AI Is Crossing the Frontier of Human Knowledge | Kevin Weil

0:00Kevin Weil:You have no excuse if you've got an interesting idea. You can now create anything that you can think of. The models can now solve problems that humans have never solved before. Going beyond the frontier of human knowledge. That's how AI, I think, and AGI will really change our lives. Why not try and accelerate science, bring about the science of 2050, but in 2030 instead?

0:20Speedrun:Most people think of AI as a productivity tool. Kevin Wheel thinks the biggest impact may be somewhere else entirely. Formerly CPO and Vice President of Science at OpenAI, Wheel is focused on a future where AI doesn't just help people write documents or generate code, but contributes directly to scientific discovery itself. The idea is ambitious. Use AI to accelerate breakthroughs in mathematics, medicine, materials science, and other fields that shape the future of human progress. In this conversation, Kevin discusses frontier science, robotic labs, AI reasoning, startup opportunities, and why he believes some of the most important consequences of AI may come from expanding humanity's ability to discover new knowledge.

1:08Speedrun:All right. This is an incredible time to be alive, I think. You helped build and scale some of the most important technology companies of the last decade, Facebook, Instagram, and Twitter, and now you're doing that at OpenAI. I did ask ChatGPT what it thinks about you. And so as you would expect, it was very complimentary, but you're also like a very accomplished guy. So Kevin is thoughtful, low ego, and unusually grounded for someone who's been at the center of so many high stakes products. So how did you, like, as you looked at those, all those four opportunities, plus many others that were amazing, what gave you the confidence that this is the type of company, this is the team I should be working with.

1:48Kevin Weil:Yeah, number one advice, marry up. It was my wife originally, actually, I was in grad school doing a physics degree. And I met my now wife, who was a Mayfield fellow at Stanford and actually worked at Andreessen for a little while. And she was the one that kind of opened my eyes to everything, startups in the valley and all that. I grew up in Seattle. My dad was an engineer at Microsoft for a long time. So I grew up programming, but still was just like, you know, math and physics, math and physics as I went through grad school. And it was my wife, Elizabeth. She introduced me to Twitter back in the day because she and Jessica Verrilli knew each other from Stanford.

2:29And after seven years at Twitter as it grew, she and Kevin Systrom were also Mayfield Fellows together at Stanford.

2:37Kevin Weil:So that's how that connection happened. And so, you know, a bunch of these things were just my wife, me just following the coattails of my wife. I used to call Sam periodically before, whenever I would like be thinking about doing something new. Sam and I didn't know each other super well, but we knew each other well enough to, you know, to do a quick phone call. And because he always has his like hands in lots of different things. He's like doing fusion startups and all of this stuff. I remember talking to him like in 2020 or something. And he was like, you know, AI will not replace blue collar jobs first.

3:20Kevin Weil:It'll replace white collar jobs first. Coding is going to be one of the big things for AI. This was 2020. And none of us used AI particularly much, at least not outside of like classical ML models that, you know, ranking your feed and stuff. And I just remember being like, yeah, whatever, dude. you know, sure, but like, let's talk about something. And so anyways, the open AI thing happened because I called him and this time he was like, actually, you know what? We have this role open. You should come talk to us. As soon as I did, I was just like, I don't care, like I'll work for free. I don't, just, this is the most interesting thing in the world.

3:57Kevin Weil:And if you give me an offer, I'm coming. So fortunately he did.

4:01Speedrun:And then the original mandate as CPO with Sam was to what? And what are you sort of most happy with what you accomplished during that time?

4:12Kevin Weil:Yeah, originally it was CPO, so leading our consumer products, B2B products, developer products, et cetera. And I mean, man, I think we grew like a weed. I've never seen anything grow that quickly in my entire life. And I think brought, you know, kind of brought AI to a whole bunch of the world. So very proud of that work. A few months ago, it was getting clear that our models could not just be great inside ChatGPT or inside Codex, which I think is an incredible product, but were at the level that they could start to answer frontier scientific problems. Like the models can now solve problems that humans have never solved before.

4:55Kevin Weil:So a lot of people like the criticism of AI is, oh, well, it's just bringing together different ideas from different places and summarizing them for you to give you an answer. I can't actually do novel thinking. But we've now seen there have been, I don't know, what, 10 or 12 just in January, 10 or 12 open mathematics problems solved, mostly by GPT 5.2, now a few recently by Gemini. Models are going beyond the frontier of human knowledge. And I wouldn't claim yet that they are solving problems that humans can't. I think if you took enough people and applied enough mathematicians towards some of these problems, they would have figured it out.

5:36Kevin Weil:But they had not figured it out yet. The model went beyond what we had ever done as humans. And that's pretty cool. And that's today. If there's one thing I've learned over the last few years, it's that you go very quickly from models could never do this thing. It is beyond the capability of AI today to models can just barely do this thing and it kind of sucks at it and like it's wrong most of the time. But you get these glimmers of like, ooh, they can almost, you know, do this. Maybe it only works five or 10 % of the time. And then six to 12 months later, it's like models are great at this thing.

6:11Kevin Weil:And I would never, I wouldn't, I would always use AI anytime I ever do that again. In eval language, you go very quickly from zero to five or 10 % to 80%. We are clearly in that middle phase with frontier science and AI, where you have all these glimmers of like, wow, it can do something that we never thought AI could do. So what more interesting place to apply it than science? I think it may be the most tangible way that we all feel the impact of AGI. if we dropped like GPT-9 inside of chat GPT for you today I'm sure it would be awesome but maybe even more awesome would be that we have all of these new materials and we have superconductivity and we understand the nature of the universe and we have personalized medicine that's how AI I think and AGI will really change our lives so why not try and accelerate science and bring about the science of 2050 but in 2030 instead.

7:11Kevin Weil:And that's our goal.

7:12Speedrun:You know, when did you have your Claude weekend moment or your Codex weekend moment where you're like, wait a second, the world is fundamentally shifted and everything that I've done before is going to be very different going forward.

7:24Kevin Weil:Yeah. So I have a very different take on this. I think it's awesome and exciting. Like when you, when you take something that has historically been the craft of a relatively small number of people, there aren't that many people in the world that know how to program, I don't know, like 30 million people, maybe, and you expand it by a couple orders of magnitude, you get an explosion of creativity because lots of people have ideas. And sometimes they weren't, they didn't have any route to actually implement those ideas. And you go everything from like thinking about people that can start companies now that didn't used to be able to start them all the way through to like, I remember sitting and talking with, this was like a little bit post COVID with this city official somewhere, you know, like small city.

8:09Kevin Weil:And he was telling me about the kinds of programs that they wish they had been able to, to run and operate just like basic kind of information awareness things for the people that lived in this small city. It's like, I just didn't have any way to do it. The data was sitting there. I could, you know, draw what I wanted it to look like, but I had no way of getting it done because he wasn't an expert and they didn't have, you know, however many thousands of dollars to hire a person to do it for the city. Now what would you do? You would like enter a prompt into Codex and it would be done. So I think these kinds of things are awesome.

8:44Kevin Weil:I'll say personally, we just launched Prism like what, a week ago. Anybody see that? Use Prism, tried it. It's like an AI native environment for scientists to do scientific writing and collaboration. So if you're using LaTeX, things like that, because it's a small product in a small team, I've been spending a bunch of time, uh, like writing code, fixing bugs, et cetera, which I haven't done in a bunch of years. And it's super fun. It also, uh, I, I went through this transition in the middle of this, you know, cause most of my day is meetings and you know, we've got a bunch of stuff going on. I went through this transition where I, I, I was sitting in, I don't know, I remember, I think it was like Fiji, a meeting with Fiji and I closed my laptop because we're all trying to be better about not multitasking because that's one of the worst habits that people at OpenAI have.

9:33Kevin Weil:Everyone multitasks all the time. I had not gotten a codex job running before I closed my laptop. And I was like, shit, I just wasted an hour. Not because of the meeting, but just because I could have been multitasking during that hour. My codex agent could have been fixing a bug or implementing a feature or doing something for me. and now I have to sit in this meeting just like, you know, like a farmer being in a meeting. Come on. And so, you know, the same thing before you go to bed at night, you're like, okay, what like really hard task can I give Codex and just let it chunk away for like 10 hours?

10:14Kevin Weil:So like that to me is a different world and if you're really good at it, you are not just juggling one job. You've got, you know, three or four things running in parallel across different work trees and like, I don't know, what a cool world.

10:26Speedrun:Yeah. The creativity that's going to be unleashed is just incredible. And I think about it as like, if you were a handcraft furniture maker, like 100 years ago or 150 years ago, industrial revolution hits, all of a sudden there's factory mass produced furniture that's better than yours, or at least equivalent. And you're saying, wait a second, my whole world has changed, but now you can design something that can be used by so many more people all at once. But like it's this disorientation that people feel because it's so disruptive and so quick at the same time. I don't know.

10:58Kevin Weil:You think we'll end up with like because now you still people still value custom made furniture. I think we'll end up with like bespoke websites. Yeah. Like this was done by a human.

11:08Speedrun:Exactly. Exactly. Same as like, you know, people will be like, I drive cars like as a hobby. Right. Right. Right. You're like, get off the road. You're dangerous. Yeah, stay over there in your little area for human drivers. With things moving so quickly, you gave an example of how you stay up with things. You're doing work with your team. You're staying close to your team. Do you have any other ways that you're both able to lead the team, lead the strategy, but also stay close to all the developments that are happening within your company, within other companies? That feels like more than a full-time job in its own right.

11:43Kevin Weil:Yeah, I mean, the industry is just moving insanely fast, right? I've never seen anything like it. It's exhilarating, and it's fun, and it's also, it's a lot to keep up with. But I think that, I just think this moment kind of selects for people who are high agency, because you can now create anything that you can think of, and you have no excuse if you've got an interesting idea not to, like, get, you know, Codex thinking about it while you do something else, whatever you were originally going to do in the morning. keep doing that but have Codex working on your idea in parallel and sometimes you'll wake up in the morning have an idea and have a thing implemented by the time you're done with the day in addition to doing what you thought you were going to do during the day so like people that are high agency people that are really curious people that learn quickly those skills are more valuable than ever in this moment and you know kind of whatever the future holds I think those skills are going to see us through.

12:37Speedrun:Yeah I've heard you talk about a vision of both the experiment design, but also then the experimentation itself and the validation that was very captivating to me. Maybe you could share it here a little bit more.

12:50Kevin Weil:This OpenAI for Science group is only a few months old, although in some sense, of course, OpenAI has always cared deeply about science. So I feel like even though we're a smallish team, we kind of have the might of all of OpenAI research at our back because every researcher at OpenAI cares about science and scientific data has been one of the ways that we have have improved our models for a long time. But so we started thinking about math, physics, theoretical computer science, because you can do everything in silico. You have like closed loop systems that you can optimize. And, you know, part of this is teaching the models to answer really hard scientific problems, teaching them to think not for like 10 minutes or maybe the hour that you can get GPT-5 Pro to think if you ask it a really hard question, but teaching models to stay on track for a day, two days, a week, two months at a time to answer even harder problems.

13:49Kevin Weil:Because just like you and me, if you could give me problems that I couldn't solve in 20 minutes, but I could give in two hours. Same is true of the models. The more time they think, the more impressive problems they can solve. So you start with things like math and physics, but then some of the biggest ways, the most important ways that accelerating science is going to feed back into all of our lives in a positive way is through stuff that we can feel in real life. It's like, you know, one of our relatives survives cancer because we've made advances in medicine. You have new devices and materials because we've been able to make advances in material science.

14:25Kevin Weil:And those things require labs. You can't do those just in silico. Although I think the importance of simulation is going to go up pretty meaningfully because you will be able to apply huge amounts of compute to these problems. But then you're still going to need experimental validation. You're going to need to try things in the real world. I think it's going to be a while before we have a model that can, you know, first principles go from like a quark all the way through a model of a cell, all the way through human biology. Like experiment matters. So, you know, you start to think about how you do that at scale.

15:00Kevin Weil:There's lots of opportunity to partner with existing labs, and we will, but I think there are really interesting, you know, the science of the future will definitely involve robotic labs and reinforcement learning loops that go through the real world where the model is thinking, maybe running a simulation, thinking some more, refining the experiment that it can run using the best possible parameters, and then sending that to a bunch of robotic labs, which by the way you can scale horizontally having the experiments run in real life the results come back to the model the model thinks, runs more simulations thinks and you have this you have like multiple loops you have tight loops with the model thinking and the simulation you have longer loops that go through the real world that's how a lot of science is going to be done in the future and that is its own form of acceleration you think you have robotic labs that you can scale horizontally that can run 24 hours a day they're not grad students pipetting things that need to take breaks and sleep.

16:00Kevin Weil:And then the grad students can do things that are much more leveraging of what makes us human than pipetting things. So I'm quite optimistic about where this goes and about our ability as a society to accelerate the pace of science very meaningfully.

16:18Speedrun:How far away do you think we are from that? What are the key technologies? Is it robotics? What needs to happen for us to unlock that fully?

16:27Kevin Weil:I mean, some of that piece is robotics. It's already happening, right? I was just giving examples of the model solving open math problems. And there are certainly robotic labs out there already. It's all kind of in the early adopter phase. But at the pace that we're all moving, I don't think it's long. and it's so clearly the right thing for many fields that this is not you know we are not the only ones to have this idea there are a lot of people that have this idea there are a lot of interesting startups building things along this these lines and you know the world is just moving

17:02Speedrun:so fast it can't be long right right you've shipped products used by hundreds of millions or billions of people um what was a product decision you were nervous about that uh ended up being right.

17:14Kevin Weil:You know, the fun thing with product decisions at that scale is I, any example I give, I bet there are people in the audience who were like, no, no, no, no, you got that one wrong. Very few are unambiguously right. Probably one was like ranking the Twitter feed, which was extremely controversial back in the day. Twitter used to be completely real time. And the most, you know, the thing you saw at the top of your Twitter feed was the thing that was tweeted one second ago. And the next one was the one that was tweeted four seconds ago. And if your, you know, spouse or your best friend happened to tweet an hour ago, like, yeah, too bad.

17:50Kevin Weil:You were never going to see it. It's going to get totally buried. But there were a lot of people that said, this is the magic of Twitter. How could you possibly do that? You know, you're becoming Facebook now. So that was very controversial at the time, although it seemed in some sense, like, how could you not want, you know, you do care about different people's stuff more than other people's stuff. How could you not want ranking if we could do it well? And if we could bring the right balance of recency and everything else. So that was one. And I think Facebook saw the same thing when they originally put out the news feed.

18:24Kevin Weil:You have a bunch of people that are super upset, but then the metrics tell you an incredibly positive story, like double digit positive kind of thing. And so, you know, and then you just, you can just keep making that better and keep getting wins there. It was interesting trying to figure out how exactly we would, when we rolled out a one preview, the first reasoning model, what was the right kind of UX paradigm for a model that would not give you an immediate answer? Like all the other previous chat models, you type in an answer or you type in a question, you basically get an answer right away.

19:03Kevin Weil:There aren't a lot of experiences online where you have to wait like that. So, and the model is doing this interesting thing with its chain of thought in the meantime, which we didn't want to expose completely because we didn't want to, because you can distill that and, you know, basically copy our model, which, you know, for a bunch of geopolitical reasons, we didn't want to have happen. But you want to show some. So it was very interesting trying to figure out how you, was the model that people would go away and just pop back in whenever it was done? Or were they going to watch? And if they were going to watch, what would we show?

19:43Kevin Weil:And how do we balance not giving too much information that would lead to model distillation but would be interesting? So that was an interesting experiment. And it's surprising sometimes building stuff with models, a reasonable analogy for how should I, you know, handle the UX of this current situation is how would I want a human to behave, you know, in a similar situation? And like, if you ask me a question that I can immediately answer, you know, GPT-4 style, then I will just immediately answer, right? If you ask me a question that I need to think about, I don't immediately start babbling my entire chain of thought, right?

20:22Kevin Weil:I don't just like spew out whatever goes through my mind. I also don't like completely turn around and go mute and like do nothing until I come back to you with an answer a minute and a half later. You know, I might say like, huh, okay, that's an interesting question. Let me think. And then you kind of give little like CliffsNotes as you think, well, it could be this. No. And so that ultimately is kind of what the model does, right? It gives you kind of like periodic updates of what it's thinking about as it's thinking, which is interesting in and of itself. And then it comes back with the answer.

20:57Kevin Weil:So we tried to model it a little bit off that.