Sam Altman CEO of OpenAI: ChatGPT, Future of AI, AGI, and China

0:08open ai shocked the world uh last november with chat gpt and um open ai is not only creating models It's creating the future. So Sam, it's an honor to have you on the podcast. Thanks a lot for having me. It's great to be here. How does it feel to spearhead this revolution? It's definitely a little surreal. It's like a very exciting moment in the history of technology. And to get to work with the people who are creating this is a great honor. And I can't imagine anything more exciting to be doing. No, I can't imagine. It's definitely a lot. I can see that. Now, big picture, what's the vision of the world where humans and AI coexist?

0:55Well, you know, one thing that we believe is that you have to answer that question empirically. There's been a lot of philosophizing about it for a long time. Very smart people have had very strong opinions. I think they've all been wrong, and it's just a question of how wrong. And the course that a technology takes is difficult to predict in advance. I love that Alan Kay quote that the best way to predict the future is to invent it. And so what we're trying to do is see where the technology takes us, deploy it into the world to actually understand how people are using it, where the risks are, where the benefits are, what people want, how they'd like it to evolve.

1:37And then sort of co-evolve the technology with society. And, you know, I think if you asked people five or 10 years ago what the deployment of powerful AI into the world was going to look like, they wouldn't have guessed that it looks like this. People had very different ideas at the time, but this was what turned out to be where the technology leads and where the science leads. And so we try to follow that. And how far into the future can you see now? uh the next few years seem pretty clear to us you know we kind of know where these models are going to go we have a roadmap we're very excited about uh we can imagine both the science and technology but also the product a few years out and beyond that you know we're gonna learn a lot we'll be a lot smarter in two years than we are today yeah and and what kind of uh you know holy shit moments have you had lately?

2:33Well, remember that we've been thinking about this and playing around with this technology for a long time. So the world has had to catch up very quickly. But we have less holy shit moments because we've been expecting this and we've been building it for a while. It doesn't feel as discontinuous to us. And what kind of big things have you seen since chatty patty four well we

3:05the biggest ones have not been about new technology or new models but about the breadth of use cases the world is finding to do this so the holy shit moments have not been like oh now the model can do this now we now we figured out that because again you know somewhat expected that. But seeing how much people are coming to rely on these models to do their work in their current form, which is very imperfect and broken, you know, we're the first to say, these models are still not very good. They hallucinate a lot. They're not very smart. They have all these problems. And yet, people are using their human ingenuity to figure out how to work around that and still leverage these tools.

3:45And so watching people that are remaking their workflows for a world with LLMs has been big. And some examples of new things you've seen, new user cases, applications? You know, a common one is around how developers are changing their workflow to, you know, spend like half their time in ChatGPT. You hear people say they feel like two or three or sometimes more times productive than before. an uncommon one is i met a guy who runs a laundromat business as like a one person thing and uses chat gpt for um coming up with marketing copy uh dealing with like customer service uh helping review legal documents he had a long list of things and he's like i got a virtual employee in every category that was pretty cool and what about things like uh brain implants and getting it to help with speech and so on which we just saw recently um i'm very excited about neural interfaces but i am not currently super excited about brain implants i don't feel ready to want one of those i would love a device uh that could like read my mind and but i'd like it to do that without having to put a hole in my skull and i think that's possible how oh there's many technologies depending on what you'd want but you know there's there's a whole bunch of companies working on trying to sort of like read out the words you're thinking without requiring a physical implant.

5:18Now, a few years ago, nobody had heard about OpenAI. Now, everybody's heard about it. You are, you know, one of the most famous people on earth. But the people, so how many people are you at OpenAI now? 500 or so. And what do these 500 people actually do? it's a mix so there's a large crew that's just doing the research like trying to figure out how we get from the model we have today which is very far from an AGI to an AGI and all of the pieces that have to come together there so scaling the models up coming up with new methods that that whole process there's a team that makes the product and figures out also how to scale it there's a sort of traditional Silicon Valley tech company go-to-market team.

6:06There's a very complex legal and policy team that does all of the work you'd imagine there. Yeah. And so your priorities as a CEO now, how do you spend your time?

6:24I kind of think about the buckets of what we have to do in research, product, and compute on the technical side. And that's sort of the work that I think I enjoy the most and where I can contribute the most. and then I spend some of my time on policy and sort of social impact issues for lack of a better word and then the other things I spend less time on but we have great people that run the other functions now your mission has been to ensure that artificial well the general intelligence benefits all of humanity what's the biggest challenge to this you think

7:11So I, a couple of thoughts there. One, I'm reasonably optimistic about solving the technical alignment problem. We still have a lot of work to do, but, you know, I feel like I feel better and better over time, not worse and worse. The social part of that problem, you know, how do we decide whose values we align to? Who gets to set the rules for this? How much flexibility are we going to give to each individual user and each individual country? we think the answer is quite a lot but that comes with some other challenges um in terms of how they're going to use these systems that's all going to be uh you know difficult to put it lightly for society to agree on and and then how we share the benefits of this what we use these systems for uh that's also going to be difficult to to agree on um kind of the the buckets i think about here are we've got to decide what global governance over these systems as they get super powerful is going to look like, and everybody's got to play a role in that.

8:19We've got to decide how we're going to share the access to these systems, and we've got to decide how we're going to share the benefits of them. There's a lot of people who are excited about things like UBI, and I'm one of them, but I have no delusion that UBI is a full solution or even the most important part of the solution. People don't just want handouts of money from an AGI. They want increased agency. They want to be able to be architects of the future. They want to be able to do more than they could before. And figuring out how to do that while addressing all of the sort of, let's call them disruptive challenges, I think that's going to be very important but very difficult.

9:01How far out is true AGI?

9:07I don't know how to put a number on it. I also think we're getting close enough that the definition really matters and people mean very different things when they say it. But I would say that I expect by the end of this decade for us to have extremely powerful systems that change the way we currently think about the world. and and you say we've got different definitions what is what is your definition of general intelligence so one that matters you know there's like kind of the open ai official definitions and then there's one that's very important to me personally when we have a system that can figure out new scientific knowledge that humans on their own could not i would call that an AGI.

9:56And that you think we may have by the end of this decade? Well, I kind of tried to like soften that a little bit just by saying we'll have systems that like really change the way the world works. The new science may take a little bit longer or maybe not. What's the end game here? Are we just all of us going to work a lot less?

10:19um you know i want to be people i think we'll all work differently i think we'll still many of us will still work very hard but differently every technological revolution um people say that we're just gonna do less work in the future and we just find that we want a higher standard of living and new and different things and also that we find new kinds of work we really enjoy you know neither you nor i have to work and i bet we both work pretty hard and i love i mean i just love my job i love my job and i feel very blessed so the definition of work what we work on why we work the reasons for it i expect that all to change what we do i expect to change but i love what i do and i expect people in the future to love even more what they do because there'll be new amazing things to work on that we can hardly imagine right now and less boring stuff yeah i'm all for getting rid of the boring stuff like i think like everybody should love that that's maybe one thing we could say in the future is everybody will do things that they love you won't have to do things you don't i think most people probably don't love their jobs right now um i believe you just traveled the world and met with a lot of people and users uh what what was your main what was your main takeaway

11:43the level of excitement about the future and what this technology is going to do for people around the world in super different cultures and super different contexts was just very very different than i expected

12:02like it was it was like overwhelming in the in the best way any any difference between geographies or industries yeah like you know in in the developing world um people are just focused on what this can do economically right now uh and in the more developed world there's much more of a conversation about what the downsides are going to be and how this is going to disrupt things. And there's still excitement, but it's tempered more by fear. That was a striking change, a difference. Do you think it will lift up the poor part of the world? Yeah, I really do. I think it's going to make everybody richer, but I think it impacts, positively impacts poor people the most.

12:49And I think this is true for most kinds of technology. But it should be particularly true for the democratization of intelligence. You or I can afford to pay a super highly compensated expert if we need help, but a lot of people can't. And to the degree that we can make, say, great medical advice available to everyone, you and I benefit from that too, but less than people who just can't afford it at all right now. And what would potentially prevent this from happening?

13:29Well, we could be wrong about the trajectory that technology is on. I think we are on a very smooth exponential curve that has much, much further to go. But, you know, we could be like missing something. We could be drinking our own Kool-Aid. We could hit a brick wall soon. I don't think we're going to. I think we have some remarkable progress ahead of us in the next few years. But But we could somehow be wrong for a reason we don't understand yet.

13:56What is it doing to the global balance of power?

14:06I don't know how that's going to shift. I'm not sure anyone does, but I certainly don't think that's something that I'm particularly well qualified to weigh in on. but it just seems like it's being it's so key now to the the weapon race the medical race the self-driving vehicle race just all these races but it's also available pretty broadly you know like one of the things that we think is important is that we make gpt4 extremely widely available um even if that means people are going to use it for things that we might not always feel are the best things to do with it. But we have a goal of globally democratizing this technology.

14:58And as far as we know, GPT-4 is the most capable model in the world right now. And it is available to anyone who wants to pay what I think are the very cheap API rates. Now, anyone is not quite there. We block a handful of countries that the US has embargoes with or whatever. But it's pretty available to the world. But in order to develop it further, you need, well, you need the right chips, right? And they are not available. You need the right chips. You also need the right people. And I would argue that's even more of a constraint. Like there are people who can go copy GPT-4, right? Like we did it fine.

15:39Once people know how to do it and know that it's possible, a lot of people will replicate it. but what matters is how you're going to get to like gpt6 and 7 and also even more than that how you're going to get the next set of very different ideas that take you on a different trajectory like everyone knows how to climb this one hill and we're going to go figure out the next hill to climb and there's not a lot of people in the world that can do that uh but we're committed to making that as widely available as we can do we know where china is here we don't maybe someone doesn't want here. Do you think there's a chance that, well, like they did with weapons, they just suddenly bang, they had the supersonic rockets.

16:21We didn't even know they existed, right? Could that happen? Yeah, totally. It could. I mean, we're going to work as hard as we can to make sure that we stay in the lead, but we're a little in the dark. So Mark Andreessen, for instance, he thinks we should stuff it into everything. and you know as part of the geopolitical fight what do you think stuff it into everything mean just like put it everywhere yeah that's happening and i think that's great like without revealing something i shouldn't the amount of gpt4 usage and the number of people companies that are integrating it into different ways is staggering it's awesome some examples if you had to reveal something

17:15uh i mean like you know car makers are putting it into cars and i was like all right that sounds like a gimmick and then i got to try a demo of it and i was like wow this being able to just talk to my car and control it in a sophisticated way entirely by voice actually totally changes my experience of how I like use a car in a way that I would not have believed was so powerful. So for instance, you see it in a car, what do you say?

17:45This is probably where I don't want to like reveal a partner's plans, but you can imagine a lot of things that you might say, like the basic stuff is easy. Like, I need to go here and I'd like to listen to this music. And also, can you make it colder? sounds good do you depend on uh newer and even more powerful chips than what we have now i mean how much quicker do how much more complex the chips need to be than h100 or the latest things from nvidia um yeah of course like there's the ways the ways that we can keep making these models better are we can come up with better algorithms or just more efficient implementations or both.

18:32We can have better chips and we can have more of them. And we plan to do all three things and they multiply together. And do you think it's the chip makers who will end up with the profits there?

18:49They will end up with profits. I wouldn't say the profits. I think there's many people who are going to share this massive economic boon.

19:03How much does it cost to train these models? I mean, how much have you spent on pre-training models? We don't really talk about exact numbers, but quite a lot. Yeah. And what's the challenge of spending so much money pre-training and then it lasts for a relatively short period of time, in a way you have to depreciate the whole investment because you need to invest more in the next generation. I mean, how do you think about this? That's true. I don't think there are going to be as many massive pre-trained models in the world as people think. I think there will be a handful and then a lot of people are going to fine tune on top of that or whatever.

19:50So how do you read the competitive part of it that i think is important is like you know when we did gpt4 um we did we produced this artifact and people use it and it generates all this economic value and um you're right that does depreciate fast but in the process of that we learn so much about how to go we push the frontier of research so far forward and we learn so much that'll be critical to us being able to go do GPT-5 someday or whatever. That it's like, you're not just depreciating the CapEx one time for the model. You have generated a huge amount of new IP to help you keep making better models.

20:32So the way you read the competitive landscape now, what does it look like?

20:40I mean, there are going to be many people making great models. will be one of them. We'll contribute our EGI to the world, to society, among others. And I think that's fine. And we'll all... We'll run different experiments. We'll try setting... We'll have different features, different capabilities. We'll have different opinions about what the rules of a model should be. And through the magic of competition and users deciding what they want, we'll get to a very good place. How far ahead do you think you are a competition? I don't know. I don't think about it that much, to be honest. Our customers are very happy.

21:24They are desperate for more features and more capacity and us to be able to deliver our service in all of these little better ways. And we're very focused on that. I'm sure Google will have something good here at some point. But I think they're racing to catch up with where we are. and we're thinking very far ahead of that. So normally in the software business, you have something which is very cheap where you ship a lot of it or something which is very expensive and you don't ship so much. Here you could potentially ship something. And I can see you smiling here. Here you can potentially ship a lot of areas.

22:03Exactly. So tell us, how is this going to work? You know, I'll tell you, One of the most fun things about this job is we are past the point as a company. I am past the point as like a CEO running this company where there's like a roadmap to follow. We're just doing a bunch of things that are like outside of the standard Silicon Valley received wisdom. And so we get to just say, well, we're going to figure it out and we're going to try things. And if we got it wrong, like who cares? There was no like, it's not like we like screwed up something that was already figured out. I mean, back to our very founding, like most big tech companies are a, they start as a product company and eventually they bolt on a research lab that doesn't work very well.

22:51And we started as a research lab and then bolted on a product company that didn't work very well. And now we're making that better and better. I'm sorry, product company, you mean Microsoft? No, no. I mean like having to figure out how to ship the API in ChatGPT. Yeah. like we started we really did just start as a research lab and then one day we're like we're gonna make a product and then we're gonna make another product and now that product is like the fastest growing product in history or whatever and we weren't set up for that is the usage of Chattipity decelerating nope

23:28I think it maybe took like a little bit of a flat line during the summer which happens for lots of products but it is going up

23:39tell us about the relationship with microsoft how does that work

23:47um i mean at a high level they build us computers we train models and then we both use them and it's a pretty clear and great partnership are you have you are your goals aligned yeah they really are um one of i mean there's like there's of course areas where we are not perfectly aligned and like i don't like any partnership in life or business or whatever i won't pretend it's perfect but it is very good and we are aligned at the highest levels which is really important and the the misalignments that come up at the sort of lower levels once in a while you know like no contract in the world is what makes a partnership good like what makes a partnership good is that when those things happen you know satya and kevin and i talk and you figure it out and you know there's like a good spirit of compromise over a long time now they've been one of the initiators and i mean together with you in terms of self-regulating this space what can this type of thing be self-regulated not entirely um i think it needs to start that way and i think that's also kind of like how you figure out a better answer but like governments are going to have to do their own thing here and You know, we can provide input to that, but we don't.

25:22We're not like the elected decision makers of society, and we're very aware of that. And what can governments do? Anything they want. And I think people forget this. Like, governments have quite a lot of power. They just have to decide to use it. Yeah, but I mean, so let's say now Europe decides that they're going to regulate you really harshly. I mean, are you just going to say goodbye, Europe, no? Possibly. I don't think that's what's going to happen. I think we have a very productive conversation. I think Europe will regulate AI, but reasonably, not very harshly. And what is a reasonable regulation?

26:03What is that level? I think there's many ways it could.

26:13I think there's many ways that it could go that would all be reasonable. But, you know, like to give one specific example, and I'm surprised this is controversial at all, but a regulatory thing that's coming up a lot in Europe and elsewhere is that if you're using an AI, you've got to disclose it. So if you're talking to like a bot and not a person, you need to know that. that seems like a super reasonable and important thing to do to me for a bunch of reasons given what's starting to happen um to my surprise there's some people who really hate that idea but i'd say that's like a very very um reasonable regulation i agree i agree do you think we'll get global regulation is there any um shape or form that can happen?

27:06I think we're going to get it for only the most powerful systems. So, you know, I think like individual countries or blocks of countries are not going to give up their right to self-determine for like, you know, what can a model say and not say and how do we think about the free speech rules and whatever. But for technology that is capable of causing grievous harm to the entire world, like we have done before with nuclear weapons or a small number of other examples, yeah, I think we are going to come together and get good global regulation. But given how embedded it now is in everything, as we spoke about, you know, weapons, your car, you're sitting in your car and it's like super cool and it's cold and hot and music and this and that and, you know, and you are a Chinese car company and you want to compete the Americans, why do you want to have a regulation on this?

27:53Well, GPT-4, I don't think needs global regulation, nor should it have it. I'm talking about like what happens when you get to GPT-10 and it is, you know, say smarter than all of humans put together.

28:07and that's where you think we'll get it that's when i think we'll get it when you have the cost of intelligence coming down so dramatically like it is now what is it going to do to productivity in the world i mean it's supposed to go up a lot right that's what theory tells us and that's what i think so so um i've told everybody in in our company that hey we should we should improve our productivity by 10 of the next 12 months all of us and that's and you know how i got the number did you ask chat gbt no i just took it i just took it straight out of the air do you think what do you think about that number is it low or high under ambitious what should productivity increase by how do you measure the stuff we do

29:06that's not very good measurement but just the kind of stuff that I produce how much of your company writes code uh if uh 15 well people in in technology probably 15 20 of us

29:28more actually but okay let's let's say that's 20 writing code i think an overall goal of like you know 20 productivity increase in a 12-month period is appropriately ambitious given the tool and given the tools that we will launch over the next 12 months. Okay. Sounds like I should up the game here a bit. I think so. Yeah, I'll just tell everybody. You told me to, so that's fine. It's better to set a goal that is slightly too ambitious than significantly under-ambitious, in my opinion. Yeah. Now, is there an inherent limitation to what AI can achieve? I mean, is there a point of no further progress?

30:11I couldn't come up with any reasonable explanation of why that should be the case.

30:23You say that most people overestimate risk and underestimate reward. What do you mean by that?

30:33um you know there's a lot of people that don't go start the company or take the job they want to take or try a product idea because they think it's too risky and then if you really ask them like all right can we unpack that can you explain what what the risk is and what's going to go wrong it's like well the company might fail okay and then what you know well then i have to go back to my own job, my old job. All right, that seems reasonable. And they're like, well, you know, but I'll be a little embarrassed. And I'm like, oh, is that, you know, what's the call? I think like people view that as a super risky thing.

31:11And they view staying in a job where they're not really progressing or learning more or doing new things for 20 years as not risky at all. And to me, that seems catastrophically risky, you know, to like miss out on 20 years of your very limited life and energy to try to do the thing you actually want to do.

31:35That seems really risky. But it's not thought of that way. Talking about staying in your job, what, so the leaders and the CEOs. So how is AI going to change the way leaders need to act and behave? Well, hopefully it's going to do my job. Hopefully the first thing we do with AGI is let it run OpenAI and I can go sit on the beach. That'd be great. We don't want to do that for long, but right now it sounds really nice. How do you develop the people in your company? How do you develop your leaders?

32:13um i think developing leaders tend to fail at the same set of things most of the time you know they don't they don't spend enough of their time hiring talent and developing their own teams they don't spend enough of their time articulating and communicating the vision of their team they don't spend enough of their time thinking strategically because they get bogged down in the details. And so when I like put a new person in a very senior role, which I always try to do with promotions, I mean, I'm willing to hire externally, but I'd always, always rather promote internally. I have the moment for dinner or go for a walk or sit down or something and say like, here are the ways you're going to screw up.

32:55I'm going to tell you all of them right now. You're going to totally ignore me on this and not believe me, or at least not do them because you're going to think you know better, you know, not make these mistakes. But I'm going to put this in writing and hand it to you. And we're going to talk about it in three months and in six months. And, you know, eventually I think you'll come around and they always ignore me and always come around. And I think just like letting people recognize that for themselves, but telling them up front so that it's at least in their mind is very important. What's the most common way leaders grew up?

33:32failing to recruit slash promote, and then failing to build a good delegation process. And then as a consequence of those, not having enough time to set strategy because they're too bogged down in the day-to-day and they can't get out of that downward spiral. What does your delegation process look like? Two things. Number one, high quality people. number two setting the training wheels at the right height and increasing them over time as people learn more and I build up more trust is that the way to manage geniuses

34:15researchers that's a different thing I was like talking about how to like executives that run the thing what about the geniuses the prima donnas yeah

34:30explain, well, pick really great people, explain the general direction of travel and the resources that we have available and kind of at a high level where we need to get to to get to the next level. So, you know, we have to achieve this to go get the next 10 times bigger computer or whatever.

34:50And, you know, provide like the most mild input on it would be really great if we could pursue this research direction and this would be really helpful and then step back. So we kind of like, you know, we set a very high level vision for the company and what we want to achieve. And beyond that, researchers get just a huge amount of freedom. Do you think companies generally are too detailed in the remit they give the teams? Yes. I mean, at least for our kind of thing. I think managing – we talked earlier about having to rediscover a bunch of things. I say this realizing it's going to come across as arrogant, and I don't mean it that way, but I think it's an important point.

35:42there used to be great research that happened in companies in silicon valley you know xerox park being the obvious example there have not been for a long time and we really had to rediscover that we made many screw-ups along the way to learn how to run a research effort well and how you balance letting people go off and do whatever towards trying to get the company to point in the same direction. And then over time, how to get to a culture where people will try lots of things, but realize where the promising directions are and on their own want to come together to say, let's put all of our firepower behind this one idea because it seems like it's really working.

36:24You know, I'd love to tell you, we always knew language models were going to work. That was absolutely not the case. We had a lot of other ideas about what might work. But when we realized the language models were going to work, we were able to get the entire research trust or almost entire research brain trust to get behind it. I'm slightly surprised you say that there was no innovation culture in Silicon Valley because that's a bit contrary to what I thought. So who do you want to explain this? Yeah, there's a product innovation culture for sure, a good one. But like, I mean, again, I hate to say this because it sounds so arrogant, but like before OpenAI, what was the last really great scientific breakthrough that came out of a Silicon Valley company.

37:07And why did that happen? What happened there? Well, we got a little lucky. No, I don't mean you. Why did this culture disappear in Silicon Valley, you think? I have spent so much time reflecting on that question. I don't fully understand it. I think it got so easy to make a super valuable company and people got so impatient on timelines and return horizons that a lot of the capital went to these things that could just fairly reliably multiply money in a short period of time by just saying we're going to take the magic of the technology we have now, the internet, mobile phones, whatever, and apply it to every industry.

38:03That sucked up a lot of talent, very understandably. Now, you had some, what should we say, your co-founders are pretty into big, hairy goals, right?

38:19Yeah. I mean, we're trying to make AGI. I think that's the biggest, hairiest goal in the world. So not so many companies have those kind of co-founders and people with that kind of track record. and that type of talent magnet funding capabilities and so on, how important was that? You mean Elon by this, right? Yeah. Yeah. And some of the other people you worked with in the beginning? Well, there were six co-founders, Elon and me, Greg and Ilya, and John and Wojciak. And, you know, Elon was definitely a talent magnet and attention magnet for sure. And also just like has some real superpowers that were super helpful to us in those early days, aside from all of those things, you know, contributed in ways that we're very grateful for.

39:11But the rest of us were like pretty, pretty unknown. And I mean, maybe I was like somewhat known in technology circles because I was running Y Combinator, but not like a, not in a major way. And so we just had to grind it out. But that was a good and valuable process. What is your superpower?

39:42I think I'm good at thinking very long-term and not being sort of constrained in common wisdom.

39:58yeah you said in 2000 i think i'm good at evaluating talent that was like a really helpful thing to learn from y combinator you said in 2016 that long-term thinking is a competitive advantage because almost no one does it yeah i mean when we started open ai and said we're going to build agi everybody was like that's insane a it's 50 years away and b it's like you know the wrong thing to even be thinking about you should be thinking about this, how to improve this one thing this year. And, you know, also this is like unethical to even say you're working on it because it's like such a science fiction and you're going to lead to another AI winter because it's too much hype.

40:35And we just said, it's going to take us a while, but we're going to go figure out how to do it. You said you also good at assessing talent. How do you do it?

40:46I don't know. I don't, I can't like, I have a lot of practice. But I don't have words for it. I can't tell you here's the five questions I ask or here's the one thing I always look for.

41:07But assessing if someone is smart and if they have a track record of getting things done and if they have novel ideas that they're passionate about, But I think you can learn how to do that through thousands of conversations, even if it's hard to explain. Why is Europe so behind generally when it comes to innovation and innovative culture? I'd ask you that. I don't know. Why is it? Well, is it, first of all? Well, I guess it is. look at where the big tech companies are, where the big innovations come. It's certainly behind. It's certainly very behind in like hyperscale software companies. There's no question there.

41:54It's big fear of failure. It's a cultural thing. There are a lot of things going into that cocktail, I think. The fear of failure thing and the kind of like the cultural environment or backdrop there is huge, no doubt. We funded a lot of European people at YC, and a thing they would always say is, they cannot get used to the fact that in Silicon Valley, failure is tolerated. I've failed at stuff big time, and I'm sure I'll fail at stuff in the future. What's the biggest failure so far? Well, I mean, monetarily-wise, I've made a lot of big investments that have gone to total, like just, you know, zero, like crater in the ground.

42:45But in terms of like time and psychological impact on me, I did a startup from when I was like 19 to 26, worked unbelievably hard to consume my life and failed at that. And that was like quite painful and quite demoralizing. And it's like, you know, you learn to get back up after stuff like that, but it's hard. How do you get back up?

43:12um i mean one of the key insights for me was realizing that although i thought this was like terribly embarrassing and shameful uh no one but me spent much time thinking about it who do you ask for advice like personally my strategy is not to just have like one person that i go to with everything and a lot of people do that. They have one mentor that they go to for every big decision. But my strategy is to talk to a ton of different people when I'm facing a big decision and try to synthesize the input from all of that. So if I'm facing a real major strategic challenge for OpenAI, kind of one of these bet the company things, I would bet that counting people internal and external of the company, I'd talk to 50 people about it.

44:09And probably out of 30 of those conversations, I would hear something interesting or learn something that updates my thinking. And that's my strategy. So now, outside AI, what are you the most excited about? Fusion. I think we're going to get Fusion to work very soon. And I think my model, if you boil everything down to get to abundance in the world, the two biggest, most important things are bringing the cost of intelligence way down and bringing the cost and amount of energy way down. And I think AI is the best way to do the former and fusion is the best way to do the latter. and in a world where we look at energy that's like less than a penny per kilowatt hour and more importantly, we can have as much as we want and it's totally clean.

45:03That's a big deal. Do you think it's going to solve the climate problem? Yes. We'll have to use it to do other things. Like we'll have to use some of it to capture carbon because we've already done so much damage. But yes, I do. What about crypto?

45:20I am excited for the vision of crypto. and it has so far failed to deliver on that promise.

45:31But you have plans? It's not something I'm spending that much time. OpenAI has taken over my whole life, so I can have a lot of plans about OpenAI, and there's other projects that I've invested in or helped start that I feel bad because I don't have much time to offer them anymore, but they're all run by super capable people, and I assume they'll figure it out. What do you read? Um, the thing that has unfortunately gone the most by the wayside for me recently has been free time and thus reading. So I don't, I don't get to read much these days. Uh, I used to be a voracious reader and, uh, there was like one year where I read, you know, not fully, but like more than a skim.

46:19I read 50 textbooks and that was like an unbelievable experience. but I don't like this last year I have not read many books

46:32what's the one book young people should read

46:41that's a great question picking one is really hard

46:50I don't think

46:54man that's such a good question

47:01I don't think it's the same for every young person and like coming up with a generic singular recommendation here is super hard

47:16I don't think I can give a faithful answer on this one good now we are we are fast forwarding here oh you know what can I actually I do have this is not the one for every young person but I wish a lot more people would read The Beginning of Infinity early on in their early on in their career or their lives The Beginning of Infinity The Beginning of Infinity why? I think why doesn't matter we'll find it I think it's the most inspiring you can do anything you can solve any problem and it's important to go off and do that it's a very like i felt it was like a very expansive book of the way i thought about the world well sam i think that's a very suitable place to uh to go in for landing now last one so um fast forward a couple decades um people sit down and reflect on sam altman's impact on the tech world and society what what do you hope what do you hope they'll say what do you what do you hope your legacy will be you know i'll think about that when i'm like at the end of my career like right now i my days are spent like trying to figure out why this executive is mad at this one and why this product is delayed and like why our network on our you know big new training computer is not working and who screwed that up and how to fix it and it's like very caught up in the like annoying tactical problems there is no room to think about legacy, we're just trying to go off and build this thing Fantastic Well, good luck with that It's been an absolutely fantastic conversation and all the best of luck and go get them Great talking to you Thank you for having me Wow, that was great