Sam Altman on Building OpenAI & Betting on the Impossible

I just brought up Tobi Lütke and the fact that I recorded with him previously. Why do you say that you think he's one of the most interesting CEOs right now?

One of the things that struck me the most about Tobi is that, in the very early days of AI and then at every moment along the curve of its development, he has been the most forward-leaning CEO. He's in there writing the software himself. He is experimenting with it. He sends us extremely detailed feedback on the product offering and on the capabilities of the models.

Before anybody else was saying this, he was like, "We are not an NPC company, and thus we are going to adopt agents. Otherwise, we're totally screwed. We're going to build it ourselves." Every time I talk to him, he is at the edge of what anyone, CEO or not, is doing. He builds it himself. He understands it. He has a great, deep feel for it, and he is always 6–8 months ahead of any other CEO. Do you remember when he wrote this is probably like a year and a half ago maybe 2024 he wrote that letter saying that like the first thing you have to do is see if AI can solve your problem and then even back then it was I don't know 18 months 24 months ago people went crazy they thought it was ridiculous ridiculous this is my point like he's just consistently been ahead he's been correct he's leaned in he's very no [__] so there's like no hype there's nothing other than like here's what it can really do right now here's what I think it'll be able to do soon here's how I'm going to push the company here and just like extremely deep understanding of where it's at.

I never even thought of how much of a benefit it must be for somebody in your position to have someone like that giving you intense, very direct, and clear product feedback. A lot of people send product feedback. He is the only person at the intersection of being the CEO of a large company and providing extremely accurate, detailed, cutting-edge feedback.

He told me—I don't know if it was on the episode we recorded or if it was afterward—but he was very adamant. He's like, "We're going to look back on 2026." I actually want your opinion on this. I didn't even think to talk to you about it. "We're going to look back on 2026 as the year that every business was up for grabs."

He said somebody was going to build the AI-native version of Shopify. And he said, "It's going to be me." So, at night, he's literally trying to rebuild it.

If you started from scratch, what would you do with the current technology? That was the other thing I was going to say about him: he does it himself. He is using these tools himself. He's writing software himself. He's trying the models himself. He's trying to reimagine his workflows himself.

Most CEOs, when you get to that level, have teams of people managing teams of people who are trying to implement the thing. They're trying to make you happy, and they're trying to smooth the rough edges. I think it's very hard to get the feel for something if you're not actually doing it, and he does it so hands-on all night long, as far as I can tell.

I'm not sure if 2026 will be the year that every business feels up for grabs. I might disagree with him a little bit there, but I get the spirit of it, and I do understand that it feels like that's happening. Do you think that's even possible—whether it's 2026 or 2046?

I mean, obviously not literally every business. I think there are some things that are very anti-AI. The better AI gets, the more some businesses that have nothing to do with AI will be harder to compete with, because we'll really want these authentic, nontechnological experiences, or we'll care more about sports teams or whatever. So, no, not everything, but I think there will be many software businesses that are very much up for grabs.

Would you disagree on the timeline, then?

I disagree on the timeline. I think it's going to take a little bit longer.

Okay. Can you say more about that?

I love startups. I think startups are the coolest thing in the economy, and I've spent my career trying to really understand them. I thought when we got to GPT-4, which was back in 2023, that very quickly after that there was going to be much more disruption in software, with businesses being up for grabs right away, than it turned out to be.

I think I was wrong about a few things. One of them, in terms of the speed, is that the economy just has so much inertia. People keep doing the same things they're doing. They keep buying from the same companies. They keep wanting to use their tools in the same way.

I think this is actually a positive in many ways, and it's going to make this big transition in front of us go smoother and slower. I'm grateful for it. But I think it means we've all been too ambitious with our timelines. Even with this incredible technology—I think AI is one of the most incredible technologies humanity has ever invented—society and the economy will adapt more slowly.

Yeah. It's funny. We were talking before we started recording about all these parallels to history. Obviously, I read history for a living. When you were just talking, I wasn't even thinking about OpenAI, AI, and Sam Altman. I was thinking about reading this biography of Larry Ellison.

In the 1980s, he was just like, "Guys, this isn't a software problem. It's a people problem. We have to convince them. We can install software, but they're not using it. We have to change their behavior. The technology is there; we have to adapt humans so they actually start using the technology."

My own example of this was after Netflix came out and started shipping DVDs, even before it started streaming. It was amazing to me that people still went to Blockbuster. It was incredible to me. I drove by a Blockbuster on my way to and from school, and it was amazing to me that people still did it.

That is an example that has stuck in my head of the force of habit and the way people do things. Changing behavior is just much harder than the technology nerds realize. So, if we go back to this uproar of Tobi writing that open letter, or that letter to the people inside his company, you're adopting this faster than anybody else because you're partially inventing it, right?

Is there something where you're actually shocked at your own behavior? You know there's a better way to do this. You're even creating the product that could be better, and yet you still can't get over this force of habit?

100%.

Okay.

I love that you asked me this before. I have been waiting for this question. The thing that feels most psychologically inconsistent about myself is that, for 20 years, I have been using computers the same way.

I now have a magic thing called Codex. So do you. So does everybody. That means I should completely be using my computer in a different way. I should not be clicking around, pasting from one messaging app to another. I should not be scrolling mindlessly through my emails and trying to figure out which one is least painful for me to open and respond to when I don't want to be dealing with it.

I should not be keeping a to-do list and doing these rote computer tasks in the same way that I have for so long. And yet there is something in my mind that is encoded to believe that doing this kind of stuff is what it means to work and what it means to be productive.

If you asked me, I would never say I like doing it that way. In fact, I would say the opposite, and I think I would mean it. But by revealed preference, I have a better way to do it now. I can do it faster. I can be using Codex for more of my day-to-day work—getting through this stack of emails, doing the stuff on my to-do list, and dealing with all these things—and I still do it that way. It makes no sense, other than that I must secretly like it or feel good about it.

What do you think is going to have to change for you to actually adopt your own product in a deeper way?

I don't really know. It's happening gradually, and this might be the right answer: These things have to happen gradually, because totally changing someone's ingrained habits and workflows is difficult.

I think there are better products we can build with this technology that will make it more seamless to do that. But right now it feels like we're all straddling these 2 worlds. We still have a computer that we can use the old way, and we have Codex, which can use our computer in this amazing new way, and we're not sure which to use when or for what.

I think this is mostly a product failure. The phase that we're in now reminds me of smartphones before the iPhone. I was an early adopter. I had a Palm Treo in 2003 or 2004 or whatever.

A lot of the technology was there. It was missing multitouch, but mostly it was missing the product ideas that made the iPhone the iPhone. I feel like we are now in a world where we have all of the technological pieces, but we have not had the iPhone moment of completely changing how someone interfaces with technology.

We were talking about Tobi. He's out here building these things himself, right? We have a mutual friend in Josh Kushner. He says there's a big comparison to be made between the way Steve Jobs thought and the way you run your company.

Steve Jobs obviously wasn't the one writing the code. He wasn't building the hardware, but he was like, "I am patient zero."

I am making products that I myself want to use. Essentially, everything that we saw with Apple was basically what he wanted. There’s this great story in one of the books where they were supposed to have a meeting on, I think, one of the new MacBook laptops.

The team prepares this huge presentation for Steve, and they’re really nervous because of his commanding presence. He walks in, and they think it’s going to be an hour-long meeting. They show him the laptop, and he’s like, “On, off.” He presses the button, it comes on, and he presses off immediately. Then he tries to open the MacBook, and there’s a delay. He goes, “Make this”—meaning the MacBook—“like that,” and then walks out of the room. That’s the whole meeting. There are a lot of examples in the history of Apple like that.

How do you approach it? How do you improve the product? Are you just doing it through your own needs? How do you think about this?

Most of my effort right now is on research and compute. I would love to be able to spend more time on product. We have great people thinking about the product here, but the most important thing that we can do is create smart models and be able to run them efficiently and abundantly for a lot of people. If we can get that right, I believe that everything else will follow.

Philosophically, I’m very inclined to say: try to find the high-leverage, difficult problem that will continue the exponential. For us, this is models and compute. I also just think those are problems that naturally suit me.

Why do they naturally suit you?

To scale compute in the way that we’re doing this requires a complex supply chain. There are a lot of interesting partnerships to figure out, which I like doing. There are interesting financial challenges of how you’re going to finance what is probably already, or at least rapidly becoming, the most expensive infrastructure project in history.

The technology questions that go into building out compute at this scale—from designing your own chip to the supply chain of fabs and people that make racks, to the power systems for these things—I’ve always been interested in energy. They all come together. So there are a lot of problems that are interesting across technology, business, policy, supply chain, and logistics, altogether around building compute at this kind of scale.

So I used to be a startup investor, and the thing in my career that I have found closest to startup investing is managing a research program. There are all these ways in which they’re really different, too. The average researcher and the average founder, I think, on the surface look different for obvious reasons, but there are a lot of similarities about how you find the nonconsensus bets, how you decide where to have conviction, how you understand what exponential growth looks like, how you manage outlier talent, and how you identify it.

This is the research building that we’re in, and it’s where I sit.

Great. Say more about the parallels between what you learned in startup investing and doing research.

One big one is the power law. People talk about this all the time in investing, which is: you have to reprogram your brain, because we don’t seem naturally well-suited to think this way. Your best investment will outperform all of your other investments put together. Your second-best investment will outperform everything else put together after that. AI research, at least, is like that as well.

When we started, people thought it was totally unlikely or almost impossible that AGI was possible.

What year is this?

2015. 2015. I mean, we just got hammered by all of the intellectual giants of the field for saying that we were going after AGI. And then, when we started really focusing on large language models, we got hammered again, with people saying, “This is completely ridiculous.”

I understood, at least from my startup background—and I think other people understood it in other ways—that high-risk bets are okay as long as you take the ones where, if they work, they’re super valuable. Research looks this way, too. The kind of people that make great researchers are sort of nonconsensus, fresh-approach, high-energy, sort of nonstandard. “Nonstandard” is the word that keeps coming to mind.

You’ve got to say more about nonstandard. Can you be more specific? Are they spiky?

You don’t want to fund a founder who has a very slightly different take on the same idea that the last 1,000 people you talked to have tried to convince you—and maybe convince themselves—that somehow they’re completely different and doing something totally new. But it’s mostly trying to fit in with the herd, be on the same track as everybody else, and do what they’re supposed to do, which is start a startup. They’ve heard Peter Thiel say enough times that there’s something they’re supposed to be doing differently, so they kind of try to emulate that, but they don’t really mean it.

It’s very clear to me when you have someone who just thinks differently than most other people and is willing to stand by convictions that are very unpopular, may well be wrong, but if right, they’re at least going to be really right, versus someone who is like a thin veneer on the same idea that everybody else has.

In late 2015, when we were starting OpenAI, there were very few AGI efforts in the world. There was DeepMind, and 1 or 2 others that I can think of. It was a very nonconsensus thing to do. In that same year, there were probably many, many thousands of founders starting photo-sharing apps. I’ll pick on it just because it came to mind, but there are other categories, too. That was probably not as good a thing to do.

Today, a lot of people want to start AI labs. There are a handful of people—2, 3, whatever—doing something completely new. That actually wasn’t possible until AI got this good, but it doesn’t seem like a good idea yet. That is the thing that, as a startup investor, I always wanted to fund, and the thing that mostly worked for me.

There’s a similar thing for researchers. There were a lot of researchers that would chase whatever the last thing was that worked, and there were a small number of researchers that had high conviction toward a new idea. I think we were and are the best research lab for those people.

I remember talking to Dario about this a few months ago, and he thought, in terms of chasing after the way that you guys are, there are the big 3 players that the money—the capital—requires. There’s not going to be a 4th bigger player. But he was like, there’s this 10%—I forgot what the number was. I’ll just make it up. Say 10% chance that there’s just some monk researcher who’s going to approach it in a way, from an angle we’ve never even considered.

Totally. I don’t know how to put a number on it, but there is some chance. I don’t think I’m making the number up, but it was like a small percentage.

There is some chance of that for sure, and I love that. I think that’s why stuff stays exciting.

I want to tell you about the presenting sponsor of this podcast, RAMP. I have been reading a lot about SpaceX lately. SpaceX is one of the most valuable businesses in the world, and one of the main themes in the history of SpaceX is constantly attacking and questioning your cost. RAMP helps many of the most innovative businesses in the world do exactly that. The median company running on ramp cuts their expenses by 5%. And one thing SpaceX has demonstrated is that a religious dedication to controlling costs can help actually increase revenue because you can pursue opportunities you couldn't otherwise. And we see that in the RAMP data, too. The median company running on RAMP also grows their revenue by 16%. So, when you're running your business on RAMP and your competitors are not, you have a massive competitive advantage that compounds over time. Ramp is the only platform designed to make your finance team faster and happier. Many of the top founders and CEOs I know run their business on RAMP. I run my business on RAMP and you should too. Go to ramp.com to learn how they can help your business save time, save money, and grow revenue. That is ramp.com. I found one of my all-time favorite quotes when I was reading the book 0 to1. The quote says, "The single most powerful pattern I have noticed is that successful people find value [music] in unexpected places. And they do this by thinking about business from first principles instead of formulas. That is exactly what apploven [music] has done with their advertising platform. Apploven connects you with over a billion potential new customers [music] inside mobile games. Apploven allows you to capture undivided attention. Apploven ads are full screen video ads that are watched for an average of 35 seconds. That is retention that blows other ad platforms out of the water. And you can launch on Apploven in minutes. You set the goal and Apploven achieves it. There's no complex setup, no expertise needed, and Apploven scales quickly. They can put your ads in front of over a billion potential customers. Other businesses have seen immediate results, have [music] scaled to hundreds of thousands of dollars of spend per day and increase their revenue by millions. So, you want to get started quickly before all of your competitors are on Apploven. And you can do that by going to apploven.com. That's apploven.com.

So, what is confusing to me? You went from founder to investor to back to founder. But why, in 2015, what got you interested in artificial intelligence to begin with that you’re saying, “Hey, this is such a nonconsensus thing”?

People think I'm fucking crazy. I'm going to do it anyway.

Well, I had been interested in AI my whole life. I was a very nerdy kid. I was the kind of kid who spent Friday nights playing on my computer, watching science fiction, and reading science fiction. I always thought that AI would be the most amazing, craziest thing. I never thought I would actually get to work on it, but I always loved it.

I even came to college to study it. I worked in the AI lab the summer between my freshman and sophomore year, and nothing was working. In fact, very memorably, a professor told me, “You could try all of these things. There are all these directions. The one thing we know doesn’t work is deep learning. We tried that for a long time. It’s the most guaranteed way to have a bad career.”

I was an impressionable freshman in college, so I assumed that was true. I pursued these other things. It was clear to me at the time—this was around 2005—that AI was not working. I happened to accidentally get into startups, but then I very much fell in love with it.

I wouldn’t even call it a career detour, because it was super helpful. Looking back, becoming a startup investor was great. The normal career path in Silicon Valley—or a common path—is that you’re a founder, then you semi-retire and become an investor.

We don’t want that. I heard you say that you’re going to work on this for the rest of your career.

Yeah, hopefully.

We’re going to come on the show multiple times. I’m going to hold you to this. We don’t need more founders that retire and invest. We have—

I don’t think we have too many investors in this. But what I was going to say is, the fact that I got to go in the other direction—I was an investor first and then I ran a company—is pretty unusual.

Very unusual.

And I’m super grateful for it, because you get this unbelievable set of learnings and pattern matching if you really study and watch companies as an investor. That has been super helpful to me running OpenAI, but it’s the opposite of the normal direction, so it’s a very rare thing. I strongly recommend it.

Why is it helpful? If you’re running a company, you have faced some number of similar decisions in your past—some number of crux decisions—and you’ve seen what works and what doesn’t. If you have to make a high-stakes strategy shift or fire an executive in a really messy way, you have whatever your own limited previous experience was over the last 5 or 10 years that you’ve been doing it.

But as an investor, you watch all the crux moments. You don’t get the operating practice that you do just day in and day out running a company, but you’ve seen a lot of the big crux moments. You see those all day long. The wealth of the data set that I had there was awesome.

So that plays in your head when you have a decision to make?

Yeah. I’m like, “Oh, this is what happened when this company had a similar thing,” or, “I saw this founder make this mistake,” or, “This founder got it really right.”

We were talking about how you studied the Industrial Revolution. We were talking about some great biographies that we both read earlier. My friend Daniel Gross says this about me. I think the benefit of doing this project on my other podcast, Founders, for 10 years is that he says, “You’re like an LLM trained on history’s greatest entrepreneurs, but with the temperature turned up because you’re fucking crazy.”

I’m super passionate about it in a weird way. It’s strange to be obsessed with dead entrepreneurs, but it is helpful. I’ll be talking to a founder and they’ll talk about something they dealt with, and I’m like, “Carnegie did this, and Rockefeller did this. You might want to try this.”

Do you find that you have one big insight out of all of that, or is it just that for any given scenario, you have what all these people did and how it comes together?

I think it’s dependent on the personality of the founder, right? I’ve read your blog for years. I think you’re a great writer. You’re very succinct; the brevity is really appealing to me, and I love numbered lists.

It’s weird that we both write in the same way. I feel like I’m reading your blog and thinking, “This is the exact conclusion that I would come to based on all the reading.” There’s this handful of principles that could be applied, but it really depends on who the founder is and what they want to do.

This is what I’m trying to understand. Let’s go back to what we were talking about. You’re making a big jump, because you’re one of—from what I hear—one of the best investors of all time in Silicon Valley. You could just be rich and not really have to work, because investors are kind of lazy.

I’m just kidding, by the way. Kind of not. Having done both, I think I can say it’s much, much, much harder to run a company than being an investor.

Exactly. And that’s what people should be doing, in my opinion. So then you’re like, “Fuck that. I’m not going to take the easy route. I’m going to do the hardest thing ever.” The thing that people think is impossible. The thing I’m going to be made fun of.

Yeah.

I still need to understand. You were an interested kid, so why would AI appeal to a kid? You were living in St. Louis at the time.

I was living in St. Louis.

Why would AI appeal to you back then?

I think it appealed to every kind of computer nerd. I don’t think there was anything that unusual about me. It just felt impossible. Most people would say, “Of course, that would be the coolest thing ever,” but it’s totally impossible.

The weird thing about me was, “Okay, let’s try.” I think everybody would think it’s awesome and something to go for.

So wait, was that a personality trait of yours as a kid—that when someone told you that you couldn’t do something, your initial response was resistance?

Not resistance, but, “Are you sure? Why not? Let’s try. Let’s see what happens. Maybe I can, maybe we can.” I was a very optimistic kid. The more something seemed impossible, the more intrigued I was.

The idea that we could invent a technology that would let us do everything else, that would empower people in a way that no other single technology could, always seemed innately incredibly appealing to me. I wanted that thing. I wanted to be able to do everything else.

I think another thing that was a personality trait as long as I can remember is that it’s interesting to really give people a lot more power and a lot more ability. In some sense, this is the whole arc of technology. I was always a technology nerd, but AI is the strongest version of that I can imagine.

What did you think it would enable back then? When you were a kid, you thought, “This seems like a cool technology. I want to do X. I can’t do X unless AI is invented.”

It’s always hard to remember how much of this is what I actually thought at the time versus what you’re trying to build right now—how much my current work has colored my memories of it.

For sure, as a kid I was very into robots. We had a robotics club in my school. The robots at the time were laughably primitive. I even remember that at summer camp we had this little turtle that you could control with a computer. It moved on the floor or on the table, and I thought that was the coolest thing.

There was something about physical stuff moving, controlled by a computer, that I always thought was amazing. Now I’m extremely interested in what AI can do to advance scientific discovery. In my memory as an adult, I think I thought that was cool as a kid, too, but it feels implausible, so I assume that’s an example of where my memories have gotten more colored.

The fact that we can have AI discover new physics and cure diseases, and what it’s already doing for math—I think this will be one of the most important areas, even more important than automation of other tasks that AI can do, just to help us understand more things.

We were talking earlier about The Beginning of Infinity. Rereading that book from today’s vantage point, I’m like, “Man, AI is really going to help us do this important thing of understanding everything, or as much as we can.”

I was definitely interested in the Star Trek version of huge prosperity and abundance, and what AI could do to drive that. Maybe the memory of being interested in science is more real. I also loved science and this idea that, because we were smart, we could figure out how to understand the world, make predictions, and do things that we couldn’t without this deep understanding. I don’t know—that seems innately awesome.

It’s interesting how consistent over time what humans want from AI is, because something you’re describing is very similar. I just reread the biography of Claude Shannon for the second time, and I had forgotten—because I hadn’t read the book in maybe 5 years—that Claude Shannon and Alan Turing used to meet every day for coffee when they were both at Bell Labs.

This was in the 1940s, and they would just talk about AI. They were both obsessed with it. They thought it was inevitable back then, and they thought it was going to happen 15 years from then.

So, like, by 1955, they thought we were going to have computers—which didn’t exist, right? They had the analog versions that were going to be smarter than humans. Anybody who thought that wasn’t going to occur, they thought was absolutely ridiculous. And they were like, “Well, what would you want the computer to do?” He was like, “Solve math problems, write poetry, cure diseases.” You hear this over and over again.

I have read a bunch of things that those guys wrote at the time, and I am so sad they are not here to see it because they were so right about everything. We’re finally at the moment where AI is solving novel math problems. It is discovering other stuff. You can argue about how good it is or not—I would say not very good—but it is writing poetry. It gets to what these guys, I think, would have said: “All right, you’ve done it. This is it. We’ve got it.” And that would have been so cool.

Yeah. This is the weird thing where everybody’s just like, “Oh, it’ll never do X.” I talk to people in the music industry, and they’re like, “It’s never going to make great music.” And then they’re like, “Well, do you think it’s going to make a podcast?” I’m like, “Of course it’s going to. It’s going to do everything that we can do, at least—I would say better than—even right now, better than what we can do.”

It’s a very bizarre thing where it’s like, it will never surpass what is happening at this current point that I happen to be alive. There’s a deep human psychological flaw there. But here is, I think, a more interesting question: Let’s say it does make a great podcast. Two AIs are having a more interesting conversation than you and I are. Do you think people will care, or will they want the one with the real people because we’re all obsessed with people?

Yeah, for me, this is more interesting. It’s like, “Oh, these 2 people that I may be predisposed to like or dislike are having a conversation that’s interesting to me.” I think for, like, strict reference—maybe my other podcast, where I’m just saying, “Hey, these are some interesting ideas I read in this book”—that could maybe get disrupted, whatever the case is.

But especially for people who were born before this happened, maybe it’s different for your son, but for me, I think humans are always going to be drawn to humans. I really deeply believe that. I think there are a lot of other jobs that could face significant transition, but stuff that’s about people, stuff that’s about people’s connection, connection to people, and people liking other people—that stuff feels like it gets more valuable in the post-AI world, not less.

I may actually be the wrong person to talk about this, though, because I deeply desire, even though my entire work is digital and broadcast all over the world, more of an analog life. I like reading physical books. When I was talking to Kelly, I was like, “I don’t want to get on a Zoom call. Call me, or we’ll talk in person.” I like physical stuff.

I’m like that, too. I don’t read e-books.

Yeah.

I don’t like Zoom meetings. I like to be with people in the real world. I definitely think there’s a subset of weirdos—and there are probably a lot of them who live in this city—who don’t like humans and only want to communicate with computers. But it’s like, I think that’s a tiny percentage of humanity. I think it’s a tiny percentage of humanity.

This is why I think the world is, on the whole, not going to be that different, even with superintelligence. People are still going to be very fundamentally wired to care about other people, to want to be around other people, and to interact with other people. There will be some people who just get obsessed with the models and think humans are in the way, or a danger to be contended with, or whatever, and for most people it’ll be the whole point.

I think it’s very important, and when we do find people like that, they need to be called out and we need to make sure they don’t acquire power.

I certainly agree with that. Maybe the 2 big risks that I’m most worried about with AI, which are a little bit in tension, are, 1, a loss of control, where AI somehow just becomes too powerful in a way that we can’t guarantee the control we want, and the other is power getting too centralized, where you have 1 company, model, or person with too much power.

In both of these, the fundamental thing is that I think it’s a very antihuman position for either of these things to happen. The right approach is to say, “We want people deeply in control of the future. We want people deeply empowered. People are the whole point of all this.” We are not going to sit here and gradually hand over control to an AI model because we don’t trust or like people. It’s a very misanthropic thing to say we’re going to put all of our trust in this model and let it have all the power and decision-making over the world.

But I think there are some people in the world who think that’s the right outcome. There’s another version of this, which is: Because we don’t trust people, we have to limit who gets access to this technology and how they can use it. All these terrible things could happen, and out of fear of those, we’re going to concentrate power in the hands of a few companies. We’re not going to let other people use this, but we’ll give them some benefits.

My caricature of this is that I think there are some people in the AI field who effectively say, “We’re going to give the world a cure for all disease, and we’re going to make stuff really cheap in exchange for people giving up their autonomy, impact over the future, and power.” And also, in the name of safety, just absolutely rampant inequality. There will be people who have access to huge amounts of wealth and power, and other people just get pretty good everything. This is a terrible sales pitch. This is a very antihuman sales pitch that somehow people feel willing to make.

Why do you think they feel willing to make that?

I think it’s fear and power. When people talk about the risks of AI, I think there are a lot of people who are so nervous about the magnitude of those risks and get so taken by that, and feel a need to protect the world from that, that they’re like, “We should trade off a lot of liberty for safety here because this is unlike other risks we’ve seen.” But then I think that also ends up being a way to justify a lot of power-seeking behavior.

Everything I’ve read—when I was just saying what Claude Shannon was saying, or Alan Turing, at least in the books that I’ve read—it’s more optimistic: We’re going to invent things that make our lives better and can do things for us.

You are totally right that if you go back to the Claude Shannon and Alan Turing era, they talked about how wonderful AGI would be and all the things that it would do. When we started, we really had a lot of pressure from the doomers.

The part of the doomers that I agree with is that this is a powerful technology, and we should err on the side of safety and act with caution at each level of technology. The part of the doomers that I don’t agree with is that it’s an unsolvable problem.

If you go back to the beginning of OpenAI, I think there would have been 2 widely held opinions. Number 1: Not at all, and certainly not in 10 years, were we going to build something that was very AGI-like. And, conditioned on us doing it, we certainly were not going to be able to make it safe.

If you had an AI that was smarter in many ways than a lot of the smartest people—most of the smartest people—then the doomers would say, “Surely, at that point, the world would have been destroyed. The alignment thing would have failed.” There were just these very confidently held positions about what would have happened a decade on.

A decade on, we have built something that I think most people would say, at the time, would have seemed very AGI-like. A lot of good things have happened, and the kind of crazy bad predictions of the world ending have not happened. So I think that should update people’s predictions about the future. There are still higher-stakes challenges in front of us to solve.

But our approach—this is another thing I learned from startups, the way you do things—is to put things out into the world, get feedback from real customers, see where they break, and see where they don’t break. That is the way you make a good product. That is also the way you make a safe product. And we have made way more progress on AI safety than I think most people thought we would when we started.

Why? Because there are 1 billion people using our products on a weekly basis. Each time we get a new level of model, we put it out in the world and see what works, what doesn’t work, where people need us to relax the guardrails because they have good things they want to use it for, where we have alignment failures, and where we have safety-systems failures.

ChatGPT has only been out for less than 4 years, 1 billion people use it, and they use it for sensitive, important stuff. The fact that we can deliver something that is broadly considered safe—of course, there are issues with it—in that short a time frame with such a powerful technology, I think there is no way we could have done that in an ivory tower.

This is how I believe you build good, safe, robust, useful technology and products. I think it’s a great learning from Y Combinator. It would have seemed to most of the AI safety people totally impossible to get to this stage and still have the level of safety guarantees we have now. I do think it gets harder from here, but I don’t think you’re going to solve it by disconnecting yourself from reality.

Why does it get harder from here?

Because we're about as smart as the smartest people in the world, and we're about as smart as the smartest models in the world. And that's going to flip right now.

Yeah, in that direction. I think that's right. I think the models are just so incredibly capable and improving on such a steep trajectory that the unknown unknowns—maybe they don't get harder relatively, but from an absolute perspective, they seem harder.

I think we'll have to make a bunch of difficult decisions about when we delay development, when we sort of say, “Okay, you know what? Let's connect with reality now,” or, “Let's wait longer to really study this more.”

I was talking to someone recently, and something that stuck in my mind is that the FAA has helped make flying incredibly safe. Flying on the surface seems like this extremely dangerous thing, and you probably get on an airplane without giving it much thought. This was certainly not the case at the beginning of aviation. Airplanes are not that old in the long trajectory of human history.

They have extremely robust accident reporting and are extremely clear-eyed. They never try to hand-wave over something. They want to extract as much information as possible. In some sense, I think with any new technology, an approach like that works very well and is often underappreciated.

When we started deploying our models, when we said we were going to put ChatGPT in the world, we knew the model was imperfect. We knew it hallucinated, and we knew it could do these other things, but we also knew that the world had to experience this technology. We've got to learn how to make it safe, and we've got to put the power in people's hands.

We cannot just use this to impose our worldview. We cannot use this to go sit in a lab and try to think through all the impacts, which won't work anyway, because society and the models are going to co-evolve. We have to all do this together as this joint product.

We'll do very good accident reporting. We will study when something goes wrong. We will put out a very clear postmortem. We will learn as much as we can. We will not only improve our own technology and products, but we'll try to share those learnings with other people building AI. I think that's worked surprisingly well so far.

Deal is how the best founders turn the world into their talent pool. I've been studying how history's greatest founders operate for a decade. And one thing they all have in common is they understand that recruiting and hiring the very best talent is your most important priority. A players recognize other A players which is why top companies like [music] ramp, Shopify, 11 Labs, Uber and Door Dash all use deal. Many of the top founders I know have personally invested in Deal after using their product and what [music] they discovered is that deal is the best company in the world at building infrastructure for global hiring. Deal will help your business hire, pay, and manage any worker [music] anywhere in the world. So you can retain the best talent anywhere and spend the rest of your time focusing on what you do best, delivering value to your customers. [music] The founder of 11 Labs has a great description of the value deal can give your company. He said, "We built 11 Labs [music] to break down language and communication barriers. With deal enabling us to hire and support [music] exceptional talent anywhere, we can accelerate our innovation and bring more voices, stories, and ideas to [music] every corner of the world. Deal is trusted by over 40,000 businesses. Learn how they can help your business today by going to deal.com/enra." that is deal.comra.

One thing that has to be disorienting for you, because it doesn't make any sense to me, is that essentially what I focus on is just entrepreneurs and entrepreneurship. All the podcasts I make are for the benefit of entrepreneurs. I'm glad other people listen, but they're heavily focused on trying to serve entrepreneurs and trying to find useful information, whether it's in a biography of a dead entrepreneur or talking to somebody like you who other entrepreneurs can benefit from hearing, or any of the podcasts that I make.

So, in that, there are not a lot of entrepreneurs who love AI. What I'm trying to figure out, and maybe you can help me reconcile this, is: everybody uses AI, everybody hates AI. What the hell is going on there?

Well, people are always afraid of rapid socioeconomic change. We talked about the Industrial Revolution earlier. I love reading about previous technological revolutions, too, and people did not have universally warm and fuzzy feelings about the change that was happening throughout the Industrial Revolution.

I think it's probably a good feature of human society that we have some built-in inertia and some skepticism of rapid change. I think that probably helps society in times of turmoil or in times of localized craziness or whatever. Some of it is probably good, and I think it's a feature of human biology that I believe in never trying to fight too hard.

I also think a lot of the people building AI have been off saying, “There's a 25% chance we're going to destroy the world, and, yeah, we're going to race ahead to do it because otherwise those bad guys will do it first.” Or, you know, it's like, “Man, this thing is going to be really terrible, and 50% of the jobs are going to go away in the next year, and we hope you all are okay, but it seems really scary.”

We have not, as a field, done a very good job of explaining to people what the benefits are and how the downsides can be mitigated. We certainly have not done a good job—even if people have had answers like, “Well, there's going to be universal basic income,” or “Work will be optional,” or whatever—of discussing how and why it's important that people have more power and personal freedom in the world, not less.

The ability of people to influence their own future and collectively design where society is going to go, and the autonomy that comes with that, is very important. I don't think a lot of people in the AI field spend much time thinking about, reflecting on, or acknowledging how important that is to other people. They feel it for themselves, but they don't think about it as much for everyone else.

To go back to that characterization of the sales pitch from earlier—I don't even know if that's a word—

It is. The “dear peasants, we will bequeath upon you these gifts of a cure for cancer and material wealth and great entertainment, and you stop complaining. We'll make all the decisions about the future, and just trust us. We'll be benevolent dictators” thing.

Not good. Not good.

As a lover of entrepreneurs, and as a student of what has made this incredible economic miracle of recent centuries work—really empowering people to go do new stuff, push on the things they believe in, have the freedom to create companies and invent technology, and pursue ideas—there is nothing I believe in more strongly than the system around that that makes it happen.

Even if people don't see themselves as entrepreneurs and may never want to start a big company, they do understand how important that is. When you hear people implicitly or explicitly saying there's going to be less of that with AI because a small number of people are going to have the power, but they're going to make great decisions and keep everybody safe, I think that's very scary to them.

I also think that even if maybe most people don't want to start really big companies, a lot of people want to start smaller companies, and that has been hard. That has required a fair amount of privilege, luck, and resources to be able to do. We are about to see the greatest boom in people starting smaller businesses that we have ever seen. I think AI is empowering that.

For some reason, the field, including us, has not talked about that enough, even though we see all these signs of it. It's great, and we have not built enough products to accelerate that, but I think we're going to see a lot more of it.

It's funny—we talked to Tobi Lütke earlier at the end of the conversation, and I think that's in the episode. He mentioned something I hadn't thought about. He's like, “Oh, yeah, you and me are in the same business.” He's like, “We're both trying to create more entrepreneurs. He's building infrastructure for entrepreneurs, and I'm building educational and inspirational podcasts for them.”

The crazy thing with what you just said is that, out of any new industry, the AI industry is doing the worst job I've probably ever seen at explaining itself. Part of it is just the ability you guys have to get out there, talk about the stuff that you're seeing, and educate people.

There's this great book called The Intel Trinity, and it tells the story of Intel. It's called Trinity because the 3 main players are Bob Noyce, Andy Grove, and Gordon Moore. There's a great story in the book that I never forgot. They go from inventing, I think, the integrated circuit to the microprocessor, and they realize that the technology is so important that it would scare their potential customers.

So, the 3 of them stopped doing what they were doing and went out and started educating potential customers, investors, and the entire country. At one time, they were putting on more classes than the local community college had in its entire course catalog.

That's how much they made it a top priority: “We're going to get out and educate people about this new technology.” Why isn't anybody in AI doing that?

I mean, no excuses. We should be doing more. I think we've tried versions of this. We haven't gotten it quite right.

You're doing it right now. This is one of the reasons I wanted to talk to you, because I use AI all the time. I think it's fascinating, but you have such a view that mine looks like the view of an ant. There's so much stuff in your head that I want to get out. You have all this context, and you're inventing this incredible technology. I would love to know how other people are using it.

Actually, before we even get to other people, I heard you say something interesting that ties into what you just said about inventing technology. The technology is increasing rapidly, but adoption should be slow and deliberate. I heard you on another podcast saying, “Hey, I'm even considering how much I should let AI see—every single thing that's on my computer.” You want to talk about that?

With the latest generation of models, I don't want to say they feel smart enough, because I think we should always aspire for them to get smarter, but they are pretty smart. I feel more limited at this point by the amount of useful context AI has on me. I want the AI to know as much as it can to help me. I want it to be doing things I can't or don't want to do on my own.

I'm not going to read every post on our internal Slack. I'm not going to read every story a customer has to tell about where ChatGPT worked for them or failed them. I can't. I probably could read more research papers than I do, but it takes a lot of mental energy. I would love to have an AI agent that is constantly trying to be helpful to me, that can look at and understand more context than I can—or that I have time or energy to process on my own—and can help bring that context to bear and give me good advice when I have to make a decision.

I think we've focused correctly so much on model intelligence that, on the product side, we have not yet thought enough about what it means to give a model more context than any person could have and help advise that person on their big decisions.

My sense is that we are just on the precipice of being able to see a very different way of working with AI, on an axis where people just can't get this good. There are plenty of very smart people, but there is no one who can read tens of thousands of pages of context in a small number of seconds and really use that accurately. This is something AI can do that's going to be very new and an incredible supplement.

You've read all these biographies. There are probably times when you vaguely remember something that, if you could remember a specific anecdote from one of them, would really help an entrepreneur in one moment for that particular entrepreneur, right when you were talking to them. But maybe you forgot it, or maybe you don't remember it exactly right.

I built my own AI tool, so I use it internally. Do you know what it is? It's trained only on things since 2018. I've put every single note and highlight from every single book that I've ever used into this database, and I search it for that.

Then, when the work that you guys do came out, I added that. I have it trained on every note, every highlight, and all the transcripts for my episodes of Founders. I use this thing every single day to make every single episode.

I'm working on Claude—called Shannon. The Claude Shannon episode came out 2 or 3 weeks ago, whenever it was, and I'm asking questions about all of it. I was like, “Hey, what did Bob Noyce say about this? What did Rockefeller do about this?” I made the episode. I read the book. I took the notes. I don't remember because it was 7 years ago. It's incredible. This is what I mean. It's fucking awesome.

That is so cool.

Let's get into how you think about running the company. You're spending your time—you said your main focus is getting more compute and then research, right?

So you want the models to be the best in the world, but how do you think about whether you have to build your own products? You built Codex, right? I don't even know the fucking product lines. Where's all the revenue coming from?

Actually, I think we should be more of a platform company than a product company. We will build products, of course, but how many products do you have? Let's back up. How many products do you have now?

We just merged ChatGPT and Codex together. We used to have ChatGPT, Codex, and the API. Codex was sort of unfortunately named, and it was not just coding. It could do any kind of work, which confused people.

I'm confused by that.

Yes.

Okay.

As were many other people.

What I think most people want is a single interface to their own personal or their company's AGI that can help them with whatever they need, and then the ability, with an API, to build anything they want on top of it. That is the platform that we should offer to the world.

We're going to sell great AI at every point on the cost-performance curve. We will be the best. You want really high-end AI to discover science? That's great. You want really inexpensive AI to do a massive amount of work that maybe doesn't require genius-level intelligence? We've got you covered there too.

People talk about different ways of thinking about this—a new utility, a new commodity, whatever you want to call it. People want to use a lot of AI, and they want it at a low cost. They want it to be fast, work well, have their context, and be smooth. We've got you.

Then there's a single product, which is, “I need to ask the AI something.” Eventually, maybe the AI should proactively offer me things. But you will have this interface, which started as a chatbot and now also has coding agents. I think at some point it will feel like a more persistent agent—an AI that's running on whatever you need it to run on.

But that's it. I don't think we should go build every product category. I don't think we should try to compete with all our customers. I don't think we should try to subsume the entire economy. I think we should offer this platform and try to have 100 million new businesses and 8 billion people use it in all kinds of new ways.

One kind of direct interface to the product, one API for people to use it however they want. Those eventually come more and more together too. Then it's all about what people do with it, build on top of it, and whatever else.

What mistakes did you make to have to learn that? I feel like you've had to kill some good ideas and sacrifice going after the great with your full intensity and focus.

I think killing the good ideas—sacrificing the good ideas to go after the great ideas—is the hardest lesson for any entrepreneur or business to learn. It sucks to kill good ideas. No matter how much you think you're going to do it, people seem to do terribly at this. Maybe it's just the nature of who chooses to be an entrepreneur. I'm terrible at this. I know I'm bad at this.

Last year, for example, we killed Sora, which was a good product and fun and cool, but it used a lot of compute and wasn't as important as Codex, where we put the compute. We killed our web browser called Atlas, which, again, I think was a great product and was the best web browser, but it wasn't as important for us to focus on as somewhere else we could put that talent.

In a world of limited compute, limited people, and limited resources, we thought really hard and said, “You know what? General intelligence for knowledge work and eventually for science is the most important thing we can do.” Anything that goes into making that upstream of generating that intelligence—building our own chip, building our own data centers, writing good infrastructure software, and certainly training models—is really important.

But then let's just offer this AI as a service and get people to use it for intellectual pursuits, for work, for scientific discovery, and to be more productive in their personal lives. Let's have the kind of flexible, general platform and not do a lot of other things.

Anybody engaged in complicated work—and you've got to be at the top of the list of anybody alive right now—needs somebody to help organize their thoughts. It's extremely beneficial. You see this in every single biography. You see it throughout history: you need somebody to talk to.

There's actually a funny story about how extreme this can be. Charlie Munger had a thing called the orangutan theory. Have you ever heard of this? He said a relatively smart human could sit down with an orangutan, tell it all his problems, tell it everything on his mind, and then the orangutan obviously says nothing. The human leaves better off, just from the idea of being forced to put his thoughts into some kind of structure.

Now, obviously, with a very intelligent partner, Munger played this role for Buffett. Buffett's one of the most intelligent people ever to have lived—the greatest investor of all time—and still needed to organize his thoughts with somebody else.

You have almost a singular lived experience, especially for somebody as young as you are.

So I'm curious: Who plays this role in your life? Who do you go to who can even remotely empathize with what the hell you're dealing with on a day-to-day basis?

There are kind of 3 categories here. One, a lot of the researchers who have been here forever—we've all been through it together, and we've developed this set of shared language, intuition, standards, whatever you want to call it—that I have not been able to replicate with anybody outside of the company.

When it comes to the shape of what's happening and what might happen next, and where the technology is likely to go, in terms of questions of business and the world, for a long time in my career, Paul Graham and Peter Thiel have been 2 of the people I've learned the most from. In lots of different phases of my career, they're still the 2 people I go to if I have a very non-obvious problem that I'm stuck on.

After a lot of looking, I have not found anyone else who has the same kind of ability to think in a super nonlinear way. If what LLMs do is predict what word comes next, those are 2 of the people for whom I can least predict what word is going to come next. That is a super valuable skill.

You go with, "Oh, man, I feel really stuck. I've thought through all these options," and someone can tell you, "I think none of those options are good. Here's this thing that now seems totally obvious and correct that you didn't think of"—a completely different view that you haven't heard anywhere else.

Is this more of a prompt for your own thinking, as opposed to explicit advice, like, "Do X," for example?

It's often, "Here is a specific thing."

Really.

Yeah. So what would be an example that you could share from Peter? Peter's very fascinating to me.

He is very fascinating, and he's actually who I thought of. The reason I thought of this question just now is because you're like, "We had to kill these good ideas for the great. We're cutting Atlas. We're cutting compute. We have to focus." Focus is something that's very obvious when you listen to him talk about the importance of focus.

If you have something that's working, making it work better and going down this line—taking an hour away from that to explore something else—is too expensive. You should just go deeper on what's already working. There's a lot of value at the extremes.

After we launched ChatGPT, it was sort of this weird thing because people didn't really know what to use it for. It was growing super fast, but it felt very unstable, or almost like low-value growth. People were using it just because they were interested in talking to it and seeing what they could do.

There were a lot of people in the company who were like, "We've got to figure out something else. This is not sustainable value." I remember talking to him about this list of 5 or 6 other things that we could focus on instead. This was maybe 2 months after ChatGPT launched, something like that.

He was like, "It's an obvious mistake to do anything about this besides the fact that it's growing, which is rare and great." It wasn't growing as fast then as it started to grow afterward. He's like, "The power of this is the power of the Google text box. It's a text box you can type anything into, and it does the right thing."

The fact that it didn't match the current Silicon Valley wisdom—you had to have feeds, and you had to have a network effect, and you had to have a way that people were going to build up more context—was why everyone was worried. This was before we had memory. People were going to build up more context. People were going to get locked in. Were people going to have all the—" He's like, "All that stuff. People have just been chasing the Google business model for 20 years, and this is the first thing that's come up. Clearly, the empty text box worked for Google, so why don't you just double down on that?"

"It's growing. It's very flexible, and it has all of the signs, other than it doesn't fit the current Silicon Valley wisdom." I was like, "Okay," and so we went super hard on it, and it was great.

He's a simple genius, based on what he just said. Sometimes there's more complexity, but that was an example of very important, simple genius.

What about some advice from Paul Graham, or some guidance or direction he pushed you in?

The thing you just said—this is a meme for many YC founders, where you would go see him for office hours and he would say, "You know what you should do?" And he would shake his finger like this: "You know what you should do? You know what you should do?"

Sometimes the thing that came after that was great. Sometimes the thing that came after that was terrible. But the important thing was that there was a kind of creativity, an open landscape, and just a, "Let's try a lot of things."

We talked about the spirit of iterative deployment, and we talked about how, in the same way as startups, he really pushed the startup ecosystem into this world of, "You've got to ship a V1. You're embarrassingly early, and it doesn't matter if it could be much better. You'll get it much, much better because of the feedback from customers."

I don't even think I asked him before we launched, "Hey, do you think we should launch this thing?" But I knew what he would say. I knew it was still early. I knew it was still embarrassing. And I knew the right thing was to get it out and get it in front of people.

So, wait. Your mental model of Paul Graham is so complete that you don't even have this one case?

That's the one where you would say there's certainty.

Let me tell you something funny. A few months before he died, I went to Charlie Munger's house and had dinner with him. I was like, "How often do you talk to Buffett?" He goes, "Never." I go, "What?" He goes, "We talked every day for hours and hours. Buffett can just pretend to pick up the phone to call me, and he already knows what I'm going to say."

That obviously comes after 65 years of working closely together, but I thought it was hilarious.

That is hilarious. That is really a funny story.

No, there are many times that I couldn't predict what he was going to say, which is why I think it's valuable. But in terms of launching when you're embarrassed of a product, I know what he's going to say there. That has been—I won't say the most valuable piece of tactical YC advice, but it's been up there.

I'm astonished, looking back at all of my data points from YC founders over the years, how much the ability to move fast and be iterative correlates with success.

Okay, you've mentioned YC way too many times in this conversation. I have to explore this because we talked before: Listen, I'm not a journalist. I'm an enthusiast. I don't have a list of questions. I'm like, "I have a world-class founder across from me. I want to know what the hell is in this person's mind, and I want to extract information out selfishly for me."

Why? I'm shocked at how much you reference it in conversations, and how impactful going through YC and then running YC—being affiliated with it—has been in your life. Can you expound on this?

There is some band that wasn't that successful. They didn't sell that many albums, but they influenced all of the musicians that came after.

I think it's called The Band.

Rick Rubin literally told this story. I think it might be the Velvet Underground, but you know the idea I'm getting at.

It's not fair to talk about YC in this way, because YC, measured by traditional metrics and market cap created or whatever, would be one of the handful of most valuable tech companies. But the degree to which YC totally influenced everything that has happened in the last 20 years of the tech industry and startups and entrepreneurship, whatever—I think that's only sort of understood.

OpenAI is an example of that, not just from how we have shipped our products in the world, but from the philosophy of how we run our research lab. I think if you go talk to many of the other people running this generation of large tech companies, they would tell you similar stories, even if they didn't go through YC.

But what's happening there? Is it an operating system that YC is giving you? Because you hear, you know, "Do these 5 things," or whatever. Or is it more like a philosophy of building companies? This is the confusing part for me as an outsider.

I think it's 2 major things. There is some of the operating system of what to do, but I think it was the philosophy of how to run companies: the idea of iterative deployment, technical people in charge, and being willing to bet on young people with a lot of energy and ambition but maybe less experience throughout all levels of the company.

Then there was also the related change to the whole ecosystem that happened in the pre-YC tech ecosystem. If we ran the clock back to 2004 and then projected technology forward to 2016, but not anything else about the shape of the startup ecosystem—what it meant to be an entrepreneur, how capital flowed, who got to run companies, all of those things—I do not think OpenAI would have been possible.

I think the changes that YC induced in the whole ecosystem—more leverage going to founders, young technical founders having the ability to raise lots of capital, and the ability to work on ambitious things without a very proven résumé—made it possible.

So, this is kind of a big change. Is this all tied to the fact that you think there was a benefit in you going from founder to investor for a long period of time, back to founder?

There are all those benefits too, and I wouldn’t say I really went from founder to investor to founder, because the first time I was a founder didn’t really work out that well. It’s like—

A company—you learned some lessons from failure, but I think you learned way more from success. Oh, hold on. We’re not moving on from that. You’ve got to say more about that.

There’s some—I can’t believe it. I think it’s in some great Russian novel. I’m very embarrassed not to know this. It starts with, “All unhappy families are unhappy in their own way. All happy families are the same.”

Yeah.

That explains why it jumped into my mind. But I think this is really true. When I look at the lessons of where I have failed at something, I learned something generic about grit and determination and something not to do. But most things don’t work, so there are a lot of reasons why things don’t work, and I think it’s harder to put together the correct causation.

When I’ve had something really work, when I understand what parts of Y Combinator really worked or what parts of OpenAI really worked, trying to apply those lessons going forward has been much more helpful to me than trying to apply the anti-lessons of what didn’t work. You should, of course, learn as much as you can from every data point. Learn from the failures, learn from the successes.

But in my own experience, when I have tried to apply those lessons, the lessons I learned from success were very good, and I should have applied those more. The lessons I learned from failure were either fairly generic and I kind of already knew them, or got in the way of something else. And I think this is generally true for a lot of people.

Yeah. But isn’t it like we already kind of know what we should do or what we should avoid, but it’s the reminder—the constant reminder? The best description of my other podcast, Founders, I ever heard is that it’s church for entrepreneurs. If you really think about it, I used to drop it on Sundays, and I should go back to doing that.

I’m really just telling the same story. The same personality type has appeared throughout history. It’s just that now this person happens to be building ships, and this person built technology, but they live in different times. It’s the same personality. That’s for sure.

I’m kind of obsessed with this idea of things that last for a long period of time. Companies—the best companies can last a long time, but not as long as cities. Cities and countries don’t last as long as religions. Out of all the man-made things, what has lasted longer? I can’t think of anything other than religion.

I grew up—my mom was a fundamentalist Christian, so I was forced to go to church my entire life. I started analyzing what all the main religions in the world have in common. We have a shared base of knowledge, usually some kind of book, right? We meet with like-minded fellow believers at regular intervals.

It’s not like I go to church on Sunday and we say, “Okay, we talked about Jesus last week, but let’s talk about this other guy.” No, we go back to the same books and the same stories over and over again. So I read your blog, and you even said something about when YC ended. You were repeating the same thing. You were telling it to them all the time, and then they leave the church, to continue the analogy, and they stop doing the same stuff.

It’s not even the lessons. It’s the constant reminder that this is important.

I extremely strongly agree with that. But I think it is better to be reminded of the things like, “Talk to your users more,” “Ship products earlier,” “Get more feedback,” “Hold a higher bar for who you recruit and who you hire,” and “Move more quickly.” But it’s the positives that I think are good.

You have one of the greatest tweets. I save it to my phone. You’re like, “Skip the conferences, the [expletive] dinners, everything else. Just essentially make the product and sell the product. If you’re not making it, you’re not selling it.” That’s all you actually have to do. And I think it goes back to that simple genius.

Somebody asked me yesterday, “What’s your historical equivalent for Sam?” There’s usually some kind of historical equivalent for every founder I meet. I can say, “That guy’s kind of like Vanderbilt. That guy’s like Rockefeller, Carnegie, or any of these people.” And I was like, “What’s your historical equivalent for Sam?” I was like, “There isn’t one. I can’t think of one, because I don’t know him well enough. I don’t understand how he thinks yet.”

What do you think now?

Well, this is the first of hopefully 8 conversations. So I’ll tell you on conversation 7, but this is very rare. I just talked to Doug Leone, and he talked about 1 guy that he hired who was an associate in venture capital and then left and founded Nubank, one of the most successful companies. I’m like, “I’ve never heard of that. Doug, have you?” And he dedicated his life to this. He goes, “No, that’s the only one.”

So again, very rare. Mostly, people go from founder, sell their business unfortunately, and then become an investor, as opposed to running the business till you die, which is my preferred method of things.

What I’m curious about is this: when you just said you learned more from successes, right? Was it the successes because you were exposed to 10,000 different companies in that decade or decade and a half that you were doing this, and you saw obviously maybe the half-dozen or the dozen that were the best in the world? Are you taking their successes as instructive as well?

No, no, totally. I think people do. I mean, you’re an incredible student, but there are a lot of pretty good students of entrepreneurship. People often try to look for the lessons from the things that really worked, and as you said, it’s kind of the same thing over and over again, just done in different industries. But you have to be reminded of it a lot, and it’s unglamorous.

I had no understanding going into this conversation. I think I have a slightly better understanding going into this. I told you before we started, this is just for my own edification, but even the people who influenced you—it’s just like Peter Thiel saying, “No, dummy. Obviously, this is working. Why are you doing anything else but the thing that is working?”

There have to be examples where you’re like, “Hey, I’ve given this advice to other founders a million times,” and then you catch yourself: “Oh, shit, I’m not even applying my own advice at this point in time.”

Totally, yeah. I’ll give many examples of that. I think it’s also instructive to ask: What was new? What didn’t you have advice for?

The thing that was really different about OpenAI than anything that I had pattern-matched before is that it was 4.5 years from when we started the company until we launched our first product.

The opposite of YC advice, right?

Yes.

Okay.

Yes. Although there were all these ways in which managing a research team was similar to selecting and advising founders, learning what it took to whatever degree we learned—because I don’t think we did it perfectly—how you manage through this part of the world where you don’t have the external signal from customers, and you’re just trying to do what would normally be the catastrophic startup advice of not shipping a product for 4.5 years.

That was very difficult, and we tried all of these things about how we replaced the signal of “Do customers actually like the product?” with “Is our research actually working?” One of the things that worked, actually, was during the Dota 2 days, when we were trying to use RL to beat people at a video game. We put up a leaderboard, and people could just see how different ideas were performing. That was objective and real, and people wanted to go up on that.

But we had to try all of these things to basically simulate end users, and that was a totally interesting new problem I had no pattern-matching for.

How did you work your way through that? What was your thinking? How did you do this?

We asked a bunch of people who had been at great research labs of the past. OpenAI started at a time when everybody in Silicon Valley, as their vanity project—including me—wanted to start a research lab. There were all these books about the heyday of Bell Labs or Xerox PARC that were very popular. Everybody was talking about this. There was a huge amount of discussion. In fact, I even see one of the books over there about Bell Labs.

But there were not a ton of people who had, in living memory, actually done it. So we talked a lot to Alan Kay. We talked to a handful of other people, and we got some advice from them about what made a really good research lab. Some of it was really good. Some of it didn’t translate as well to the current moment.

Well, you also didn’t have this giant monopolistic profit-printing machine like Bell Labs had. Bell Labs was spun out independently. Polaroid did a lot more research when they had essentially a monopoly on photography.

I just read the biography of the founder of Honda, the guy who created the most successful motor vehicle of all time. The Honda Cub has sold uninterrupted for like 60 years—millions of these vehicles. His whole thing was that he arrived at the same conclusion Bell Labs did: he thought the research and development had to actually be separate. It was spun out of the company and had separate ownership, just like Bell Labs did.

We did not have that.

When I think back to those early days, I mostly feel like I was trying and failing to raise money. That's my dominant memory of the early days of OpenAI. It took so much effort and was so frustrating. I wish we had some sort of cash machine like that.

I remember one of my clearest memories of all of OpenAI. It was announced coming out of 2015, but the first day was right after New Year's in 2016. Twelve of us—or 11 of us—showed up at Greg Brockman's apartment around 9:30 on a Monday or Tuesday morning, something like that. Let's say it was January 4th, and everybody was there. It had been this big effort, and everybody walked in with a lot of excitement. It felt like the first day of school, or whatever.

Then very quickly, people looked around the room and were sort of like, “Well, what do we do now?” Someone said, “Okay, we should get a whiteboard.” Greg got someone to go off and find a whiteboard. The whiteboard came, and we looked around again. “What are we supposed to do now?” You just felt the energy in the room collapse. None of us knew what to do.

It wasn't like building a product startup. It wasn't like, “Let's build this product. Let's talk to customers.” It was like, “Okay, we said we want to make AGI. Maybe we should write some papers.” “Okay, let's write some papers.” “Maybe we should think about some ideas.” “Okay, let's think about some ideas.” Everybody's got their moments of “I have no idea what I'm doing.” That was one of mine. We had just launched this thing, and none of us had any idea what we were going to do.

So we did what we knew how to do, and eventually we figured out that a lot of things didn't work. Eventually, we figured out a kind of rhythm for making and then evaluating research bets. It was far from perfect, obviously, but we did find a gradient that we could kind of progress along. We figured out how to get the resources that very smart people needed and how to make sure that we weren't completely getting lost in the wilderness.

Over some number of years, mostly through chaotic stumbling, we eventually made most of the big discoveries. What started as the unsupervised sentiment neuron paper turned into GPT-1 and then eventually GPT-whatever. The scaling laws work that gave us the confidence not only to buy the compute but also the understanding of how to scale up our models sort of came together through this process, along with many other things.

Through this process, we learned that things like the idea of leaderboards worked. We also learned the incredible power of external demos for an eminent person that the researchers really wanted to impress. Then we learned a bunch of things that didn't work, like fake deadlines.

That has to be so disorienting to live through that experience. You're 12 people in an apartment, you don't even have a whiteboard, and you don't know what to do. Fast-forward a decade, and you have a billion people using it.

Very strange experience.

Do you keep a journal?

When my first kid was born, I would get home at the end of the day, be rocking him to sleep, and just talk to him or whatever. I just needed to come up with things to talk about, so I would tell him about my day, what we were struggling with, what I was worried about, and what was happening. It was kind of fun for me to do, and I thought, “This is sort of interesting, and someday it'll be interesting for him to have this.”

So I started writing him a letter every Sunday. I would talk, just to talk, and then I would write it down. I only ever did 8 of them or something.

How many kids do you have?

Two.

Okay. Bezos has this great line about building Amazon: “We're trying to do stuff that we can tell our grandkids about, that we're proud of,” right? And those things are hard. The fact that you were writing to your son—

Oh, man.

Keep writing the letters. And if you don't do that—this is real quick, just because I've read enough books about this—most of the time, guess what? Founders don't write autobiographies when they're 40. They write them when they're 70, and they're looking back and they wish they could do it again. So much has been lost to the sands of time. They all repeat this. They're like, “I wish I journaled.”

So even if you don't do it, you have enough resources. What I would do is have a book written, even if it's for internal purposes only. Have you ever read The Little Kingdom by Michael Moritz?

I've never read it.

Oh, you have to. It's like the first 6-year history of Apple, written by Michael Moritz. Isn't it crazy that he wrote that book?

He's a phenomenal writer.

Crazy writer winds up being one of the best venture capitalists of all time, I guess. But the point is that the book ends before Steve has even been kicked out of Apple yet. So you get what actually happened. You're going to want this. You might not want it now, but you're damn sure going to want it when you're 60 or 70.

The thing that was so interesting was the mindset of writing to your kid. You really can't hide behind anything. You're like, “I really care what my kid's going to think about me, so this thing happened. I didn't feel great about it. Better do it differently next week.” It was an extremely interesting mental framework. Maybe I'll find some way to do it again.

Oh, maybe you're going to do it.

Okay, Sam, thanks for taking the time. This was awesome, man. Appreciate it.