The equation, I think, for starting a robotic business has changed and will continue to change at an accelerating pace because the upfront cost is not that high anymore. Everyone's sort of spending a lot of time in the digital world and it feels like now is the time to start thinking about the world of atoms. You literally just gave people the playbook for how to build a vertical robotic company. This has really been our mission from the start is to create that Cambrian explosion. It still like blows my mind.
I didn't know if this would exist even in my entire lifetime. Welcome back to another episode of the Litecone. Today we have a very special guest, Juan Vong. He's one of the co-founders of physical intelligence, which we think might be the robotics AI lab that brings about the GPT-1 moment for all of robotics. Juan, thank you for joining us.
Pleasure to be here has been a long time admirer of YC and our mission is to build a model that can control any robot, to do any tasks that is physically capable of and to do so as such a high level of performance that can be useful to people in all walks of life. And so GPT-1 for robotics, what is it? You know, is the chat GPT moment for robotic real. Our perspective here is that we want to build a model that's really intelligence. We want to build a platform that allows us to externalize that intelligence to the rest of world and allow them to use it to build very interesting applications in all sorts of vertical and robotics.
And we think that it's going to be more like a peeling and audience analogy where you start from a really strong base model that have all sorts of common sense knowledge and already works to some extent on your robot. You have then a mixed autonomy system, very similar, for example, to a autonomous driving car today. And then you actually deploy the system to do a real job. That system might make me sick. It's okay.
And then over time by actually exposing the system to the complexity and the edge case of the real world, that system get incrementally, even just slightly better over time every day. And one day you wake up and you suddenly have a system that is just fully autonomous and just provide tremendous value. I'd be helpful to give the audience a bit of a mini history lesson on why robotics is so hard. And there's been a lot of breakthroughs in the last two years. And I mean, just to simplify the robotics problem is three pillars semantics, which I think we got a lot of analogs and with language models that somehow reported into robotics, then you have the planning.
And then the last thing is control, which needs to be done in real time and interact with the environment that changes. So I guess through the seminal papers that a lot of the team of Pyronbotics published that gave you the inkling that the GPT one moment is near. And that started in 2024. Yeah. So the dream to build general purpose robots like robots has been a long time dream.
I think in humanity, you know, we're not the first to say that our mission is to build a model that can work on any robot. And we're really fortunate to be in this moment in history where we feel that it's possible to kind of walk back a little bit a few years before. There was, I think the first is Saikhan, which to me was the first demonstration of language model and how you can bring all of the common sense knowledge in language model into robotics. And therefore, that significantly kind of reduces the need to collect robot specific data. So for example, if you have a task of, oh, I want to go to the YC office, the recorded podcast, what is that?
And you can ask a language model, you know, just show me the steps and show me the plan. And that worked incredibly well. And then the way you kind of language model, if you will, in robotics, is it start at the planning level at the semantic level. And then the control problem, you know, at the end of the day, you still need a mechanism to convert the plan into low level action that can actually actuate the robot. And that bring us to Palm E and that bring us to RT2, which stands for robotic transformer 2.
And what is to work really show is that if you start from a vision language model, that is really powerful and you kind of use robotic data to adapt this model to speak robot language, if you will, then you see a lot of transfer from the kind of knowledge that exists in the language, what the innovation language model down to the low level action. Again, one of my favorite examples when we did the RT2 project was you can have a picture of the library on the table, a picture of Taylor Swift, you have a picture of the Queen of England, and you can ask the robot, you know, pick up the co-can and move it to Taylor Swift, even though the concept of Taylor Swift doesn't exist in the robot data at all and that work, you can do other examples such as kind of spatial reasoning that doesn't exist in the robot data at all.
Like for example, move the dinosaur next to the red car and these are also completely unseen objects in robot data. And so that was RT2 and that was Palm E. Now RT2 and Palm E are single embodiment exercise. Just for the audience, single embodiment meaning you work for a very specific robot. It worked for a very specific robot.
In robot, you can ask the question, how do you scale, especially how do you scale data collections? One of the insights that we had back then was, you know, maybe the data from one robot is not that different from another robot's anyway. If you have enough robots in your training data, maybe what the model learn is into control one specific robot and what the model learn is something that's more abstract, which is, how do I kind of learn a general notion of what it means to control any particular robotic platform and therefore I will be better at controlling any particular platform. And that brings us to what we call open cross embodiment and robotic transformr X. That was a big paper because it was the first that showed potential scaling laws that apply to robotics because now you could start training all these models across multiple kinds of hardware, not just one, which has never been done in robotics ever before.
Because from all the research labs, they would all train with a very specific set of sensor actuators and motors and it was all very finicky with that particular hardware, right? Yeah. One of the really interesting results from open cross embodiment and let me provide the context here is that you can take, let's say, 10 different robot platform, collect data from them, train a policy and really optimize the policy to work well on that platform. So let's say you have that, you have 10 different platforms, 10 different policies. And now if you simply take the data and absorb it into a model that is high capacity enough to really absorb that data and you can compare, you have this generalist, right, that learn to control how to attend different robots, you can compare it to the specialist that has been optimized to work well on a particular embodiment.
How does it compare? And the interesting result from open X is it was 50% better. And that was really surprising because in robotics, it's hard enough to get your model to work on one particular robot platform. And one of the reasons why I say that we're really fortunate to be in this moment in robotics is because open X was really only possible because of the support that we received from the robot community. It was a huge collaboration across the robot community.
And the reason why that's really important is there is this joke in the robotic grad school that if you want to add two years to your PhD, just work on any robot platform. You know, by that logic, if you want to have 10 robot platforms, that's 20 years. Why is that? It takes like a year or two to just get the platform up and running to even collect the data. Yeah.
Is it fair to say that the data set that was created from embodiment X is similar to the scale of an impact that ImageNet did for vision because it was huge and it was the first large data set across multiple hardware, huge collaboration? I still think that ImageNet was more impactful in the vision community. And the reason for that is a few. The first is that ImageNet also allowed for reproducible evaluation, right? You know, OpenX as an effort was more about making data available for kind of people to use and evaluation is a really difficult problem in robotics that OpenX did not solve.
And the second is I think OpenX is a drop in the bucket at this point in the robotic community. If you measure in the kind of the scale and the volume and the diversity of data that the community is collecting, I think OpenX at this point is a drop in the bucket. I mean, I guess we started talking about sort of GP1, but even GP1, that was sort of this moment where you can prove, Alec Radford figured out that there was a neuron based on a very specific input and output and then that allowed the scaling loss to sort of take hold. The biggest problem in robotics I've heard is basically actually exactly what we've been talking about is the data problem. The language you could bootstrap off of the sum total of what you could get off the internet, which is actually quite a lot.
Can you give us a sense for scale? Is it petabytes? What do you think is necessary as an input to the true GP1 of robotics? Yeah, so the data scarcity problem in robotics is a few ways to look at it. The first way is that it's really two problem in this guy.
There is a generation data generation problem and there's data capture problem. The difference is that the data capture is that there might already be lots of robotic data that is being generated, but there's just never been really an incentive to capture it, to make it easy for the gestures in training. And that's one of the goal that OpenX was trying to solve, which is if you have robotic data, it's a really good idea to capture it and make it possible to train on. The second way to look at it is that robotic is very different from that. There is not an internet of robotic data that you can use.
So you see this kind of very operationally heavy effort to collect data. And that's the question of is it going to scale? Well, the way that I look at it is let's take the USGDP, 24 trillion US dollars. Let's say we actually solve robotics, a model that can control any robot to do any task. That can mask maybe contribute 10% to USGDP.
Well, that's already a massive number. And I think that promise is one of the reasons that warrants the investment into data collections in robotics. And the third way to look at it is we're very focused on cross-volume. In cross-volume, there is the data collection aspect of as well, which is to really make sure that your model and your organizations and infrastructure are set up to consume data from many different sources of robots. And that actually allowed you to scale easier.
For example, if I were to contrast our approach compared to, let's say, a company that have a particular hardware platform that they optimized for and they scale, it's not an approach that really allowed people to scale because it's just much harder to figure out how to manufacture a thousand units of something for now, compared to making sure that you yourself are ready to absorb data from a thousand different types of robots that are already in the community. I mean, it's a crazy problem, isn't it? The hardware itself, even within the same design of embodiment, if there's a hardware run that goes awry or one of the servos is slightly different, you see it in the data, right?
And then how do you control for that? Yeah, so I think we were doing an inventory of robots in the company. We were so shocked to find out that there are no two robot platforms that are the same. And if you ask people in the right community, sometimes there's debate about multi-robot versus single-robot. And the argument is that single-robot is simpler to scale.
And actually, that's not how it plays out in practice. It's even if you have a single robot that you're optimizing for, over time that platform is going to drift. You know, maybe you want to make hardware change or you have software change. You end up in a situation where it's much harder for you to reuse old data because, you know, in machine learning, if you want to generalize from a distribution, you would like many sample from a distribution. And if you just have one robot platform that have a major change every three months, maybe you have a few data points from that distribution.
Whereas if you start from the hypothesis that if you have many robot platform in your fleet, your model is going to learn something more abstract, which is how do I control a robot, not any particular robot, then the model will be able to ingest data from, you know, a slightly different robot better. And actually, we're starting to see immersion property in this kind of robot-large relation or what we're doing, where you start to see interesting transfer between different data sources. For example, today it's possible to perform tasks, zero shot, zero shot meaning you don't collect any data. And these are the tasks that last year might have required like hundreds and hundreds of hours. What are some examples?
Yeah, do we have any videos? We can see that like- So, you know, I get some flak when I come back because this is not published, hopefully this will come out soon. So I want to reserve the excitement for that and I'm building up the excitement a little bit. So hopefully this will come out soon. These are not simple tasks.
These are like actually difficult tasks that just last year require like hundreds of hours of data collections. You hear it here on like-home first that there's some emergent properties that are going to come out of pie. Can you give us a like the flavor of the tasks? It's really easy to fool yourself. And so we wanted to test across like, feel different tasks or different flavors.
A task that requires precision, tasks that require reasoning with multiple objects in the scene. It all seems to have this property. That's really nice. So it does seems like that's something that's kind of a more general property that emerge rather than we just, you know, got lucky and suddenly the model start working on one particular task. And so we're going to be able to understand where we are now in terms of like what's working and how well it's working.
Like we're not quite at the chat GPT moment yet. Like where are we? And I think you brought some videos that you're going to show us to like help everybody visualize what the current state of the art actually looks like. I think where we are is I think if you have a task where it's okay for the robot to make a mistake and it's possible for you to set up a mix autonomy system where you have a person that takes over when the robot make a mistake and provide corrections. It is possible to get to a level of performance where it starts to make sense to think about scaling robot deployment.
And the example that I specifically want to highlight here is this blog post that we've talked about. And it's great that these are both from YC Company. I want to provide a little bit of context here first. The context is that Pi is a primarily research organization. We want to focus on building the best model.
But we also want to not be tunnel vision. We want to make sure that the model that we built actually going to be useful and actually perform tasks that people in society cares about. And one of the really good ways for us to do so is to partner really closely with a company that want to get the robot out there today. And the way that these relationships work is that we treat each other like we're on the same team, very free flow of information. And we design a system that try to get the best performance for the tasks that these companies care about.
So let me talk about we first. What you're seeing in this video is a system that we built together, folding really diverse item of laundry in a real laundromat and condition. You can see people walking outside. And why this task is difficult is because there's just infinite possibility of observation space, like, you know, clodings are deformable and not two item of clothing here are the same. And these are also unseen.
These are not like clothing item that I've seen in the training data. Yeah. And I think they're some of the most cracked people out of Apple I've ever met. Carrie was the partner for me. Yeah.
I want to like explain like what we did and what their like company is. Yeah. I mean, they're actually shipping their first robots into the home. We sort of talked about it as being able to do household tasks like this. And I think they were very inspired by physical intelligences for stemos with laundry folding.
So it's actually a total trip to hear about it a year ago. We were talking about them doing it and then now to see them do it, working hand in hand with you is really awesome. I think this is a great example of like, you know, you need the model smarts, you need the data collection and then the hardware and the sort of system integration all working together is just hard to nail. Yeah. And to get back to your question about why robotic is hard, it's really it is a really hard system problems.
And you need everything to work well and work together to get this result. And like we've such an incredible team for us to work with to get this result. And it actually didn't even take us that long to get this result. It was roughly what we set a goal and maybe it was two weeks afterwards where we got a model that was a model and a system that was good enough at performing this task. It still like blows my mind to see a robot actually folding laundry because I remember until basically I didn't know if this would exist even in my entire lifetime.
It's like folding laundry. I mean, it's always been like the Turing test for robotics because there's no way to like deterministically program a system the way that you did like create AI to do this because this space is like so infinite. And like we've shown that it's possible for us to do it. Everyone can do this. Like robots will be able to do everything.
It's only a matter of like improving it from here. There was a funny story where when we first published Pi Zero, people thought of us as a laundry company because the demo was just focused on laundry and actually picking home tasks, especially tasks that has to do with the formable objects are intentional choice on our end. We're not just after the home, we really want to make it broadly applicable, but picking home tasks for us to start with has a few benefits. Like one is relatable. You know, you can see the laundry folding demo and you can kind of like grok how this is going to be useful and you can get a sense of why it's hard.
And the second is that it's really easy to set up to test generalization. You can talk about Ultra which is your company, Jared, demo of it. Yeah, this is Ultra. The thing that I love about this video is you see it's bright outside and you see this is for X speed and it's 100 minutes. If I scroll to the end, the sun has set.
Oh, wow. That was one of the big problems in robotics where it would be so sensitive to the environment in lighting and mess up the vision system, the semantics and part of it. Yeah. And the interesting thing here is that it is possible to get to the level of autonomy that the robot is just performing the task. This is autonomy at scale.
Like this is ready to be scaled. Because this task is less familiar than laundry folding. Do you want to explain what the robot is doing here and what Ultra is like doing as a company? Ultra is a company that want to make it really easy to adapt robot to new tasks. And right now they're focusing on logistics based, which is really important because there's lots of labor of footage in logistics.
And the task that we focus on together here is, you know, if you're on a lot of the data, you can just add an item from Amazon, you sometimes get this soft pouch that item gets shipped from. And the task here is you have a tray of these items here and the robot is supposed to pick one of them at a time and place it inside this pouch. The machine wouldn't close it and then pick up the pouch and put it on the left here to be ready for shipping. Now this heart is hard because there are many different types of objects that can be in this tray. And the opening here is actually very narrow.
So you see this interesting example of the robot kind of nudging the item to go into the pouch. And that's really hard. Like that's required. Very good understanding of the scene and like very precise motion to nudge the object into pouch. The other thing that's hard about this task is the level of autonomy that's required.
Like this is running for an entire day. There is still human intervention, I want to say, in this like full day operation. But the level of intervention is actually quite minimal. This is not just like some like demonstration, right? This is actually recorded in an actual e-commerce warehouse where they're actually shipping real products to real customers.
This isn't just like a lab. This is packaging real customer, real order for customer to be shipped out in a real warehouse. So this is real operations. So I think this is really cool because I think when people think about robots, they tend to think of the consumer use cases like we, if that's what we're familiar with in our daily life, what I find really interesting is that there's like a million applications like this ultra thing that you wouldn't think of as I was like, oh, who packs the like soft pouch of things that you get from like Amazon? Well, there's some person like who does that?
And this is like a dog that we can have a robot to do. The interesting thing about the approach is that you're converting it from a very difficult engineering problem into a operation problems of how do I identify the use case and how do I collect the right data, which is in some sense more scalable because you can build a system that allows you to collect data from any different tasks. So you know, it's not a problem of how do I scale data collection rather than, you know, every new task, how do I design a really difficult engineering system to solve it? Why see started school is back. We're hand selecting the most promising builders in the world and flying them out to San Francisco for July 25th and 26th to discuss the cutting edge of tech.
Apply now for a spot. Okay. Back to the video. I think one thing that the audience may not know is that you have a very unique technical insight that in the past, robotics folks would have kind of gas and be shocked because robots need to run in real time. A lot of times all of the compute runs in on device, but you guys have done something very different.
Can you tell us more about that so that this works in real time with large models and really well. So the context here is that we talked to many companies that would like to deploy robots in one of the first question we get is what compute units should we get on the robot. You know, it's expensive. It's going to increase the bond cost and they're worried that it's going to go out in fashion very quickly because the model changed tomorrow. It's bigger.
How do I make sure that the hardware that I'm going to commit to today is going to be viable for a couple of years. It's very difficult questions. People are often really surprised when I tell them that almost all of the robot evaluation that we've run at Pi today, including the really complicated demo that we have shown making coffee, folding laundry, mobile robots navigating around, the model actually hosted in the cloud. And you know, this is not like a cloud is in a server in the office. It's a real cloud.
The model is hosted in the data center somewhere and within this high frequency control loop that is controlling the robot, the robot is actually querying an API endpoint that hosts the model, sending it images and language command and getting back action that then executed directly on the robot. And this is surprising because of precisely the reason that you mentioned, you know, how do you actually make it work? This is why it's really important for Pi to couple system, hardware and model development and research very tightly together because it allows us to solve this problem. So for example, one of the insight that we have here is that you can actually bury the inference time within the robot control loop because, you know, if I'm a robot, I have enough action for me to execute for the next hundred milliseconds.
So there's no reason for me to wait until I finish executing that action to ask my model for a different action. I can do it as fast as inference, essentially. And so, you know, maybe when I only have 50 milliseconds of action worth left, I can ask for an access of action. And when the current 50 milliseconds is over, I have something that's ready for me to continue with my next 100 milliseconds. And so that's one of the inside the other kind of algorithmic improvement.
We refer to them as real time chunking. So I can actually do it in a way that's going to be a delay in how long it takes to query the model on the cloud, basically. Like the problem here, if I get a little bit more technical, is an action chunk is a sequence of action that I can execute on the robot. So, you know, it's not just one action. And if I have an action chunk that I can execute for only 100 milliseconds and 50 milliseconds in, I want to predict another action chunk.
And I'm going to transition to that new action chunk after my current 50 milliseconds is over. How do I make sure the two are consistent? Like, how do I make sure that if I'm moving this way, the next action chunk is going to continue to allow me to continue to be smoothly moving this way? And pre-compute. Yeah, you can pre-compute.
And like that's one of the hours make improvement that will make inference using models in the cloud possible. I studied computer engineering, so I'm not really an algorithms person. But when it comes to systems like that, like pipelining, like, get me all over that. That sounds great. That's so interesting.
I mean, this simplifies. It's a real choice because it simplifies so much of the system for the robots. You don't need all these clunky, I don't know. People have two operating systems that sometimes were for robots, embedded RTOS and then the regular one and all these complex giant compute and power. And this is what the initial versions of a way more used to run basically a server on the track.
And you can't afford to do that with general day robotics, which is brilliant that you figure out how to do it. Yeah, you don't have to. I mean, you can do things. Some of it obviously has to be some compute there, but a lot of the compute can happen elsewhere. And then there must be a video of this thing that we're looking at in the top left, like, how much of that is sort of like video feedback, how much of it is like local processed.
I mean, is there any compute locally on this robot or is it just like a dumb video camera that streams out of the cloud? For this, I'm not 100% sure, but I am inclined to believe that it's just a dumb computer. Like for this specific video, I don't remember, but I'm just 100% confident that we can make this work with a dumb computer on the robot. One interesting thing about our collaboration with Weven Ultra is one of never seen the robot in person. Oh, wow.
Two is I have very little idea about how the robot actually works. Interesting. And that's a very intentional choice. I want to stay away from that as far as possible. I also don't know how to collect data.
Like I intentionally don't ask them this question to understand whether it's possible for an organization like Pi to parrotute into their existing system and to work really closely with them on the thing that actually matters to get the system to work and not have to learn about how they've set up their system because in a way that's like a more scalable recipe. Yeah, you completely decouple a lot of the hardware control loop choices from the semantics and planning, which is just works. Just brilliant. Yeah, I mean, I'm really surprised. It works.
And when we started the company, we thought that real deployment is going to be in the conversation like five years into the life of the company because the problem is really hard. And what two years in, and this is the result that we have in real deployment and scaling the robot is a really serious consideration today. And so the pace of progress has just been very pleasantly much faster than we expected originally. Often on this podcast, we talk about what all this means for startup founders. I think that might be an interesting question for us to explore here.
So if you imagine someone was listening to this podcast, and you're like a college student that's studying computer science and they think robots are really cool and they want to do something like this, how should they get started and what are the skills that they need? Do they need to be a mechanical engineer to be able to build a robot like this? Can they just buy an off the shelf like robot arm and camera? System and like, well, and look, yeah, you're off the line. Yeah.
Before I actually answer your question, let me provide a few more context. The first is that robotic is traditional really hard because it's an extremely vertically integrated business. You need to have your own customer relationship, your hardware, your autonomy stack, your own safety and your own everything. And the barrier to entry is just really high because of that. And one of the things that we're trying to change is that we're trying to provide a foundation of physical intelligence that the community can build on top of that allow them to onboard autonomy onto their robot and their task much quicker than before.
So that's the first, you know, we want to provide that kind of seat of intelligence that allow people to move much faster so that they can focus on other problems. The second thing is that though I think the recipe for starting a vertical robotic business today is one, have a really good understanding of the existing workflow because the robot system needs to fit into this thing workflow. And the second is to be very meticulous about identifying where the opportunity is. If there's a workflow that needs X number of work today, where is the robot when you insert it's going to make it a bit different. And two is to really be scrappy when it comes to hardware and data collections.
You don't need an incredibly expensive robot that is capable of very precise motion today to do this task. And the reason why is this model really reactive and so they can compensate for some of the accuracy in the actual robot movement and to ensure that you have the ability to collect data and to run evaluation, especially evaluation in real deployment. The next step after that is to get a mixed autonomy system that allows you to get to the point where it's breaking economically. Right, even economically. Right, even economically.
Because the reason why that's important is because it allows you to then scale the number of robot. I just lose money in every robot. It's very hard to scale. But that's been historically one of the biggest challenges for robotic companies that they go into the growth stage is just the payback cap period. It just doesn't make sense.
Yes. So the equation, I think for starting a robot business has changed and will continue to change at an accelerating pace because the upfront cost is not that high anymore. And now what is the upfront cost? The upfront cost is much cheaper hardware, ability to collect data, ability to collect evaluation and ability to understand the use case to see where they should insert the robot. It's not about having incredibly expensive hardware.
It's not about having your own proprietary, I think autonomy, classical stack, any more to do this task. And so it allows companies to focus on the component that will actually allow them to differentiate themselves from the rest of the space. Now that you sort of unbundled it and you no longer need to build this fully vertically integrated company in order to build robotics company. Are we on the precipice of a Cambrian explosion of vertical robotics companies where there's going to be like a thousand companies like Ultra going after every like menial job in the economy and like getting a deep understanding of the customer, building a robot that can solve that problem, doing that like mixed human machine deployment until it like can run fully autonomously and building a company in every sector.
Is that the future that you see people building on top of Pi? It's funny that you mentioned Cambrian explosion because when we wrote this blog post, there was that term that was very kind of like hotly debated. We are I think academics at heart and we want to be kind of very major when we communicate. But myself personally, I believe there's going to be a Cambrian explosion of a robotic company across the entire world, across many many different verticals. Just because it's just so much cheaper to build and it doesn't require someone with 20 years of experience in robotic to start anymore.
It requires someone that is really scrappy, that can move really quickly, can do the system integration, can understand customer what they want to start the deployment. I mean what's coming up for me is obviously we work with a lot of robotics companies and meet a lot of founders and it feels like there's this continuum. One is to use an analogy to personal computing. You could argue that industrial robotics today is basically like mainframe or main computer level. If you look back in the 70s, huge public companies like Digital Computer that just did these very very expensive deployments but they were very specialized and it was all extreme enterprise.
The idea of a personal computer was ridiculous. It took the Altair and Apple One and Apple Two and then IBM PC XT to create personal computing. Then the traditional advice for robotics for many years is go after dirty and dangerous. Of course those are the industrial cases. You have these giant Tesla robots in the Gigafactory and things like that.
It feels like what you said around profitability is really really big. Does that mean that the people who do the vertical robot kambrane explosion moment or the people who are first in that, it sounds like they would be the first to be profitable and not dirty and dangerous? I think this is already happening today. I think we have the fortune of having lots of visibility into the robotic community because people would like to talk to us, people would like to learn what it's like to build a foundation model for a robot and people would like to know how do I get the same level of economy. There are so many companies and businesses that we talk to and I would love to put the robot into that space that it's okay for the robot to make a mistake and they just need it so much.
I really believe that the recipe that I mentioned earlier of identifying what a robot can fit in, focus on keyboard hardware, collect data, run evaluation, mix economy, break even, scale robots will work across many different verticals. I'm seeing it play out today and it's just incredibly exciting to see. And this is pretty cool that you literally just gave people the playbook for how to build a vertical robotics company. Like this is a playbook they could possibly be following successfully hundreds or thousands of times. And the reason why I want to mention it is because I do want to see that computer and explosions and we want to help enable it.
For pie, if we talk about why pie is going to fail, it's probably going to be because the problem is way too hard. It may be a take 50 more years to solve the robot problem and not a couple of years, 5, 10. And so we want to enable the community. We want to accelerate progress and that's why we're very open. We publish our research.
We open source, pie zero and pie five. And people also shock when they ask me, is there any difference between pie zero and pie zero and pie five that you open source versus the model that we use until only pie zero and pie five. And the answer was, I actually know it's the same model. The pre-trained model weights that you're using that we open source is also the pre-trained model weights that we use for pie zero and pie five. And so we really want to help accelerate progress in the community and to create that concrete and explosions.
Yeah, that's very inspiring. I mean, I feel like that's everyone's sort of spending a lot of time in the digital world. And it feels like now is the time to start thinking about the world of atoms. And this is sort of the perfect mix of actually, like, you know, how do you take a electron and turn it into abundance in the atoms world. And I think about Dario Amadeus, I say, all watched over by machines of loving grace.
And when you really think about the perfect manifestation of that, it's not like, you know, perfect agents that look over you just like in the electronic world. It's, you know, actually something a little bit more akin to what we're seeing here. Yeah. And this has really been our mission from the start is to create that concrete and explosion. And you know, this is why we choose to focus on the model because we believe that is the bar neck to just really make robot useful across many different different tasks in the world.
And that's why we also focus on cross environment, you know, success for us is not defined as only our model on our robot performing tasks that is useful. The surface area for success is actually much larger, which is our model performing really useful tasks on somebody else's robot up there. Maybe we don't even know what that robot is like in a way that's like useful to the end consumer. Can we maybe talk a little bit about like the humans behind the robots here? How do they get started?
Like who are your co-founders and how do you get together and what skills you each bring to such a complex problem? Sometimes the joke I make here is that the human behind the robots are also robots. Yeah, so high is a very, I would say, an untraditional company. We have like larger than average founding teams and some of us work really closely together when we were at the robotic team at Google. And the robotics team at Google was, I think, a really, really great environment for seeing the sign of life and creating the relationships in the community that allow the robotic community and like these advances to flourish.
There is Lockheed, which we met when we were thinking about starting the company and it has just been really instrumental in making sure that we're a good business. And there is Adnan, our hardware lead that came over from Andro. And Adnan has a really difficult job because if you want to work on Cross- NYU and remember my joke about how you had two years of your grad school, you bring on one more robots, the hardware problem and the operational problem for us is how do we build, improve and scale a fleet of, how to judge genius robots and it's just not one robot platform. And because we built the organization from scratch in the beginning to support that, I think we're able to do it, but it's just a really hard problem.
And because there's just like no two different robots in the fleet, how do you make sure that everything runs smoothly? We're really good at Diva and Conger, if you ask. How many co-founders are there in total? We have Brian, we have Chelsea, Sergey, myself, Lockheed and Adnan. Is it just necessary to have any co-founders to solve the problems because of this?
Or was it a case like you were already sort of like a unit together, you'd already work together and you just, whatever you started, you would all have... Yeah, one common question that we have is why you band together? And the first is that we really enjoy each other company. We spend a lot of time at work and it's in some sense give meaning to life. And so we really want to enjoy the relationship we have at work.
And the second is that any one of us could have started a company and be successful. The problem is just so incredibly hard. And the chances of success is just so much higher that we band together and we can divide and conquer the problems. And that's I think one of the main reasons why the progress has been much faster than we expected. What are the differences of you working before in either academia or a big industry, big company, like Google and as opposed to now in a startup?
This is the first time for a lot of you doing a startup, right? Yeah, this is the first time for a lot of us. One of the really surprising thing that we learned when we started the company is that the infrastructure for supporting large-scale general purpose robot, which is not there. And this starts from the software itself. How do you collect data?
What device you use to collect data? How do you manage data? How do you annotate the data? How do you get visibility into the data? How do you run evaluation?
How do you build operational process? Like there was a company that offered these kind of services, which is very different from software and we were really surprised to find out. And so we end up writing a lot of the software at Pi. I think this is another area of incredible opportunity of building services for a robot company. If you can offer remote tele-to-of, for example, if you can offer data collections, if you can offer annotation service.
Because these are functions that doesn't need to be repeated from one company to the next. So I think there's lots of opportunity to build support for growing robotic business. So that's one surprising thing that I learned. And the second is I think one of the reasons why we have managed to achieve such progress is that there is a really tight loop of collaboration in the entire lifecycle of model development. Going from what tasks you collect data for, you collect data for the tasks.
How do you do it? What hardware do you use? Once after you collect data, how do you get visibility of your actual data quality? How do you then make sure that you can easily train on that data? After you train on that, how do you run evaluation?
Evaluation is really hard to run in robotic because it scales super-linearly to model capability. That's how you have a model that can perform a two-minute task. Running evaluation for that is very different from running evaluation for a task at 20 minutes. It's not 10 times harder. It's more than 10 times harder.
After you run evaluation, how do you can distill the learning from the evaluation to know how to improve the model further? One of the really side project I would love to take on is to build an automated robotic research scientist, which is really one of the boronect we have today because this is a really difficult skill set that requires intuition about the entire stack. I would love it if there is a model that can ingest multi-model data such as this and analyze similar modes, understanding is the robot performing this way because of the data that was collected and annotated and trained the model. And then suggest idea and try them to figure out if those hypothesis are correct or if that's something I would love to have and dramatically unlock this.
Sometimes I make a joke in a company that we should record all of the meetings and train the model to make prediction about what is the next set of experiment. You totally could. What if it's Open Claw and Obsidian and Markdown Files and a brain.md with ontology that's custom to your use case and what if it's 100 open clause in the background that you orchestrate? I think there's two sides of this. The first is that we already see a little bit of a side of life where for simple failure modes and during evaluation, if you can describe the way that the robot failed in text very clearly, then you can ask that language model to make very reasonable recommendations about what the next step is.
But the flip side is that this only works for simple cases today. The reason why that's the case is because I think it's pretty fundamental limitation of the model that we have today, which is that they are not at the core model that take action in the world and see the consequences of its own action, especially action that changes the physical world. And so I think this kind of very fundamental understanding about how the physical world works is missing from the really large-filtration model. And I think that's one of the ingredients missing to be able to build this automated robot research scientist. What's interesting about Open Claw?
I don't know. I mean, basically it can go and they can just do things, which is interesting. And at that point, it's on the research lab to provide CLI MCP endpoints to the things that might control robots or reconfigure your rooms. I mean, I think Carpathi feels like he's starting to talk a bunch about this, where if you mix auto research plus what he's been talking about with Markdown files, it might just happen in the open. There's a sort of sense that you have to make something much, much more complicated to make it work.
But what if that's just wrong? What if we just have Markdown files and agents and you could make it yourself with literally cloud code and MCP today? What if it's not an algorithm problem? It's just literally an integration challenge. We have a version of this internally that I use a lot.
That was a point when I was spending an embarrassingly large amount of money on API queries. And my team was like, what are you doing? I'm not a guy at Y Combinator right now. So to give you an example, we have a cloud scale that is essentially serving the role of a pre-training on call today. So we have these pre-training runs that are really large.
It's very, I think, a difficult exercise to keep them alive, for them to continue to churn, because there's so many things that can go wrong. And we have a prototype of pre-training on call that kind of babysit the run and have the permission to take action, to remedy error that it see. And one of the surprising outcomes of that exercise is that it leads to about 50% improvement in compute usage, like just overall compute utilization for that large pre-training run, which is huge for us. And, you know, this is just a small, simple prototype. But I believe that there's a lot more to be done.
Korn, this is incredible. Thank you so much for everything. Thank you for making physical intelligence. Thank you for showing us these incredible demos. And honestly, the thing that gives me the most hope is this idea that there's an entity, there's a research lab out there that is focused on giving this to the world, you know, about to create this Cambrian explosion of robotic startup.
So someone watching right now will be inspired by this and, you know, start playing with your models and they might create a robot that touches billions of people's lives for the good. Thank you for having me. It's been a pleasure to listen to the one take away that I want you to have is I think robotic has changed a lot. And the cost of building in robotic has decreased and I think will continue to dramatically decrease. And it also requires a very different kind of scrappy skill set that young startup like meets we hope to enable really an explosion of many, many different robotic use case and you know, always reach out to us if you want to collaborate.
Thanks, man. Thank you. Thank you.