OpenAI Set To Surprise Everyone By Launching Thier Own Humanoid Robots

So one of the craziest things that OpenAI have recently been discussing is the fact that they are considering developing a humanoid robot. This is the information, according to the information. And apparently the company has recently considered developing their own humanoid robot, according to two people with direct knowledge of the discussions. Now, over the past year, they speak about how OpenAI has dropped not so subtle hints about its revived interest in robotics. And we can actually look at the companies that OpenAI have invested in. For example, if we take a look at physical intelligence. This is a remarkable company that I covered earlier this year that I don't think has gotten enough credit. You can see here, Jeff Bezos at OpenAI have invested in robot startup Physical Intelligence at a $2.4 billion valuation. If you're wondering why you've probably never heard of this company, that's because this company, Physical Intelligence, is. Isn't a humanoid robotics company. This is a company that is focused on autonomous robots that can pretty much do anything with two hands. Now, when we take a look at what the physical intelligence robots actually do, these are ones that can pretty much do a lot of tasks autonomously that previously we thought weren't really possible. Now, this is something that I think is remarkable because I do remember that when I first covered a robotics company that was doing this, it was called Mobile Aloha, and that was a company that I thought was doing some incredible work. And it was. It actually responded, you know, like the audience actually thought so as well. And in that video got tons and tons of views. So this is a company that is able to do a variety of different tasks autonomously. Right now this is something that is 100% autonomous. This is not teleoperated. And this is one of the companies that OpenAI have invested in. And it seems like they are backing this company now, in addition to backing this autonomous company. And this isn't all the company can do in terms of just packing robots. They can do laundry, just. Just a variety of different things. And trust me, those policies that they generate can generalize to other robot platforms as well, which has stark implications. Like, for example, in this demo, where we're actually seeing the robot clear away a table, it's navigating the trash, and it's literally putting everything in the bin. I mean, imagine in 10 years when you have your robot server come up and clean the table for you. I mean, a company like this backed by the likes of OpenAI and Amazon, wouldn't be surprised. Now, another company that you are probably more familiar with is, of course, the company called Figure. Now, this is a company that is valued at $2.6 billion, as many companies, including OpenAI, have, of course, joined funding. Now, the crazy thing about this is that this company is actually powered by OpenAI. And in their recent demo, we got to see the first glimpse at what humanoid robots might look like if OpenAI have created them, their selves. This is something that gained huge virality in the Internet space. And this is something that I think genuinely brought huge humanoids into the main sphere, as I saw a lot of people that weren't really a part of the AI robotics space talking about this for the first time.

Hey, Figure one, what do you see right now?

I see a red apple on a plate in the center of the table, a drying rack with cups and a plate, and you standing nearby with your hand on the table.

Great. Can I have something to eat?

Sure thing.

Great.

Can you explain why you did what you just did while you pick up this trash.

On it? So I gave you the apple because it's the only edible item I could provide you with from the table.

Great. So based on the scene right now, where do you think the dishes in front of you go next?

The dishes on the table, like that plate and cup are likely to go into the drying rack next.

Great.

Can you put them there?

Of course.

All right. So how do you think you did?

I think I did pretty well. The apple found its new owner, the trash is gone, and the tableware is right where it belongs.

I agree. Thanks, Figure one.

You're welcome. If there's anything else you need help with, just let me know.

Now, I do think that this update will be one of the most fascinating updates, because this demo was one of the ones where we actually truly got to see how large language models can truly enhance the experience of robots and them interacting with the real world. It's really interesting to see robots do a variety of different actions in physical environments. But when you can actually talk to a robot and can actually get feedback on what it's doing and understand why it's making the choices it has, then that is something that is just, you know, elevating the level of human to robot interaction. So it will be interesting to see if this is something that OpenAI continues to work on with these companies, because the last time we saw a demo like this, we all know exactly what we saw. Now, in addition, figure isn't the only company backed by OpenAI. We also recently have 1X Robotics being backed by OpenAI in $100 million round in the race to Humanoid robots. Now, the reason that this robot here was one of my favorites in 2024 was because this robotics demo was just absolutely incredible. And one of the standout features for me was not just how smooth the robot moved or how human like it was, you know, operating was the fact that you couldn't even hear this robot at all. A lot of the robots sometimes you hear like the large gears moving and stuff like that. But this robot was so, so quiet. Like it was just so Apple esque in the way that it moved and the way that it seemed. So I think this one here is going to be super fascinating because it doesn't really have the robotic, you know, look, it has more of a sleek and, you know, friendly design. And that's something that they really try to work on when designing the robot. You can see that it is, you know, covered in cloth. And this is another company that is working closely with OpenAI. And I'm wondering if they will embed their large language models in them. Now, interestingly, we do know that OpenAI recently restarted its robotics team and two months ago they revived their robotics team that it disbanded years ago. Okay, so that is a very interesting thing. Now some of you guys might be wondering, well, why did OpenAI decide to of course leave their robotics team in the first place? Why was it disbanded? And Ilya Sutskever actually talks about this in a Dwarkish Patel interview. He gives us an incredible insight with as to why exactly OpenAI chose to abandon robotics. And he here he explains that it was mainly due to data.

Back then, it really wasn't possible to continue working in robotics because there was so little data. Like back then, if you wanted to do on robotics, if you wanted to work on robotics, you needed to become a robotics company. You needed to really have a giant group of people working on building robots and maintaining them and having. And even then, like if you only if you're going to have a hundred robots, it's a giant operations already, but you're not going to get that much data. So in a world where most of the progress comes from the combination of compute and data, right, that's where we've been, where it was the combination of compute and data that drove the progress. There was no path to data from robotics. So back in the day when you made a decision to stop working in robotics, there was no path forward.

Is there one now?

So I'd say that now it is possible to create a path forward, but one needs to really commit to the, to the, to the task of robotics. You really need to say, I'm going to build like many thousands, tens of thousands, hundreds of thousands of robots and somehow collect data from them and find a gradual path where the robots are doing something slightly more useful, and then the data that they get from these robots and then the data that that is obtained and used to train the models and they do something slightly more useful. So you could imagine this kind of gradual path of improvement. You build more robots, they do more things, you collect more data and so on. But you really need to be committed to this path. If you say, I want to make robotics happen, that's what you need to do. I believe that there are companies who are thinking about such doing exactly that. But I think that you need to really love robots and need to be really willing to solve all the physical and logistical problems of dealing with them. It's not the same as software at all.

Now, in addition to that, there was also the researcher at OpenAI, Wojciech Saremba. And this is, you know, one of the founding members of OpenAI. He actually spoke in an interview in 2021 on weights and Biases podcast about how there were just too many issues with robotics. And it was really interesting to see the kinds of issues he spoke about. And do you want to know why that interview was fascinating to watch once again, is because when you take a look at the issues that he speaks about here and he mentions things like, you know, data and, you know, stuff like, which I want you guys to take a look at quickly, do you think that there's more progress in the short term, or will it be sort of the last thing that we solve on the path to AGI.

So there are two possibilities for me, like a few possibilities. So one is if someone would be able to actually in natural way to collect a lot of data, I think that might be the capabilities.

And so now, of course, we could have that situation where the data issue is solved. We could have simulated robots and then immediately deployed, deploy that to the real world. When we have simulators like Genesis that allow us to place robots in environments in the simulated world and then map those one to one. We do know that this is something that could result in something amazing in terms of speeding up robotics. You can see right here that they talk about the fact that this is much faster than Nvidia's Isaac Gym and it is 20 times faster than these GPU accelerators. So this is something maybe opening eyes, looking at it and like, okay, maybe now that the data issue and the speed is solved, maybe we can start hopping back into robotics and developing something alongside these other companies. Now of course you might be thinking that are they competing with Tesla? Well, I think OpenAI is one of those companies that is on path to be a trillion dollar company if they continue at the pace that they're going. And when we look at the size of the humanoid robot market, we can see that the humanoid robot under development by Tesla could multiply the company's valuation. And Robotaxis make Tesla about a $5 trillion companies. And the Optimus robot could make Tesla a $25 trillion company. When you think about the fact that these are basically humans duplicated on a infinite scale, we have to realize how much economic value we could pull from that. And I think Maybe, just maybe, OpenAI is trying to grab a piece of that pie. And something to consider is the fact that recently we did see Google DeepMind join forces with Apptronic's robotic team and they are going to be combining the best in class AI to with cutting edge robotics hardware to create AI powered humanoid robots. And it says get ready for a new era of helpful robots in dynamic environments. So overall we can see that there is going to be some competition from these top labs in terms of the kind of products that they're going to be working on. And I think maybe OpenAI wants their own. I personally don't think OpenAI will create their own humanoid robots. I think they're personally working on a million different things and I think that might slow them down. But I'm always ready to be surprised considering the fact that OpenAI are extremely talented company. But it will be interesting to see what they're going on right now. Of course it is just discussions, so nothing is concrete, but it will be interesting to see what potentially does occur. Hardware is of course really hard and they did leave it at the first time for a reason. So will be interesting to see if they actually manage to come back.