever been generated by technology in human history. So I think that's going to be really great.

Well, cyborgs outcompete AI. So if the flesh and blood is the hard part, why not tack on devices to people? Common in movies, but also in the real world. I'm wearing eyeglasses. There are hearing aids. You can imagine much more. Isn't that an easier way to get the work done?

You know, when, when AI started beating humans at chess, there was a short period of time where the very best thing of all was a human and an AI playing chess together. The AI would say, here's 6 moves. The human would pick the best one of those. Um, and that was better. That could beat an unassisted— that merged version, or that— not, that's not really a merge. That teamed up version could beat an unassisted AI. And I don't know exactly how long that stayed true for, and people loved that fact, but it didn't stay true for long. Um, the humans started making the AI worse than the AI was. Playing alone as it got smarter. And I think we will learn that we're just not that smart. Like that, you know, the size of a human brain has all of these biological limitations, but you can make a really big computer with very fast interconnects between the chips. And I suspect we will find that many versions of that merge you talk about are still not better. Now there may be other things that some people may choose to do, like maybe you upload your entire brain, and then you can add a bunch of stuff to it in the computer. But I think going the other direction, biology will be more of a limitation. And we have to, like, that's where we have to make a decision about the kind of future we want.

Let's say we institute a universal basic income. Since it seems from the data that citizens often hate the idea of immigrants just getting money for free and forever, isn't there a risk that a UBI would lead to a choking off of immigration in this and other countries? And that we would not on net end up helping humanity?

I think the argument against a universal basic income that I am most sympathetic to is that it is a good reason to lower immigration. I think increasing immigration, particularly high-skill immigration into the US, is one of the most important policy things we can do. So, I agree, that is a negative, but I think the positives of UBI are so high that I would take the trade, although I'd work as hard as I can at a minimum. To increase the level of high-skill immigration. I think if you believe talent is equally distributed around the world, which I certainly do, 5% of the founders we want are in the US. Speaking in our own self-interest as the US here, not as what's best for the globe for a minute, I'd like all the other 95% here.

Potential founders in other countries. I agree that talent is equally distributed around the world, but cultural background is not.

For sure.—

and if people from other places, including within this country, don't from early ages get the conceptual frameworks and positive reaffirmations from their parents, what is it the less developed parts of the world need to actually be contributing the proportionate share of founders, given that they do have the talent?

Nothing at all. They already are. If you look at the successful tech companies in the United States, the percentage of them founded by a first or second generation immigrant.

But they have very common backgrounds even though they're from different places. They're from parents who treated them a certain way when they were young.

Oh, I thought the question was specifically about immigrants.

So someone like Vitalik, who was not born in this country, he nonetheless has the perfect background for being a founder, and most people in the world do not.

The most important thing you can get to be a founder is to pick your parents well. That's a really sad statement about the world. If I weren't working at OpenAI, That is a problem I would like to work on. I suspect the answer is education, but that's only part of it. I do think the two great gifts I got in my life were the love of my parents and a great education. And I'd like to figure out how a lot more people get that. I think that's actually one benefit of a UBI. But I think that'd be transformative to the amount of human potential unlocked in the world. And just from a social justice perspective, it's so obviously the right thing to do.

Now, you grew up in St. Louis, right?

Yes.

I asked Larry Summers this exact same question. If you had, say, $200 million to distribute, you could do with it whatever you wanted to help St. Louis, just St. Louis, not the world. What would you do to help St. Louis?

This is very specific to St. Louis, but I think it may be true of other cities in the Midwest that were once growing and prosperous and now are less so. I'm such a believer that momentum is at all levels of organization, an individual's life, a company, a city, a country, is so important. Even though cities seem eternal and indestructible in a way that not many things, maybe religions also are, St. Louis has lost momentum. And there's been the talent drain that you would expect from that. And so I would spend all of the money trying to get a few really interesting high-growth startups that I was confident in to go move to the city of St. Louis and build enough of an ecosystem around them to make them successful and let that sort of regenerate the talent pool momentum of the city.

And what exactly is the check written for? There's a bonus for companies who—

So the thing that people normally do, which is predictably awful, is cities will open these coworking spaces. And say, "Hey startups, we've got this really cool coworking space and there's free espresso and colorful couches and come." And you get the people you deserve when you do that. So I'm very anti-coworking spaces, but I'll save that rant for another time. No, literally what I would do is make a venture capital fund, get really good GPs to do the advice and fund the companies to come be in St. Louis.

And you would give them special treatment, or you would make them jump through all the same hoops as in any other context?

I'd give them like a higher valuation or something. Um, but no, I would like make them be really great companies, but I would just invest in them on the condition they move to St. Louis. And I would do it like, you know, we talked earlier about the importance of a network effect. Rather than do this for one or two companies at a time, I would do it for enough at once and get them all in one part of the city to form something like what Y Combinator has done here.

Why is nuclear energy so unpopular in many parts of the world?

I think it is because most people are very bad at intuiting risk. They will happily breathe in coal smoke all day and then someday die of cancer and not connect that together. But if there's a nuclear meltdown and 3 people die, they will march against nuclear energy forever. Nuclear energy is by far the safest form of energy humanity has ever created. Much safer than solar panels because people don't die when they fall off the roofs when they're installing them, for example. But the disaster, like the way the world was introduced to nuclear power is an image that no one will ever forget of a mushroom cloud over Japan. It feels to some people like, you know, I've thought a lot about why the world turned against science.. And one answer of many that I am willing to believe is that image. And that we learned that, like, maybe some technology is too powerful for people to have. And that was like, people are more convinced by imagery than facts. And that was— that's the other side of nuclear energy.

If nuclear fusion has had a number of false starts, as you know, if it really starts to work, and I believe you're optimistic about that, What will it look like? How will we know? And how will we distinguish it from the previous false starts?

Well, we'll just get out more electrons than we put in. I believe we will get net gain nuclear fusion working in the next decade.

This is hot or cold?

Hot. It may not turn out to be economical. You know, we're still at the point where there's errors in the exponents, and that's always hard. It may turn out to just debug the engineering. Take so long that it doesn't matter. You know, if it like, it could be the kind of thing where we get fusion working in 2022 and the first commercial system isn't deployed until 2032 or longer. These are going to be complicated, not complicated like a fission reactor, but still complicated systems. But I believe we will get net gain to work. And I believe there's a good chance that the engineering won't be as hard as we think. And that the regulatory framework will be much easier than we think. And the economics will be great. And, you know, at any given moment in history, there is one cheapest source of energy, and whatever that is, in a small number of decades, usually takes over a significant part of the generating capacity of the world.

Will China do it first?

Whoever does it first, I'll be happy with. I don't— they don't tell me as much about the Chinese projects, so I don't know, but I'd be delighted if they did.

If we in some basic fundamental way hack biology through tech. How am I as a customer or patient most likely to notice this first? What's the New York Times headline going to say? What will my doctor give me?

I think we've hacked biology already to an incredible degree.

You know, life expectancy is just rising at a constant rate, right? It's even gone down for a few years in the U.S. That's not totally encouraging. It's not the fault of tech, but tech is not at risk.

Healthcare has been a catastrophic mess. I, I meant our ability to do other, like to make GMOs and, you know, we can do a lot with synthetic biology today. You know, one thing I was meeting with a company today that is looking at, and NYC had put out a request for more of this, about how you genetically modify bacteria in the ocean to pull carbon out of the atmosphere and stop climate change. That wouldn't have been something that anyone would have seriously talked about 10 years ago. And the fact that we can even have that conversation now I think is awesome. We may decide not to do it for good reasons. You know, we talked about growing up in St. Louis. One thing that my dad was really into when I was a kid was the Human Genome Project. WashU, which we live near, was one of the centers. And they sequenced a human for a billion dollars. And it was this huge deal. And I was a kid. It wasn't that long ago. You can sequence a human now for $100. You can take a sample of your blood and you can sequence every piece of DNA and RNA in there. And any bacteria or virus in your body, we have a perfect diagnostics for. So if you have an infectious disease, I guess unless it's a prion, we can just sequence everything and we have a perfect diagnostic. That's amazing.

Why is the tech world found education so hard to crack?

Why can the tech world not convince parents to love and prioritize their children?

Say a little more.

There's a lot of things technology can't do. There are a lot of things that require human connection. There are a lot of things that require people. And I think technology can do some things, but I don't think the biggest problem with our educational system is a technology one.

It's a human one.

For sure.

What is the biggest—

and by the way, one thing that I'm excited about is if AI really does unemploy 80% of the world, maybe we get a great teacher. Two, yeah, for every child in the world. Not just a teacher, but like, uh, someone to just like care about and love and make that kid the best they can be.

What is the biggest problem in the US financial system that you think can be fixed or ameliorated by a tech company or tech revolution?

Well, we all come at this with our own bias, but in terms of where I'm choosing to spend my own time, yeah, a lot more economic growth by funding a lot more great startups, building AGI and figuring out how to distribute that fairly among citizens of the world, and nuclear fusion. I would take any of those.

The, the financial system, banking, payments.

Oh, that, literally that. I thought you meant like the economic system. Like one, it's amazing to me, just studying history, how much the cost of energy correlates with quality of life.

Yes.

And so if like for the economic system, that'd be a huge win. For the banking system, I don't know. It's not Bitcoin, I'll say that. I don't feel uniquely or particularly well qualified to answer this, but I suspect it is doing something along the lines of providing fair financial infrastructure for poor people.

So cheaper loans, more responsible lending practices.

Something like that is what I would guess, but I haven't studied this.

What's the best-case scenario for crypto? I'm not suggesting you're predicting it, but how might it go best?

So I'm excited about universal basic income. I don't think countries are going to implement that anytime soon. I can imagine a crypto system where you see something that is more powerful than any government on Earth, where you actually figure out a way to give every person on Earth a coin, and then you make this gigantic network that everyone believes in, and you can do redistribution outside the control of any government. And, you know, we talked a bunch of times tonight about network effects. That would be the most powerful network effect the world has ever seen economically, and I think that would be cool.

Much of economic and social activity ever moves into virtual reality? Will we use crypto in virtual reality or just trade dollars?

I'm only laughing because whenever people put two buzzwords together, I've learned from YC interviews to really get skeptical.

Good ideas should have like one big thing that really works, right?

One big thing.

Yes.

One big thing. I don't know. I don't think that would— I think we could still use dollars.

What are we going to do with all the empty big box stores?

I've been thinking about that.

Yeah.

So it's an interesting question. I don't have an answer.

Community centers, all these tutors who are put out of work by AI, they meet up with children there.

I would take really cheap housing. If we could fill them all up with really cheap housing, I'd be pretty happy. Look, I think one of the things that's gone badly wrong in the United States is we have now pursued for a long time a policy where we want housing to be an investment, not a consumption asset, and thus grow faster than the rate of inflation. And after you let those curves compound out for many decades, you get the predictable disaster we're in now, where you basically steal the future from young people. And if you could pick one national policy that would help, I think, with redistributing economic opportunity, the cost of housing near high-quality jobs, that would be a great thing to do.

How optimistic are you about the prospects in the Bay Area of having greater freedom to build in a way that would say lower prices by 20% compared to baseline?

Well, I will caveat this by saying I think if you believe what I believe about the timeline to AGI and the effect that'll have on the world, it is hard to spend a lot of mental cycles thinking about anything else. And so I have not thought deeply about what it would take to solve really any other problem in the last few years. But I don't feel optimistic given prior performance that the Bay Area is really going to do the right thing on.

On the country as a whole? Do you think the Bay Area is especially unlikely? Will Atlanta become a new land of NIMBY, and in turn Houston, Dallas, someday Chattanooga? Or how do you see this evolving?

Yeah, at some point, and actually I think it may be close, the network effects that have made the Bay Area so powerful for technology startups, the costs will overcome that. And I think that'll be a real opportunity for other cities.

Other than the Bay Area, of course Israel, which is connected to VC networks here, but what's the undervalued geographic center for VC in the next generation that people should be talking about more?

My current belief about that is that it's not— now that the world has kind of woken up to this, it's not going to be one center. It's just going to sort of be more diffuse everywhere.

But here there's so much clustering, and the land prices reflect it. You want to be near other people. You're on boards of companies. You want the company to be with you. Venture capitalists often induce the company to move next to them.

For sure.

What will change that?

What I meant more is, I think instead of one new center or two new centers, there will be 30, and there will just be clusters in different places that don't quite get to the density of the Bay Area but get beyond critical mass. And I think we'll just see that diffuse throughout the world.

So let's say you visit a city abroad. You've either never been or haven't been much, and you have 2, 3 days to look around. Is this a promising center for venture capital? What is it you look for?

I have a funny metric on that, and it doesn't probably work for most people, but it is the number of times I get stopped on the street for a selfie.

It's a good metric. And it correlates positively.

At least with, like, startup energy in that city, yeah.

Yeah, yeah. Charter cities, is that a potential venture capital idea, or is that just something different?

Yeah, yeah, that's cool. I think the world, like, there is just this desperate need for a better, more affordable way to live in better functioning cities. Like, cities are so important in the current world to economic opportunity, and they deny it to so many people now that I think people are really willing to try something new. And I think there will be, there's enough customer demand there that there will be economic support for it.

As our final segment, let me ask you a few questions about what I call the Sam Altman production function. Are you game?

Sure.

What did you learn from Looped?

Uh, you cannot create a market that does not want organically to exist.

What did you do wrong as a 19-year-old?

We only have a few more minutes. Um, I didn't spend enough time thinking about what to work on. I just let myself get pulled along. And I was probably pretty obnoxious and a little bit of a jerk. I think I corrected for that quickly, but I wish I had done it faster.

In the world of tech startups, venture capital, what is weightlifting correlated with?

Weightlifting?

Weightlifting. Oh. So, Taleb tells us in New York City, weightlifting is correlated with supporting Trump. But I doubt if that's true in many—

It's not true here. I mean, I think it's correlated with successful founders. It's like fun to have numbers that go up and to the right. And like the most fun thing for me about weightlifting is like I have— I am— I'm basically financially illiterate. I can't build an Excel model for anything. I can't read a balance sheet. But my Excel model for weightlifting is beautiful. Because they're numbers that go up and to the right, and it's really fun to like, you know, play around with that.

What makes you a good poker player?

So learning— I played a lot of poker in college, and I think I learned more about life and business from that than I learned in college. I would not say I'm a great poker player, but I'm pretty good. And the thing that makes me, I think, good about that is getting good at quickly evaluating risk.

Now, I'm from the Northeast. When I watch racing cars, I see a bunch of little things on TV go around, around the track, and I'm totally bored. What am I missing?

Well, it's not that fun to watch, but it's very fun to drive.

Fun to drive.

There are very few activities, I think, that are high enough adrenaline to totally stop thinking about work, and racing cars is certainly one of them. But watching it is not that fun.

Is regret rational?

Yes, and I think it is rational both to live your life aiming to never regret anything. Like, I try— if I ever am making a decision and I feel like I'm gonna regret not doing something, or I have regretted not doing something, I try to go do that thing right then. But on the other hand, I think like if you can't change it, if it's done, then it is maybe not that productive to regret, but it is a deeply human thing and I think sometimes helpful.

What is a source of beauty or potential beauty in your life that you would like to have more of or plan on having more of soon?

For the first time I think since I was a kid, I have really been able in the last few years to take more time to A, just do nothing with people that I love, and B, marvel at everything in life and figure out how to look at things for the first time again. And that has been a source of great beauty.

And what's the binding constraint for those of us who want to do more of that? Is it we need to devote more time or somehow open our minds? What's the path to get there? Is it a free lunch?

For me, it was finding someone to really explain that to me and kind of getting to a point in my life where I was ready to do it.

So like beauty mentors?

Sort of, yeah.

Now, one of your greatest achievements, you said, was reinventing Y Combinator after you took it over. When the point comes, how will you do this again once it's needed?

How does the next person reinvent Y Combinator? How do I reinvent it?

Whoever it will be.

I think the key to doing that for any organization I think it is actually harder to reinvent an organization than it is to do it the first time. I think the hardest thing in business is an organization that produces repeated innovation and success because loss aversion is so high, resistance to change is so high, and when you have something that's working, putting it at risk to make it work better is not a bet that ever feels comfortable. Internalizing, truly deeply internalizing, that most things in life are not as risky as they seem. Most things in life are two-way doors. You can come back. And that even if people that you work with or in the press or whatever, like, call you idiots and say, oh, like, that doesn't mean you have to listen to them.

Can you ever appease your critics when people call you idiots?

I try not to think about them.

Try not to think of— now, last question.

Well, that's actually not true. I try to listen to every piece of criticism, assuming that it is true and assuming that I'm in the wrong. But once I have reflected on it and decided that I'm right and this person is just engaging in bad faith, then I try to put it out of my mind entirely. I put no effort into trying to appease them.

Last question. Let's say a very smart and credible 19-year-old comes up to you and says, I would like to be the Sam Altman of the next generation. Not an exact copy of what you've done, But the next, the 2.0 version, what advice would you give to that 19-year-old?

I mean, I don't know what the Sam Altman of my generation is going to look like yet. Like, I hope that I'm not done here.

Other than move to St. Louis.

No, I like— well, I think the memetic stuff is always super dangerous, but I don't think you should ever try to be somebody else. Or like, I don't think you should ever even try to take too much inspiration from someone else's life path. Like, I still hope that the most important work I have to do by far is in front of me. And I hope that's the case until the day I stop working. And I think that's what I would encourage people to do, is don't shoot for something that somebody else has already done. Shoot for something that no one has done yet.

Sam Altman, thank you very much. Thank you. We now have some time for questions at the two mics. Please keep in mind, these are not speeches from you. We are here to hear Sam, a speech or long-winded statement, I will simply cut off. So we will start with the microphone over here, please. First question.

In terms of narratives in the country, I think we're seeing a lot of fear and a lot of going— people going to simpler narratives because it makes them feel more safe. And I think you talked maybe last year, the year before, at Disrupt, and you talked about a startup that was looking to take over the role of a church. In the community, and I was gonna ask you if you have thought more about that kind of interaction between simpler narratives and maybe what they were trying to do and maybe where that might go in the future. I think the thing that has gone wrong in the tech industry, and maybe the country as a whole, is we have— we, this area of the world, have failed to present a vision of the future where everyone gets to win and that everyone's excited about. And I think we're unlikely to get good at that fast enough, so I'm actually excited for the tech industry to figure out how to collaborate with Hollywood or other industries that are good at narrative and paint a vision of the future that is optimistic, because pessimism is always easier than optimism. Demagogues always have an easier time winning elections with pessimism than optimism. And I think people can kind of sense that whether they like it or not, tech is going to have a lot of influence over their lives. And I think it's up to us to paint that positive narrative. And in fact, a lot of what I hope OpenAI does in the next couple of years is figure out how to tell the positive AI story, not the negative one.

Next question.

So I heard you speak a little bit about physicists tonight. I myself am a physics PhD student. I'm an aspiring physicist. All right, that makes two of us. When I get back to Berkeley tomorrow and I tell everybody that you might not think physics is the place to be, what should I— what should I tell the physicists to do? They're smart people, they're hardworking, but maybe they're doing the wrong thing. I actually think that there is plenty of fertile ground in the field of physics. I think most people aren't working on it. Everyone wants to solve the Grand Unified Theory. Somehow or other. That is every physicist that I've ever met's dream. Maybe that's the wrong problem to be going after. Maybe more physicists should be working on better plasma physics models so we can get nuclear fusion sooner. Or maybe more physicists should come work at OpenAI because physicists are usually the smartest people. Many people in the field have gotten trapped in various dead ends. One problem with being really smart is that really smart people tend to be super memetic. And so most physicists I've met have the same aspiration, work on the same problem. Maybe it's just too hard and there are all these other problems. And so my advice to all the physicists at Berkeley would be spend less time thinking about the problem you're working on today that you're not making progress on and more time thinking about what problem no other physicist is working on that would be really important to the world and that maybe you can solve.

A question from the iPad: Please give us your anti-coworking spaces rant.

Oh, alright, I will. So I used to say there are 3,000 YC clones in the world. I have since learned that there's 8,000 in China alone. So let's say the whole number in the world is 16,000. Every time someone decides they're going to build the next Y Combinator, the idea is always the following: it's going to be just like Y Combinator, except we're going to provide free coworking space. And I always want to say to people, like, did you ever think that maybe we thought about that and it's a feature, not a bug, that we don't have it? I think the single thing that has differentiated YC more than any other decision we've made is that we do not have a coworking space. We bring the companies together once a week, but that's it. It's enough for a community, but it is enough to build your own identity. Coworking spaces have two big classes of problems. Number 1, they are a bandpass filter. Good ideas— actually, no, great ideas are fragile. Great ideas are easy to kill. An idea in its larval stage, all the best ideas when I first heard them, sound bad. And all of us, myself included, are much more affected by what other people think of us and our ideas than we like to admit. And if you are just 4 people and your own door, and you have an idea that sounds bad but is great, you can keep that self-delusion going. And if you're in a coworking space, people laugh at you and like, no one wants to be the kid picked last at recess, so you change your idea to something that's like, sounds plausible but is never gonna matter. It's true that coworking spaces do kill off the very worst ideas, but a bandpass filter for startups is a terrible thing, 'cause they kill off the best ideas too. The other thing is the average level of ambition and willingness the willingness to work hard at a coworking space is incredibly low. And so there's just this like reversion to the mean that is not what you want in your life.

Next question.

So I know we played, uh, would be fun or not, but we didn't play overrated or underrated.

Can we do that? False one out. Antarctica, overrated or underrated?

Underrated. I went there for Christmas. Awesome.

What was great about it?

Um, it felt like being on another planet. And I think as you get older, it gets harder and harder to like truly have new perspective on things. And let me tell you, like having no internet, no other humans around, slash plants or animals, being extremely cold and having nothing to look at but like white and rock, man, was that a mental reset. That was really cool.

Another question from this side.

As AI gets better, you were talking about in chess, I think you said biology becomes a liability. As AI gets better at thinking than we are, so then what becomes the most important work of humans? Love, creativity, how we treat each other. We're going to find out. I don't know. I was talking to someone recently about all the different ways this can go. There are versions, and I think a lot of people, a lot of sort of like techno-utopists love them. Where we all just merge with AI and there's no opt-out. Either you merge with AI or you become irrelevant. And that sounds cool probably to a lot of people in this room and terrifying to 99% of people on Earth. And, you know, we talked about the need for positive stories. I don't think that's the story that most people want to hear. I don't think that's exciting. I don't think that's a future that most people are excited about. But the thing that is I think really interesting about this is like, sure, we're going to build this thing that's like smarter than humans in every way, but we get to make a lot of decisions about how to build it. You only get to make them once. You know, when the Von Neumann probes launch from Earth and eventually they never get to communicate again because of the speed of light, you better have gotten it right. But we do get to make them. And like, I don't know what the answer is. I do know I don't— should not get to deserve to make it by myself. But like maybe we as humans as a whole decide the deal we want with AI is humans get Earth and a 100 light-year radius and the AI gets the rest of the universe, but we get to live our lives as we want and maybe that's okay. So I think the option space is big, but we won't be smarter.

An iPad question. You mentioned some of the qualities of good and bad founders. Of the qualities of bad founders, which do you think are the most curable?

Arrogance. You want to be confident but not arrogant, and they look superficially to an outside observer very similar. And if a founder is intellectually honest and wants to be better, to my surprise, this is something I've changed my mind on, arrogance can often be cured if someone is otherwise well motivated.

And what cures arrogance other than you?

Getting beaten up by the market sometimes does it. But just teaching people that, like, you can operate this way if you want, and you will be unhappy and people won't want to help you, or you can make this slight tweak to the way you interact with people, the way you treat people, and the way you treat your team, and the way you listen to feedback, and get this very different result. My experience is if someone is genuinely intellectually curious, you only need to show them that one time.

Question over here.

Um, how will YC do better to support the 90 or 95 or 99% of talented people that it's maybe missing or isn't supporting at the moment? Well, I think the most important thing we can do is grow. But that will take some time. You know, it's always easy to look back at decisions and be like, ah, I knew that thing was going to work and be really important. And it's almost always bullshit when people say that. But the thing that I did personally that surprised me the most for that was in 2014, I think, I taught this class at Stanford, but also broadcast, which I, without their permission, they were very against it put online, called How to Start a Startup. And I still get, like, dozens of emails per week from people around the world who, like, watched those videos and wrote and said, "Hey, like, I never knew about this. I never knew this was an option, and this changed my life." That taught me, and then I've done a bunch of things since, that even if we can't fund every startup, we can put the information out there for every startup. And we have tried to open source everything we know about startups. And at least make that part widely available. We tried to create more communities outside of just the founders we fund. We're going to try to launch another one tomorrow. We have things like Hacker News. We tried to make a bunch of groups online. But we will continue to put as much knowledge out there and create as many communities as we can.

From Harshida on iPad: What are some systematic problems in Silicon Valley that prevent certain particular problems from even being noticed? That's a bit of my paraphrase.

I think one thing is just people don't get out of Silicon Valley enough. I don't do this as much anymore as I used to, but I used to make it a point to like, I would try to travel like 12, 16 weeks a year, and I would try to like meet people in very other contexts or just observe the world like way outside the US. And I think that gave me a broadness of perspective that was super valuable, and it's something most people don't do.

Question over here. Regarding the experiment on UBI in Finland, could you comment on the results of that?

Didn't they pull that experiment? The Finnish, I thought they stopped that experiment. Oh yeah, so I don't know how to comment on the results, but I would have loved it if there were results to discuss.

From the iPad, if pessimism is indeed the biggest handicap, What do you do to actively cultivate optimism?

You know, I think one of the most important questions that I would like an answer to is that, is optimism versus pessimism nature or nurture? And if it is nature, is that, can we fix that with CRISPR? Because it is such a bad thing to have. I hope it's a personality trait. I hope it's something that can be improved with deliberate practice. I believe in my heart that it is, but I try so hard to avoid pessimistic people in my life that I don't feel like I have enough of a dataset to answer that with any confidence.

What's the optimal percentage of pessimistic people in your social circle?

You do want some, and I have plenty, but maybe it's like 1 in 10. 1 in 50, something like that. I believe you said that the best way to create wealth is to start a startup. On the other hand, you're the successor to running Y Combinator. You were almost this apprentice-seeming figure before you took over, which seems almost like the least founderly thing to do. So what should we learn from this? Should we start a startup unless we have the option to run Y Combinator, or what? Look, if you have the option to run Y Combinator, I think you should That's a pretty good thing to do. But the way that Y Combinator generates wealth is still by startups getting started. And I think, like, as the fundamental mechanism, sure, like, I don't start them, but I do think I build something that helps them do better. And what I meant more by that statement, and you don't have to, like, start it yourself. You can join as an early employee. You can be an investor. A more precise version of that would be The best way to generate wealth is to own a significant piece of a startup. And there are many ways to do that. Like, you can choose to be an investor, you can choose to be an early employee. You don't have to be a founder. In fact, that's one thing that I think Silicon Valley gets wrong, is that like either you're a founder or you suck. That is like clearly not the case. But I do think startups are the best way to generate huge amounts of wealth quickly.

Next question.

There's a famous book which came out in the year 2000 called Bowling Alone, which documents the decline of social capital in America. Do you think that decline has continued through to the present day? And if so, what, what solutions do you think there are to solve it, either technological or otherwise? I'm not familiar with the book. Could you explain social capital in this context?

Yeah.

So social capital refers to a broad set of factors, including shared norms, values, cultural understandings that sort of act as a lubricant in a society or group. To enable it to function more effectively? Got it. Great question. There's all this data that the things that make people happy are most of all is this sense of community and belonging and sort of frequent community celebrations and things like that. And that has declined for sure badly. And I feel like even in my own life, I've watched that decline in the last 30 years. And even moving from— San Francisco is just an awful city for a sense of community.

Tyler Cowen: Why is that?

I mean, I could answer that, but it will take— Do you want a full answer on that? It's a long one.

Sure. Yeah.

I think in cities that become dominated by a single industry and cities that reward generation of wealth and financial success over a sense of sort of shared humanity and community, have a very hard time preserving that fabric of social capital, community, whatever you want to call it. Like, where I grew up, no one would ever walk by a person collapsed on the side of the street on their way to work and not do something about it. And I hope I never get used to the fact that that happens in San Francisco. I do blame the tech industry for a lot of things that have gone wrong with this city, but not all of them. But we have just over time had this unbelievable wealth generation in this small geographic space in this small period of time, and I think not been particularly thoughtful about the effects of that on the community as a whole. And because those problems are so hard and so hard to think about, I think most people just choose not to and they just accept this. I think we also have fairly incompetent city government and city management. We have people that are sort of unwilling to stand up to certainly the greatest housing crisis I hope I ever see, uh, and drug one as well. But if we have a city government that has been unwilling for the entire time I've lived here to make hard decisions that are unpopular with vocal minorities of the voting base, um, and so we have unaffordable housing, we have a homelessness epidemic, um, and we have streets that are like Like, again, there's many things I love about this city, but the fact that like, if I take my trash out late at night when it's dark and don't put on shoes, I have to like look for hypodermic needles outside my house. It's like, really? That's what, that's what we want to do here? And so I think you need like strong, powerful community leaders in a way that the city hasn't had. And you need, you need people who are willing to embrace change. You can't keep the city the way it was. That's not gonna happen. That ship has sailed. And so right now you have people who in their unwillingness to make any changes are just letting everything be awful. And the only thing that I think will work is say, okay, like the game has changed. The city has to change in big ways. And you need a leader empowered to do that. And it's gonna make some people unhappy, but the current, like the complete lack of humanity and civic connection we have And the fact that, like, yeah, I'll never get used to people not saying hi to each other on the streets. I'll never get used to sort of just ignoring people, like not even stopping to, like, call an ambulance for people. I don't think I'll ever get used to.

Last question is from me. The American Midwest, underrated or overrated?

Underrated.

Why?

You know, as a kid, it was an idyllic place to grow up. As a teenager, I did get kind of bored. And as an adult, I feel such a strong pull back. Like, people are nice to each other. There are seasons. Like, one thing that I think is surprisingly awful about San Francisco and the social fabric of the city is it's never warm at night, or it only is like 4 nights a year. And this sounds like this very minor thing, but what it means is you have no sense of life on the streets, and you have no people like sitting outside for dinner happy and talking to each other as they pass. Like, when you go outside in San Francisco, in the evenings, like you're guarded against the world because you're cold all the time. And so there's not this life on the streets of the city. And I think it actually has this huge effect on the social fabric.

Sam Altman, thank you very much.

Thank you.

Thanks for listening to Conversations with Tyler. You can subscribe to the podcast in iTunes, Stitcher, or your favorite podcast app. And if you like this podcast, please consider rating it on iTunes and leaving a review. This helps other people find the show.