O3 and the Next Leap in Reasoning with OpenAI’s Eric Mitchell and Brandon McKinzie

This week on No Priors, Elad and Sarah sit down with Eric Mitchell and Brandon McKinzie, two of the minds behind OpenAI’s O3 model. They discuss what makes O3 unique, including its focus on reasoning, the role of reinforcement learning, and how tool use enables more powerful interactions. The conversation explores the unification of model capabilities, what the next generation of human-AI interfaces could look like, and how models will continue to advance in the years ahead. Sign up for new podcasts every week. Email feedback to [email protected] Follow us on Twitter: @NoPriorsPod | @Saranormous | @EladGil | @mckbrando | @ericmitchellai Show Notes: 0:00 What is o3? 3:21 Reinforcement learning in o3 4:44 Unification of models 8:56 Why tool use helps test time scaling 11:10 Deep research 16:00 Future ways to interact with models 22:03 General purpose vs specialized models 25:30 Simulating AI interacting with the world 29:36 How will models advance?

Speaker names marked ? are inferred from speaking time and the show’s known hosts — they are not voice-verified.

• 0:05Sarah Guo?

  Hi listeners and welcome back to No Priors. Today, I'm speaking with Brandon McKinsey and Eric Mitchell, two of the minds behind OpenAI's O3 model. O3 is the latest in the line of reasoning models from OpenAI, super powerful with the ability to figure out what tools to use and then use them across multi-step tasks. We'll talk about how it was made, what's next, and how to reason about reasoning. Brandon and Eric, welcome to No Priors.

• 0:27Speaker 01

  Thanks for having us.

• 0:28Sarah Guo?

  Yeah, thanks for having us.

• 0:29Speaker 02

  Do you mind walking us through O3? What's different about it? What it was in terms of a breakthrough? In terms of like, you know, a focus on reasoning and you're adding memory and other things versus this is a core foundation model, LLM, and what that is?

• 0:41Speaker 03

  So, O3 is like our most recent model in this O-series line of models that are focused on thinking carefully before they respond. And these models are, in sort of some vaguely general sense, smarter than like models that don't think before they respond. You know, similarly to humans, it's easier to be, you know, more accurate if you think before you respond. I think the thing that is really exciting about O3 is that not only is it just smarter if you make like an apples-to-apples comparison to our previous O-series models, you know, it's just better at like giving you correct answers of math problems or factual questions about the world or whatever.

• 1:26Speaker 03

  This is true and it's great. And we, you know, will continue to train models that are smarter. But it's also very cool because it uses a lot of tools that enhance its ability to do things that are useful for you. So, yeah, like you can train a model that's really smart, but like if it can't browse the web and get up-to-date information, there's just a limitation on how much useful stuff that model can do for you. If the model can't actually write and execute code, there's just a limitation to how, you know, the sorts of things that an LLM can do efficiently, whereas like a relatively simple Python program can, you know, solve a particular problem very easily. So, not only is the model, it's on its own smarter than our previous O-series models, which is great, but it's also able to use all these tools that like

• 2:10Speaker 03

  Further enhance its abilities and whether that's doing like research on something where you want up to date information, or you want the model to do some data analysis for you, or you want the model to be able to do the data analysis and then kind of review the results and adjust course as it sees fit. Instead of you having to be so sort of prescriptive about like each step along the way, the model is sort of able to take these like high level requests, like do some due diligence on this company and, you know, maybe run some reasonable like forecasting models on so and so thing. And then, you know, write a summary for me, like the model will kind of like infer a reasonable set of actions to do on its own. So it gives you kind of like a higher level interface to to doing some of these more complicated tasks.

• 2:49Speaker 02

  That makes sense. So it sounds like basically there's like a few different changes between your core sort of GPT models where now you have something that takes a pause to think about something. So at inference time, you know, there's more compute happening and then also it can do sequential steps because it can kind of infer what are those steps and then go act on them. How did you build or train this differently from just a core foundation model or, you know, when you did, when you all did GPT two point five and four and all the various models that have come over time? What is different in terms of how you actually construct one of these?

• 3:20Speaker 01

  I guess the short answer is reinforcement learning is the biggest one. So yeah, rather than just having to predict the next token in some large pre-training corpus from, you know, you know, everywhere essentially. Now we have a more focused goal of the model solving very difficult tasks and taking as long as it needs to do to figure out the answers to those problems. Something that's like kind of magical from a user experience for me was we've in the past for our reasoning models talked a lot about test time scaling. And I think for a lot of problems, you know, without tools, test time scaling might occasionally work and but at some point the model is just kind of ranting in its in its internal chain of thought.

• 4:01Speaker 01

  And especially for like some visual perception ones, it knows that it doesn't, it's not able to see the thing that it needs and it just it just kind of like loses its mind and goes insane. And I think. Tool use is a really important component now to continuing this like test time scaling. And you can feel this when you're talking to O3, at least my, my impression when I first started using it was the longer it thinks, like I really get the impression that like I'm going to get a better result and you can kind of watch it do really intuitive things and it, it's, it's a very different experience, but being able to kind of trust that as you're waiting, like it's worth the wait and you're going to get a better result because of it and the model's not just off doing some, you know, totally irrelevant thing.

• 4:41Speaker 02

  That's cool. I think in your original post about this too, you all had a graph which basically showed that you, you looked at how long it thought versus the accuracy of the result and it was a really nice relationship. So clearly, you know, thinking more deeply about something really matters and it seems like in the long run, do you think there's just going to be a world where we have so sort of a split or bifurcation between models which are sort of fast, cheap, efficient, get certain basic tasks done. And then there's another model which you upload a legal M and A folder and it takes a day to think. And it's slow and expensive, but then it produces, you know, output that would take you a team of people. You know, a month to produce. Or how do you think about the world in terms of how, how all this is evolving or where it's heading?

• 5:24Speaker 03

  You know, I, I think for us, like unification of our models is something that, you know, Sam has talked about publicly that, you know, we have this big crazy model switcher in chat GPT and there are a lot of choices and, you know, we have

• 5:39Speaker 03

  a model that might be good at any particular thing, you know, that a user might want to do, but that's not that helpful if it's not easy for the user to figure out, well, which model should I use for that task? And so, yeah, making the models better able, you know, making this experience more intuitive is definitely something that is, is like valuable and, and something we're interested in doing and that, that applies to this, you know. Uh, question of like, uh, you know, are we gonna have like two models that, you know, people pick between or a zillion models that people pick between? Or do we put that decision, you know, inside the model? Um, I, you know, I think everyone is gonna try stuff and figure out what works well for like

• 6:23Speaker 03

  the problems they're interested in and like the users that they they have. But, um, but yeah, I mean, that that question of like, how do you, you know, make that, uh, that sort of decision be like as, you know, effective, accurate, like intuitive as possible is is definitely top of mind.

• 6:39Sarah Guo?

  Is there a reason from a research perspective to combine reasoning with pre-training? Try to, um, have more control of this? Because if you just think about it from the product perspective of like the end consumer dealing with chat GPT, like, you know, we won't get into the the naming nonsense here, but they don't care. They just want like the right answer and the amount of intelligence required to get there in as little time as possible, right?

• 7:05Speaker 01

  The ideal situation is it's like intuitive, uh, that like how long should you have to wait? You should have to wait as long as it takes for the model to like give you a correct answer. And I, uh, I, I hope we can get to a place where our models have a more precise understanding of their own level of uncertainty. Because, you know, if they, if they already know the answer, they should just kind of tell you it. And if, uh, if it takes them a day to actually figure it out, then they should, they should take a day. But you should always have a sense of like, uh, it takes exactly as long as, as it, as it needs to for that current like model's intelligence. And I, I, I feel like we're on the, the right path for that.

• 7:40Sarah Guo?

  Yeah, I wonder, um, if there isn't a bifurcation though, between like an end user product and the developer product, right? Because there are lots of companies that use, you know, um, the APIs to all of these different models and then for very specific tasks. And then on, on some of them, they might even use like open source models with really cheap inference with stuff that they control more.

• 8:00Speaker 01

  I hope you could just kind of tell the model like, hey, this is a API use case. And, uh, yeah, you really can't be over there thinking for like 10 minutes. Uh, we gotta get an answer to the user. It'd be great if their models kind of get to be more steerable, uh, like, like that as well.

• 8:15Speaker 03

  Yeah, I think it's just a general sterability question. Like, at the end of the day, if the model's smart, like you should be able to specify, like, the context of your problem and the model should.

• 8:27Speaker 03

  Do the right thing. There's going to be some limitations because, you know, maybe just figuring out, given your situation, what is like the right thing to do might require thinking in and of itself to figure out. So like, it's not that you can obviously do this perfectly, but, but yeah, pushing, you know, some, all the right parts of this into the model to make things easier for the user is like, seems is a very good goal.

• 8:52Sarah Guo?

  Can I go back to something else you said? Like, so the first guest we ever had on the podcast was actually Noam Brown.

• 8:58Speaker 01

  Oh, nice.

• 8:59Sarah Guo?

  So.

• 9:00Speaker 01

  I've heard of him.

• 9:01Sarah Guo?

  You know.

• 9:02Speaker 01

  I know.

• 9:03Sarah Guo?

  Two plus years ago. Yes. Hi, Noam. It'd be great to get some intuition from you guys for why tool use helps like test time scaling work much better.

• 9:10Speaker 01

  I can give maybe very concrete cases for like the visual reasoning side of things. The, there's a lot of cases where, and back to also the model being able to estimate its own uncertainty, you'll, you'll give it some kind of question about an image and the model will very transparently tell you what you're saying. If I like, I, I don't know, I can't really see the thing you're talking about very well. Or like, it, it almost knows like that its vision is not very good. And, but what's kind of magical is like when you give it access to a tool, it's like, okay, well, I gotta figure something out. Let's see if I can like manipulate the image or crop around here or something like this. And what that means is that it's, it's, it's like much more productive use of tokens as it's doing that. And so your test time scaling slope, you know, goes from something

• 9:52Speaker 01

  like this to, you know, something much deeper. And we've seen exactly that, like the, the test time scaling slopes for without tool use and with tool use for, for visual reasoning specifically are very noticeably different.

• 10:04Speaker 03

  Yeah. I also say like for like writing code for something like, there are a lot of things that an LLM could try to figure out on its own, but would require a lot of attempts and self verification that you could write a very simple program to do in like a verifiable and, and, you know, much faster way. So, you know, I do some research on this company and like use this type of, you know, valuation model to tell me like, you know, what the valuation should be. Like, you could have the model like try to crank through that and like fit those coefficients or whatever in its context. Or you could literally just have it like write the code to just do it the right way.

• 10:49Speaker 03

  And just know what the actual answer is. And so. Yeah, I think like part of this is you can just allocate compute a lot more efficiently because you can. Defer stuff that the model doesn't have comparative advantage to doing to a tool that is like really well suited to doing that thing.

• 11:04Speaker 02

  One of the ways I've been using. Some form of O three a lot is deep research, right? I think. That's basically a research analyst. AI that you all have built that basically. We'll go out, we'll look up things on the web, we'll synthesize information, we'll chart things for you. It's it's pretty amazing in terms of its capability set. Did you have to do anything special in terms of? You know, any form of specific reinforcement learning specifically for it to be better at that or other things that you built against it? Or how did you think about? The data training for it, the data that was used for training it like I'm just sort of curious like. How that product, if it all is a branch off of. Off of this and how you thought about building that specifically as part of this broader effort?

• 11:46Speaker 03

  I think when we. Think about like. Tool use. I think browsing is one of the most like natural. Places where, you know, you you think of as a starting point of like. Okay, like. And it's it's not always easy. I mean, like the, you know, initial kind of browsing that we. Included in GPT four a few years back, like it was hard to make it, you know, work in a way that felt like reliable and like useful.

• 12:15Speaker 03

  But, you know, in the sort of. You know, modern these days.

• 12:21Speaker 03

  Last year. You know, two years ago is ancient history.

• 12:26Speaker 03

  I think it feels like a natural place to start because it's like so widely applicable to so many types of queries like. Anything that is, you know, requires up to date information like it should help to browse for. And so. In terms of like a test bed for, hey, like. Does, you know, the way we're doing RL, like, does it really work? Or, you know, can we really get the model to learn like? Longer time horizon kind of meaningful extended behaviors like it it feels like kind of a natural. Place to start in some ways in that it, you know, also is. Fairly likely to be like useful in a relatively short amount of time. So it's like, yeah, let's try that. I mean, you know, in RL, like at the end of the day, you're defining an objective.

• 13:10Speaker 03

  And if you have an idea for like, who is going to find this most useful? Like, you know, you you might like want to tailor your the objective, you know, to who you expect to be using the thing, what you expect they're going to want, you know, what is their tolerance for? Do they want to sit through a 30 minute rollout of deep research? You know, do they? When they ask for a report, you know, do they want a page or five pages or a gazillion pages? So, um, yeah, I mean, you're you're definitely, you know, you want to tailor things to like who you think is going to be using it.

• 13:46Speaker 02

  I feel like there's a lot of almost like white collar behavior work that you or knowledge work that you all are really capturing through this sort of tooling going forward. And you mentioned software engineering is one potential area. Deep research and sort of analytical jobs is another where there's all sorts of really interesting work to be done. That's super helpful in terms of augmenting what people are doing. Are there two or three other areas that you think are the most near term, interesting applications for this? Whether OpenAI is doing it or others should do it aside. I'm just sort of curious how you think about the big application areas for this sort of technology.

• 14:15Speaker 01

  I guess my, you know, very biased one that I'm excited about is is coding and also research in general. Being able to like improve upon the velocity that we can do research at OpenAI and others can do research when they're using our tools. I think our models are getting a lot, a lot better very quickly at at being actually useful. And it seems like they're kind of reaching some kind of inflection point where they they they are useful enough to want to reach out to and and use like multiple times a day for me, for me at least, which wasn't the case. They're always like a little bit, you know, behind what I wanted them to be, especially when it comes to like navigating and using our internal code base, which is not simple.

• 14:57Speaker 01

  And it's amazing to see like more recent our our our models actually really spending a lot of time trying to understand the questions that we ask them. And coming back with things that saved me like many hours of of my own time.

• 15:12Speaker 02

  People say that's the fastest potential bootstrap, right? In terms of each model subsequently having helping to make the next model better, faster, cheaper, et cetera. And so people often argue that that's almost like a inflection point on the exponent towards superintelligence. Is basically this ability to use AI to build the next version of AI.

• 15:33Speaker 01

  Yeah, and there's so many like different components of research, too. There's it's it's not just, you know, sitting off in the ivory tower, thinking about things. But there's there's like hardware, there's, you know, various components of training and evaluation and stuff like this. And each of these can be turned into some kind of task that can be optimized and iterated over. So there's plenty of, you know, room to to squeeze out improvements.

• 15:55Sarah Guo?

  We talked about browsing and writing code, arguably the greatest tool of all, right? Especially if you're trying to figure out how to spend your compute, write more efficient code. Generating images, writing text, there are certainly like trajectories of action I think are not in there yet. Right, like reliably using a sequence of business software.

• 16:14Speaker 01

  I'm really excited about the computer use stuff. It kind of drives me crazy in some sense that our models are not already just like on my computer all day, watching what I'm doing. Well, I know that could be creepy for some people and like, I think you should be able to like opt out of that or have that opt out by default. I hate typing also. I I wish that I could just kind of like be working on something on my computer. I hit some issue and I'm just like, you know, what am I supposed to do with this? And I can just kind of ask. I think there's tons of space for, yeah, being able to improve on like how we interact with the models. And this goes back to them being able to use tools in a more intuitive way. I guess using tools closer to how we use them. It's also surprising to me how intuitively our models do use the tools we give them access to.

• 16:55Speaker 01

  It's like weirdly human like, but I guess that's not too surprising, given the data they've seen before, but yeah.

• 17:00Sarah Guo?

  I think a lot of things are weirdly human like, like my intuition for like, well, why is tool use so impactful to test time scale? Like, why is the combination so much better? Take any and any role, you can make a decision when you are trying to make progress against a task as to like, do I get external validation or do I sit and think really hard? Right, and usually you want to do like one or is more efficient than the other. And it's not always just sit in a vacuum and think really hard with what you know.

• 17:26Speaker 01

  Yeah, absolutely.

• 17:26Speaker 03

  You can seek out sort of new inputs, like it doesn't have to be this closed system anymore. And I do feel like the the closed system-ness of the models is still sort of a limitation in some ways. Like, you're not, you're not necessarily like turning this. I mean, like, I think it'd be great if the model could control my computer for sure, but in some sense, it's. There's a reason we don't go hog wild and say like, oh yes, here's like the keys to the kingdom. Like have at it. Um, there are still, you know, asymmetric costs to like the time you can save and the types of errors you can make. And so we're trying to like iteratively kind of, you know, deploy these things and like try them out and figure out like, where are they reliable?

• 18:07Speaker 03

  You know, and where are they not? Um, because yeah, like if you did just let the model control your computer, it could do some cool stuff. Like I have no doubt, but you know, do I trust it to like respond to all of the, you know, random emails that Brandon sends me? Actually, maybe for that task, it doesn't require that much intelligence, but you know, more generally like, do I, you know, do I trust it to, to do everything I'm, I'm doing? Like, you know, some things, and I'm sure like that set of things will be bigger tomorrow than it was yesterday. Um, but yeah, I think part of this is like, we limit the affordances and keep it a little bit in the like sandbox, just out of caution. Um, so that, you know,

• 18:50Speaker 03

  you don't send some crazy email to your boss or, um, you know, delete all your texts or delete your hard drive or something.

• 18:57Sarah Guo?

  Is there some sort of like, um, organizing mental model for like the tasks that one can do with, uh, you know, increasing intelligence, test time scaling and tool, improve tool use, right? Because I, I look at this and I'm like, okay, well, you have complexity of tasks and you have time scale. Um, then you have like the ability to come up with these RL rewards and environments, right? Then you have like usefulness. Um, uh, you, maybe you have some, you, of course, you have some intuition about like diversity and generalization across the different things you can be doing, but it, it's a, it seems like a very large space and scaling our, like new gen RL is not, it's just not obvious. Like how, to me, it's not obvious how you do it or how you

• 19:41Sarah Guo?

  choose the path. Is there some sort of organizing framework that, you know, you guys have that you can share?

• 19:46Speaker 03

  I mean, I don't know if there's like one organizing framework. I think there are a few like factors at least that I think about in like the very, very grand scheme of things is like, How much, like in order to solve this task, like how much uncertainty with the environment do I have to like wrestle with? Like, um, for some things where it's like, this is a purely fat, like, who was the first president of the United States? Like, there's zero, like, environment I need to interact with to like, reach the answer to this question correctly. I just need to remember the answer and say the answer. You know, if I want you to like, write some code, you know, that like, solves some problem, well, now I have to deal with a little bit of like, not purely internal model stuff,

• 20:28Speaker 03

  but also like, okay, I need to execute the code and like, that code execution environment is maybe more complicated than my model can memorize internally. So I have to do like, a little bit of like, writing code and then executing it and making sure it does what I thought it did, and then testing it, and then giving it to the user. And things get like, the, the amount of that sort of stuff outside the model that you have to like, you know, you can't just recall the answer and give it to the user, you have to like, test something and, you know, run an experiment in the world and then wait for the results of that experiment. Like, the more you have to do that, the more uncertain the results of those experiments. Like, in some sense, that's like, one of the core, like, attributes of like, what makes the tasks hard.

• 21:09Speaker 03

  Um, and I think another is like, how, you know, simulatable they are. Um, like, stuff that is really bottlenecked by like, time, like the physical world, um, is also, you know, just, just harder than stuff that we can simulate really well. You know, it's not, it's not a coincidence that, you know, so many people are interested in coding and, you know, coding agents and things. Um, and that like, you know, robotics is hard and it, you know, it's, it's slower and, you know, I used to work on robotics and like, it's frustrating in a lot of ways. I think both this like, how much of the external environment do you have to deal with? And then like, how much do you have to wrestle with the unavoidable slowness of the real world are two like, dimensions that I, I sort of think about.

• 21:53Speaker 02

  It's super interesting because if you look at, um, historically, some of these models, one of the things that I think has continued to be really impressive is the degree to which they're generalizable. And so I think when GitHub Copilot launched, it was on Codex, which was like a specialized code, code model. And then eventually that just got subsumed into these more general purpose models in terms of what a lot of people are actually using for, um, coding related applications. How do you think about that in the context of things like robotics? So do you, you know, there's like, probably a dozen different robotics foundation model companies now. Do you think that eventually just merges into the work you're doing in terms of there's just these big general purpose models that can do all sorts of things? Or do you think there's a lot of room for these standalone other types of models over time?

• 22:31Speaker 01

  I will say the one thing that's always struck me as kind of funny about us doing RL is that we don't yet do it on the most like canonical RL task of robotics. And I personally don't see any reason why we couldn't have these be the same model. I think there are certain challenges with like, I don't know, do you want your RL model to be able to like generate an hour long movie for you natively as opposed to like a tool call? Like that's where it's probably tricky to have you have more conflict between having like everything in the same set of weights. But certainly like the things you see O3 already doing in terms of like, you know,

• 23:13Speaker 01

  exploring a picture and things like that are are kind of like early signs of something like an agent exploring like an external environment. So I don't think it sounds too far fetched to me.

• 23:21Speaker 03

  Yeah, I mean, I think the thing that came up earlier of the also the like intelligence per cost thing, you know, the the real world is like an interesting litmus test because at the end of the day, like there is a, you know, frame rate in the real world you need to live on. And it doesn't matter if you get the right answer after you think for two minutes, like, you know, the ball is coming at you now and you have to catch it. Gravity's not going to wait for you. So you, you, that's an extra constraint that we get to at least softly ignore when we're talking about these purely disembodied things. That's kind of.

• 23:57Speaker 02

  It's kind of interesting though, because really small brains are very good at that. You know, so you look at a frog, you know, you start looking at different organisms and you look at sort of relative compute.

• 24:06Speaker 03

  Yeah.

• 24:07Speaker 02

  And, you know, very simple systems are very good at that, ants, you know, like, so I think that's kind of a fascinating question in terms of what's the baseline amount of capability that's actually needed for some of these real world tasks that are reasonably responsive. And nature.

• 24:23Speaker 01

  It's really tricky with with vision too. There we have so our models have some, I think, maybe famous like edge cases of where they don't do the right thing. I think Eric probably knows where I'm going with this. Like, I don't know if you ever asked, like, our models to tell you what time it is on a clock. They they really like the time 10:10. So, yeah.

• 24:42Speaker 02

  It's my favorite time, too. So that's that's usually what I tell people.

• 24:44Speaker 01

  It's like over 90% or something like that of all clocks on the Internet are 10:10. And it's because it looks like, I guess, like a happy face. It looks like nice. And but anyways, like, the what I'm getting at is like, our our visual system was developed by interacting with, you know, the external world and having to be good at like navigating things, you know, avoiding predators. And our our models have learned vision a very different type of way. And I I think it'll we'll see like a lot of really interesting things if we can get them to be kind of closing the loop by, you know, reducing their uncertainty by taking actions in the real world, just as opposed to like, thinking about stuff.

• 25:21Sarah Guo?

  Hey, Eric, you brought up the idea of like, how what in the environment can be simulated, right, as a as an input into like, how difficult will it be to improve on this? As you get to long running tasks, like, let's just take software engineering, like, there is a lot of interaction that is not just me committing code continually. It's like, I'm going to talk to other people about the project, in which case you then need to deal with the problem of like, can you reasonably simulate how other people are going to interact with you on the project in an environment? That seems really tricky, right? I'm not saying that, you know, oh, three or whatever set of foundation models now doesn't have the

• 26:04Sarah Guo?

  intelligence to respond reasonably. But like, how do you think about that simulation being true to life as a true to life, true to the real world as you? Involve human beings in an environment in theory.

• 26:17Speaker 01

  My spicy, I guess, take on that is like, I don't know if it's spicy, but 03 in some sense is already kind of simulating what it'd be like for a single person to do something with like a browser or something like that. And I don't know, train two of them together. So that you have, you know, you have two people interacting with each other. And there's no reason you can't scale this up so that models are trained to be really good at cooperating with each other. I mean, there's a lot of already existing literature on multi-agent RL. And yeah, if you want the model to be good at something like collaborating with a bunch of people, like maybe a not-too-bad starting point is making it good with collaborating with other models.

• 26:57Speaker 03

  Man, someone should do that.

• 26:58Speaker 01

  Yeah, yeah, we should really start thinking about that, Eric.

• 27:01Sarah Guo?

  I think it is, I think it's a little bit spicy because yes, the work is going on. It is interesting to hear you think that is a useful direction. I think lots of people would still like to believe, not me, like my comment was extra good on this pull request or whatever it is, right?

• 27:18Speaker 01

  Okay, I could, I could sympathize with that. Sometimes I see our models training and I'm like, oh, what are you doing? You know, like you're taking forever to figure this out. And I actually think it would be really fun if you could actually train models in an interactive way. You know, forget about just like a test time, but I think it'd be really neat to train them to do something like that. Be able to like intervene when it makes sense. And yeah, just more, more me being able to tell the model to cut it out and like in the middle of its kind of chain of thought. And it being able to learn from that on the fly, I think would be great.

• 27:46Speaker 03

  Yeah, I do think this is like the intersection of these two things where it's both like a point of contact with the external environment that is like, can be very high uncertainty. Like humans can be very unpredictable in some cases. And it's sort of limited by the tick of time in the real world. If you want to like, you know, deal with actual humans, like humans have a fixed, you know, clock cycle. You know, in their, in their head.

• 28:16Speaker 03

  So yeah, I mean, this is if you, you know, if you want to like do this in the literal sense, it's hard. And so, you know, scaling it up and, and, you know, making it work well is, is, you know, it's not obvious how to do this.

• 28:25Speaker 01

  Yeah, we are a super expensive tool call. You know, if you're a model, you can either ask me, you know, meatbag over here to, you know, help with something and I'll try to think really slowly. In the meantime, it could have like used browser and read like 100 papers on the topic and something like that. So it's, yeah, how do you model the trade off there?

• 28:42Speaker 03

  But the human part is important. I mean, I think in any research project, like my interactions with Brandon are the hardest part of the project, you know, like writing the code is, that's the easy part.

• 28:50Sarah Guo?

  Well, and there's, there's some analog from self-driving. A lot's gonna say the, you know, hanging out with me every week is the hardest part of doing this podcast, but.

• 28:58Speaker 01

  It's my favorite part. Look at how healthy their relationship is, Eric. We need to learn from this.

• 29:02Speaker 03

  No, we're honest. It's okay. We gotta work through it.

• 29:07Sarah Guo?

  In self-driving, one of the like classically hard things to do was like predict the human and the child and the dog, like agents in the environment versus like what the environment was. And, and so I, I think there's like some analogy to be drawn there. Going back to just like how you progress the O series of models from here, is it, is it a reasonable like assessment that some people have that the capabilities of the models are likely to advance in a spikier way? Because you're relying to some degree more on the creativity of research teams and like making these environments and deciding, you know, how to

• 29:49Sarah Guo?

  create these evals versus like we're scaling up on existing data set and pre-training? Is that a fair contrast?

• 29:59Speaker 01

  Spikier, like, what's the plot here? What's the like the X axis and the Y?

• 30:03Speaker 03

  Domain is the X axis and Y is capability?

• 30:06Sarah Guo?

  Yes, because you're like choosing what domains you are really creating this RL loop in.

• 30:13Speaker 03

  I mean, I think this is a very reasonable hypothesis to, to hold. I think there is some like counter evidence that I think should, you know, be factored into people's intuitions. Like, you know, Sam tweeted an example of some creative writing from one of our models that I think was.

• 30:38Speaker 03

  I'm not an expert and I'm not going to say this is like, you know, publishable or like groundbreaking, but I think it probably updated some people's intuitions on like what, you know, you can train a model to do really well. And so I think there is some structural reasons why you'll have some spikiness just because like as an organization, you have to decide like, hey, we're going to prioritize, you know, X, Y, Z stuff. And like as the models get better, the surface area of stuff you could do with them grows faster than, you know, you can potentially like say, hey, this is the niche, you know, we're going to carve out, we're going to try to do this really well. So like there, I think there's some reason for spikiness, but I think some people will

• 31:20Speaker 03

  probably go too far with this and saying like, oh yes, these models will only be really good at math and code and like not, you know, like everything else is like, you can't get better at them. And I, I think that is probably not the right intuition to have.

• 31:34Speaker 01

  Yeah. And I think probably all like major AI labs right now have some partitioning between, let's just define a bunch of data distributions. We want our models to be good at and then just like throw data at them. And then another set of people in the same companies is probably, are probably thinking about how can you kind of lift all boats at once with some like algorithmic change. And I, I think, yeah, we definitely have both of those types of efforts at, at OpenAI and

• 32:05Speaker 01

  I think especially on the data side, like there are going to naturally be things that we have a lot more data of than, than others and but ideally, yeah, we, we have plenty of efforts that will not be so reliant on the exact like subset of data we did RL on and it'll generalize better.

• 32:22Sarah Guo?

  I get pitched every week and I bet a lot does too. A company that wants to generate data for the labs in some way. And, or it's, you know, access to human experts or whatever it is, but like, you know, there's, there's infinite variations of this.

• 32:39Sarah Guo?

  If you could wave a magic wand and have like a perfect set of data, like what would it be?

• 32:46Sarah Guo?

  That you know would advance model quality today.

• 32:48Speaker 03

  This is a dodge, but like uncontaminated evals.

• 32:54Speaker 03

  Always super valuable and that's data.

• 32:58Speaker 03

  And I mean, yeah, like you want, you know, good data to train on and that's of course valuable for making the model better. But I think it is often neglected how also important it is to have high quality data, which is like a different definition of high quality when it comes to an eval.

• 33:17Speaker 03

  But yeah, the eval side is like often just as important because you don't, you need to measure stuff and like, as you know, from, you know, trying to hire people or whatever, like evaluating the capabilities of like a general like capable agent is really hard to do in like a rigorous, you know, way. So yeah, I think evals are a little underappreciated.

• 33:38Speaker 01

  But it's true evals are, I mean, especially with some of our recent models where we've kind of run out of reliable evals track because they kind of just solved a few of those. But on the, on the training side, I think it's always valuable to have

• 33:54Speaker 01

  training data that is kind of at the next frontier of model capabilities. I mean, I think a lot of the things that O3 and O4 mini could already, can already do, those types of tasks, like basic tool use, we probably aren't,

• 34:08Speaker 01

  you know, super in the need for, for new data like, like that. But I think it'd be hard to say no to a data set that's like bunch of like multi-turn user interactions and some code base that's like a million lines of code that, you know, is like a two-week research task of like adding some new feature to it that requires like multiple poll requests. I mean, I mean, like something that was like super high quality and has a ton of supervision signals for us to learn from it. Yeah, that's, I think that would be awesome to have, you know, I definitely wouldn't turn that down.

• 34:39Sarah Guo?

  You play with the models all the time, I assume a lot more than average humans do. What do you do with reasoning models that you think other people don't do enough of yet?

• 34:47Speaker 03

  Send the same prompt many, many, many times to the model and, and get an intuition for the distribution of responses you can get. I have seen, it drives me absolutely.

• 35:02Speaker 03

  Mad when people do these comparisons on Twitter or wherever and they're like, "Oh, I put the same prompt into blah, blah and blah, blah and this one was so much better." It's like, dude, you like, like, I mean, was something we talked about a bit when we were launching. It's like, yeah, O3 can do really cool things, like when it chains together a lot of tool calls. And then like sometimes for the same prompt, it won't have that, you know, moment of magic or it will, you know, just take a little, it'll do a little less work for you. And so, yeah, the like, the peak performance is really impressive, but there is a distribution of behavior. And I think people often don't appreciate that there is this distribution of outcomes

• 35:46Speaker 03

  when you put the same prompt in and getting intuition about that is useful.

• 35:50Sarah Guo?

  So as an end user, I do this and I also have a feature request for your friends in the product org. I'll ask, you know, Oliver or something, but it's just, I want a button where like, assuming my rate limits or whatever support it, I want to run the prompt automatically, like 100 times every time, even if it's really expensive. And then I want the model to rank them and just give me the top one or two.

• 36:14Speaker 03

  Interesting.

• 36:15Sarah Guo?

  And just let it be expensive.

• 36:16Speaker 02

  Or a synthesis across it, right? You could also synthesize the output and just see if there's some, although maybe you're then reverting to the mean in some sense relative to that distribution or something. But it seems kind of interesting, yeah.

• 36:27Sarah Guo?

  Maybe there's a good infrastructure reason you guys aren't giving us that button.

• 36:30Speaker 01

  Well, it's expensive, but I think it's a great suggestion, yeah.

• 36:35Speaker 03

  Yeah, I think it's a great suggestion.

• 36:36Speaker 01

  How much would you pay for that?

• 36:38Sarah Guo?

  A lot, but I'm a price insensitive user of AI.

• 36:41Speaker 01

  I see.

• 36:41Speaker 03

  Perfect.

• 36:42Sarah Guo?

  Maybe there are many of us.

• 36:45Speaker 02

  Should have Sarah tier as one of your tiers.

• 36:47Speaker 03

  Exactly, exactly.

• 36:51Speaker 01

  I really like. Sending prompts to our models that are kind of at the edge of what I expect them to be able to do, just kind of for funsies. A lot of times before I'm about to do some like programming tasks, I will just kind of ask the model to go see if it can figure it out. A lot of times with like no hope of it being able to do it. And indeed, sometimes it comes back and I just am pretty like, I'm like a disappointed father. But other times, it does it and it's amazing and it saves me like tons of time. So I kind of use our models almost like a background cue of work, where I just, I'll just like shoot off tasks to them and sometimes those will, those will stick and sometimes they won't. But in either case, like it's always a good outcome if something good happens.

• 37:31Speaker 02

  That's cool. Yeah, I do that just to feel better about myself when it doesn't work.

• 37:35Speaker 01

  Yeah, I'm still providing value.

• 37:40Speaker 02

  When it works, I feel even worse about myself. So it's very hit or miss.

• 37:44Speaker 01

  Yeah.

• 37:44Sarah Guo?

  This has been great, guys. Thank you.

• 37:46Speaker 02

  Yeah, thanks so much for coming on.

• 37:47Speaker 01

  Yeah, thanks. It was fun. Thanks for having us.

• 37:50Sarah Guo?

  Find us on Twitter at NoPriorsPod. Subscribe to our YouTube channel if you want to see our faces. Follow the show on Apple Podcasts, Spotify, or wherever you listen. That way you get a new episode every week. And sign up for emails or find transcripts for every episode at no-priors.com.