00:57:14 · Mario Gabriele, Karol Hausman
[00:57:14] S1: Do you have theses at this point on what gives you the the sort of steepest rate of learning? Whether that's, you know, I don't know, very fine tuned tasks or or something totally different? [00:57:23] S0: I think there is there is a few of those things that we that we learn. We know that the diversity of data is really, really important. We know that the quality of data is really important. We know that closing the loop with the models are very important. I just feel like these are fairly broad statements that, you know, if you had asked me this a year ago, I would probably say something similar. But we did learn a lot about what these terms actually mean. I think, you know, very often people think about the diversity of data or quality of data or or something like that. But we don't really have very good definitions of what quality of data actually is, what does it mean, or what diversity of data actually means, or how do you measure it. So I think these are like fairly deep questions and we are are getting more and more of a of a feel of what that actually is and how we can measure it and how we can optimize for it. But that's a constant push and pull on how much do you scale the existing thing, how much can you better understand it, how much can you increase the the slope of improvement there. So I think we'll just continue on the trajectory as we scale it.
00:58:32 · Mario Gabriele, Karol Hausman
[00:58:32] S1: You mentioned that, you know, there's still a need for real research in this space to to achieve the ultimate end goal, but that there's also plenty to optimize. You know, in the sort of thought experiment where we get no new research, how far do you think that that sort of takes us? Do you have some rough heuristics of what kind of a robot that that is able to produce? It's a question that we think about a lot. [00:58:57] S0: Because I think one powerful insight from the language model work is that for a long time, people thought that, you know, we still have a few ideas that are missing or we need a different architecture or we need different this or that. But it turned out that it was good enough. It was just a matter of scaling, and we just didn't have enough foresight to predict how scaling is gonna resolve some of the issues that seem like big issues of of today at small scale. So I think there is a nontrivial chance that the recipe is already there. That the recipe that we have today would work and would just solve everything. And we just need to scale scale it in the right way and execute on it really, really well. Well, I'm not certain about this at all, but I think there is... That's one question that is that is an important question that we're thinking about quite a bit. So what we're trying to do is is verify it. So start scaling the the existing recipe and see how it how it performs, how it scales, and then the same time continue to do research that, at the very least, could improve the the slope of that scale.
00:58:57 · Karol Hausman, Mario Gabriele
[00:58:57] S0: Because I think one powerful insight from the language model work is that for a long time, people thought that, you know, we still have a few ideas that are missing or we need a different architecture or we need different this or that. But it turned out that it was good enough. It was just a matter of scaling, and we just didn't have enough foresight to predict how scaling is gonna resolve some of the issues that seem like big issues of of today at small scale. So I think there is a nontrivial chance that the recipe is already there. That the recipe that we have today would work and would just solve everything. And we just need to scale scale it in the right way and execute on it really, really well. Well, I'm not certain about this at all, but I think there is... That's one question that is that is an important question that we're thinking about quite a bit. So what we're trying to do is is verify it. So start scaling the the existing recipe and see how it how it performs, how it scales, and then the same time continue to do research that, at the very least, could improve the the slope of that scale. [01:00:01] S1: NVIDIA and, you know, is probably the best example of a company that is putting a ton of money into improving its sort of simulation engines and getting that to higher and higher fidelity. How optimistic are you about that improving at a sufficiently high rate that it really, I don't know, closes the gap in some meaningful way or allows you to to get the kind of data at scale that could improve things meaningfully? So as I mentioned before, I'm
01:00:01 · Mario Gabriele, Mario Gabriele
[01:00:01] S1: NVIDIA and, you know, is probably the best example of a company that is putting a ton of money into improving its sort of simulation engines and getting that to higher and higher fidelity. How optimistic are you about that improving at a sufficiently high rate that it really, I don't know, closes the gap in some meaningful way or allows you to to get the kind of data at scale that could improve things meaningfully? So as I mentioned before, I'm [01:00:29] S2: I'm quite open minded about these things. I'm not
01:00:29 · Mario Gabriele, Karol Hausman
[01:00:29] S2: I'm quite open minded about these things. I'm not [01:00:31] S0: very pragmatic about the path that's gonna it's gonna take to get there. I think simulation over the past, you know, a decade plus has been improving like crazy. Simulation that we see today is much more realistic than what we've seen in the past and it's becoming more and more scalable. I believe that the first place where we're gonna see the impact of it is not going to be as much data collection as it's going to be evaluation. And one thing we see here at at at Physical Intelligence is that as these models become more powerful, it takes more and more time to evaluate them. So now not only you need to evaluate them for longer, you need to take more samples to distinguish between the model being, you know, at 99% success rate versus 95% rate, the success rate, and it was between, you know, 50 versus 70%. You just need more samples to be statistically significant, but the repertoire becomes broader and broader. So now you need to evaluate them on more tasks across more robots in more environments. And that trend, I think, will continue. As the models get get stronger, you will need to evaluate them for much longer and much more sophisticated ways. So I think if simulation starts to work in the way that we all hope it will, I think that's where it will... We'll see the the first impact of it, where we can start evaluating these models in simulation across diverse set of scenarios. And if they're realistic enough, that should correlate with with real world evaluations.
01:01:58 · Mario Gabriele
[01:01:58] S1: You know, we talked about some of the real inflection points for you over the course of your career. I'm sure that physical intelligence biggest inflection points are are to come. But over the past couple of years so far, what have been the... Maybe the biggest surprises or the the moments that have impressed you the most?