Limitless Podcast - Inside The World's First Robot Olympics: Wild, Real, & Unhinged

Episode Date: August 19, 2025

China just staged the first-ever “Robot Olympics” — the World Humanoid Robot Games — and we watched 12 hours so you don’t have to. Robots sprinted, boxed, played soccer, and even fo...lded laundry as 280 teams from 16 countries sent 500+ bots to Beijing. We unpack the viral falls and real wins, from Unitree’s 1500m (6:34) to the 100m controversy where an autonomous Tiangong beat a faster tele-operated run via a rule multiplier. More importantly, we break down autonomy vs. teleop, why dexterity and chores matter more than sprints, and how public competitions are becoming the benchmark—and data engine—for embodied AI. Stick around for spicy takes, future timelines, and the robot form factors that’ll actually matter at home and at work.------🌌 LIMITLESS HQ: LISTEN & FOLLOW HERE ⬇️https://limitless.bankless.com/https://x.com/LimitlessFT------TIMESTAMPS00:00 Start02:30 What On Earth Is Going On?04:17 Epic Fails06:12 Setting Speed Records10:20 How Autonomous Were They Really?14:53 Why Software Matters18:49 Folding Your Laundry23:13 Are Humanoids Final Form?27:24 Ice For That Burn?------RESOURCESJosh: https://x.com/Josh_KaleEjaaz: https://x.com/cryptopunk7213------Not financial or tax advice. See our investment disclosures here:https://www.bankless.com/disclosures⁠

Transcript
Discussion (0)
Starting point is 00:00:03 So how was your weekend, dude? It was nice. One of my best friends got married. I was in the wedding. It was hot and sweaty all weekend love. It was a good time. And the good and the bad part about weddings is I'm mostly disconnected from the world over the course of the weekend. So there is this low-level anxiety. I'm like, okay, what am I missing? And then I get back to my phone at the end of the weekend on like a Sunday night and I'm just like, I got to do it a little doom scrolling. So what did you do? How was your weekend? Anything noteworthy? Yeah. I think I did the complete opposite of that, dude. I spent the last 12 hours watching the robot Olympics. I was remote tuning into China, baby.
Starting point is 00:00:39 I was watching this entire thing. So, okay, so please explain. Yeah, okay. So this is called the world humanoid robot games, and it is the first of its kind. Think Olympics, but instead of humans competing across a range of different sports and track events, they're full-on humanoid robots.
Starting point is 00:01:00 The first of its kind, They're boxing. As you can see, they're racing on 100-meter race. They're doing the baton race. They're doing high jump competitions. It is absolutely insane. Down to the point where they had this insane opening ceremony. I don't know if you can see my screen, Josh.
Starting point is 00:01:16 If you're not watching this, if you're just listening to this, I advise you to go on YouTube or Spotify and actually watch this. It's pretty ridiculous. It's like everything you would expect from a traditional Olympic Games, except instead of humans, there's just these robots that are shockingly cute and charming. And I kind of like them. They did a good job of not making them look evil. They seem to be like stumbling around a lot, Josh.
Starting point is 00:01:38 Like I looked at some of the pictures. Did you, I'm going to pull this up. Look at this. I mean, some of the action shots here are insane. I mean, okay, firstly, let's level set here. We have this image. We have an image of a robot, you know, just taking kind of like a refuel session. So, you know, it's charging up.
Starting point is 00:01:59 I think the team is like calibrating it to do. God knows what. And then we see like some of these robots in full on action. For example, here where I think someone's going through the middle of like a flying kick in what is presumably the kickboxing event against another robot. So extremely lifelike, except it's not, Josh. It almost looks AI generated, right? Like it doesn't even look particularly real. These photos are beautiful, by the way. Like these, these look amazing. The robots look great. I really love this. So maybe we'll set some context for like what people. are actually looking at it. So, I mean, you described earlier. We're looking at the world human
Starting point is 00:02:36 robot games. This is the first ever, which is interesting because I think what, the modern Olympics started in 1896 or something like that. So it's funny, for the last like 130 or so years, we've been competing with humans. And now for the first time we've moved over to robots. And this was a pretty large scale. There were 280 teams from 16 countries and over 500 robots competed across events like track and fields, soccer, table tennis, boxing, kickboxing. And then there was also my favorite one, there was chores, like medicine sorting, material handling, cleaning of clothes and things of that nature. So it was like practical things too. So a robot would be boxing next to a robot folding clothes. And I think I loved every part about how they decided to run
Starting point is 00:03:17 this Olympics. I think one of the more fascinating parts about this is how excitable the humans got. Yeah. So for context, this happened in China in Beijing. And so you would expect some of the audiences to be kind of like nerdy and maybe kind of like niche. But from everything I could watch from the live stream, there was like a huge fandom for these things. Like I was actually quite shocked. People brought their families. People brought their kids.
Starting point is 00:03:46 There were kids walking around the track, kind of getting like fist bumping these robots, getting autographs. It was like a weirdly like human societal thing, but with these like largely mechanical robots, which I thought was like wild. Another thing was, I'm just going to be honest, I think the fails were more entertaining than the athletic prowess that these robots had, Josh.
Starting point is 00:04:07 Like, I need to show you some of these to give you some context. Okay, so here we have a winner falling down from the medal podium. Okay, cool. Lost ahead. Right. But surely you can't get any more worse than that, right? But here we have a very, you know, prominent participant. I think this is Robo Terra making a dash for the 400 meter race. And this is the first of its kind.
Starting point is 00:04:29 It's really keen to make, you know, some kind of a metal position on this and it ends up falling flat on its face. It's kind of like this. It's funny. They kind of lose their sense of like human physics and it shows the robot physics where when they fall, they just fall and kind of collapse into a puddle. Yeah. And in some of the cases like just obstacles got in the way.
Starting point is 00:04:52 As you can see here, like you just have like a head-on collision. So all in all, it was a very hacky kind of. first attempt at a robot Olympics. Like, I don't know about you, but when I think about robot Olympics, I think about these like supercharged Iron Man type looking things that are just kind of like flying everywhere, running faster than any kind of human could ever do
Starting point is 00:05:16 and throwing things much further than any kind of humans could do. But this kind of looked like a hacky start. These robots were kind of like ranging from like mini kind of three foot size little kind of like gremlins running around versus these like, honestly, these robots like that hit like seven and a half to eight feet tall, right? So, myveling some of these like NBA like basketball players. So overall a kind of very weird experience, but very, very entertaining. I wonder if the enthusiasm is around the novelty or its actual genuine interest in these robots.
Starting point is 00:05:48 Because, I mean, for me watching these videos, it is fascinating seeing these things that look kind of vaguely like humans running. And it just, there's something so bizarre about it because you always see this in the movies, but to see this in reality. is like a totally different story. So maybe it was amateur for the first one, but there were some big wins, right? Like there were some pretty interesting stats, some pretty amazing benchmarks that people won, particularly. I think one of the biggest races was the 1500 meter sprint, basically. And Unitary actually won.
Starting point is 00:06:16 And Unitary is a very popular human robotics company. And they, I mean, reportedly got six minutes and 34 seconds on the 1500 meter gold, which is fascinating because I think the world record for a human is only half of that. that time. It's three minutes and... 326. So, I mean, granted, this is version one, but they are only 2x the speed of the fastest human in the world. And you could imagine by like the second or the third one, surely they will be able to beat this. It's like when you look at these humanoid robots, they look kind of goofy, right? Like they don't quite have their legs under them. They kind of look like their children. And when I think about this, I think of full self-driving, Tesla's like
Starting point is 00:06:53 full-self driving autopilot program where at the very beginning when you first start using it and you kind of see the AI thinking almost, where it moves kind of slowly, it moves a little hesitant, it's not very confident in its movements. And then over the period of time, as it kind of learns, it gets better, it transitions from this child to like infant to like toddler, elementary, middle school, and then eventually you get to full-blown Olympic athlete. And I can very clearly see the same trajectory happening here, where, I mean, to get double the record on the first try, surely by version three, they will be kicking our ass. And then I would imagine version five is like, what, they're just running around for one minute and it's blowing every human out of the water.
Starting point is 00:07:31 We'll put it this way. Like, even if you're double the fastest human in the world, that's still quicker than like at least 80% of the human population who are like sitting on chairs like, I am looking at a laptop screen and not really exercising that often, right? So that is just, I mean, look at this clip. Yeah, so what you're telling me is I have about one year until ROWAS cannot run me. Well, yeah, yeah, it outstrip us and pretty much everything. I mean, look how Quick this thing is going. Look at this video. I wish the listeners can, like,
Starting point is 00:08:00 grab a screen sometime soon because the remote control operators for this unitary robot, that one gold, are absolutely pegging it. Like, look at them behind this robot. Just absolutely going for it. I think the thing that, like,
Starting point is 00:08:15 surprised me the most, kind of stepping back away from this, Josh, is, I'll be honest, I've kind of been a robot hater. I've kind of been looking down on these things. I've kind of been thinking, At most, they're going to be living in factories doing menial work that I couldn't care less about. Folding clothes, putting wrappers on different types of products and maybe packaging different things.
Starting point is 00:08:40 But I never really thought of them as like on a level of other kind of humans doing human like casual or competitive activities. And if I did consider them in those cases, I didn't think they'd be good enough. but I'm like kind of like looking at this and I'm thinking, oh crap, they're going to be quicker than us in about two iterations, as you said, which is at the scale this is going, probably a couple of months. So the big takeaway for me is these robots are going to be way more impactful in a much bigger way, way quicker than I thought they were going to be. Oh yeah, this is great because I strongly disagree. I think robots like everywhere, everything all the time. And it kind of ties back to the conversation we had. I mean, we just recently announced with Arvin, the CEO of Perplexion. about the role that these robots play in the world, particularly in the sense of the browser, where we talked about, well, browser kind of serves two purposes, right? It's leisure and then it's
Starting point is 00:09:31 productivity. And I think when I think of robots, I kind of think of the same thing where the things I enjoy doing as a human, well, I want to do that as a human. I want to go hang out with my friends. I want to sometimes drive a car really fast. But the productivity stuff, like the folding the clothes, is they going to get your groceries, the cleaning, the doing the dishes and the laundry. That stuff, I would love for a human way to do. And I was really glad to see those featured in this Olympics because those are the practical things that I want. I want to offload the productive parts of my life that seem tedious to humanoid. And it seems like even very early version ones are already doing that really well. So it seems clear this like this natural trend is
Starting point is 00:10:06 moving towards more and more humanoid robots as a form factor, but also different form factors. There were some other robots that weren't just humanoid. But there is one one thing I want to touch on here that I thought was interesting and is worth mentioning. It's that not all of these humanoid robots were fully autonomous. Right? So they had a little bit of help. So, Ejas, can you, do you walk us through? Like, how autonomous was this really? Because we're seeing the humanoid robots and they look like they're running quickly. But is, is there a human actually behind them who's controlling them kind of like an Xbox? Okay. I think rather than bore everyone with the facts, I want to lead with the single most important event at any Olympics. Do you know which one that is, Josh? No. Fill me in. What is the most important event? It's the 100 meter race, of course. It's a hundred meter race, of course. It's what everyone is strongly in anticipation. They are excited about, right?
Starting point is 00:10:58 Oh, is this like, is this what Hussein Bolt did when he made the record? Oh, yeah. Okay, this is a big deal. Yes, because this is the one that I actually remember. This is the big one. This is the one that, you know, you might not even be a sports officiantado, and you tune in for this race. It's where it is eight, seven to eight seconds of the quickest stuff you've ever seen and the most high adrenaline-fueled thing, right? So we had the robot equivalent of this, and it turned out to be an extremely contentious event,
Starting point is 00:11:28 because technically one robot Unitri, which, by the way, was sweeping up at all the other events as well, won first place. It crossed the finish line first, but it didn't place gold, Josh. It actually placed silver. Do you want to know why? How does that work? Yes, please share. The reason was Tian Gong robot, which is.
Starting point is 00:11:51 is a different robot company who came in third was autonomous. And because it was autonomous and because unitary that technically crossed the finish line first was remote controlled, you could technically say that Tian Gong was more advanced. And therefore, it deserved some kind of a handicap or advantage. So what the judges gave it was a 0.8% multiplier to the time that it took to cross. the finish line, which ended up placing it quicker than Unitreys robot. So it ended up getting first thing. Okay. I want to get your first gut take on this. Tell me about it. You said it feels right.
Starting point is 00:12:33 Yeah, that feels right. I think that's right. I think if you are able to do this fully autonomously, which is the end state of all these humanoid robots, you deserve a premium for that. And it's funny because watching this video, you're watching the robots run, but then you're also watching the humans follow behind them with these controllers, like sprinting towards them on the sidelines. It's this really funny, stupid-looking. thing where it's like, there's a bunch of like Chinese engineers who are just running with these controllers, I guess, manually controlling them. The most that run in a while. Exactly. So if you're watching, like, one just finished and then the second one just finished. And the second one that's
Starting point is 00:13:03 finishing is currently running at like a 45 degree angle sideways. But it finished. And it finished like fairly quickly. So I think that deserves a lot of credit where maybe they weren't the straightest, but they got there in the second fastest amount of time and fully autonomously. And that goes a long way. With the tele operation, it's, it's very easy, I would imagine, to do that. Like, the physics of a robot, like the actuators, making it move and fixing the center of gravity as it moves quickly is probably a very challenging energy for you. Don't get me wrong. But, I mean, to just have a controller with a joystick and you're just telling it where to go and keeping it center, that is much more, that's much easier than developing the entire full stack. So it seems like they've solved the hardware
Starting point is 00:13:41 stack, but I'd argue the software stack is equally as hard, developing that intelligent autonomy. And to do that, I mean, I would even venture to give them an even higher multiple. Because if you're able to do that, it's probably worth more than a 20% premium, in my eyes, at least, in terms of what's valuable. Okay. I want to say I disagree with you, Josh, on this one. Okay. Just I want to play devil's advocate on this for a second. The goal of this event, and actually all these events, is to run the quickest, jump the highest, maneuver around the obstacles the best. Now, there's a bigger question around whether these robots should be completely autonomous.
Starting point is 00:14:21 Eventually, I think they should be, right? You don't want like a human handler. But for now, I guess my naive gut take is like, it kind of makes sense. I kind of like don't mind if there's a human handler. What I care about is how quickly it is, right? And so the robot that came in third, okay, fine.
Starting point is 00:14:38 So what if it ran autonomously in a straight line? Why didn't they build it in a way that, it ran faster, right? Couldn't they do both of those things at the same time? Why not focus on the hardware stack if the goal is to win the race? Yeah, it's an interesting question because I guess we're getting into why this actually matters. I mean, beyond the memes, like why is this an important event? And I think there's two things, right? One is to kind of show off your engineering prowess as a country or a company. It's to show what you're capable of doing and creating. But then the other thing, and the interesting part to me and why I'm kind of stingy on the software stack of things is because in a way we're kind of benchmarking embodied AI in public.
Starting point is 00:15:21 So for software, we frequently talk about like benchmarks. This is how we measure the successive models. We have GPT5, we have Gemini, we run them through all these benchmarks, and that's how we kind of say how good they are. Because us as humans, we're not really that great testing because we're not really smart enough at all these areas to keep up. So we just have the smartest people we know create really difficult questions and we benchmark them. And there hasn't really been a benchmarking system for embodied AI in public. So I guess when I heard the robotic Olympics were happening, I hoped it would kind of serve as this embodied AI benchmark.
Starting point is 00:15:54 And a big part of that is converging the software stack with the hardware stack into one final package because, I mean, in my mind, at least that's the final form. We want these robots to be autonomous. You can't just have someone controlling them. Otherwise, it kind of defeats the purpose. So that's why I'm disappointed. I think you're totally right in the sense. it's not fair to make people run them fully autonomously just sit yet because we're so early
Starting point is 00:16:16 and there just hasn't been enough time to build the software stack and converge it with that hardware stack. But man, I hope like, I hope future iterations. After these engineering teams have had more time to develop the software, there are more strict standards on the software and I guess teleoperation metrics that they use that offer these discounts to incentivize people at least to build the AI part of it as well. Yeah, I agree with you largely on most of that. I think my mind kind of splits into two parallel tracks, Josh, and they are the following. On the one side, I think humans and AI should collaborate and continue to collaborate into the future.
Starting point is 00:16:57 That doesn't mean that I don't think these AIs, whether they're in robotic form or software form, shouldn't be independent. but I do believe that ultimately for humanity to survive and achieve its best, you know, optimal form, we kind of either need to merge directly with them and that might be chips in the head or in the brain or we kind of work conducively with these robots. And on that side of things, I think that's going to be great. And therefore, you know, that supports the human remote controller robot relationship that we're seeing in this race, right? And maybe I'm extrapolating too much, but bear with me. on the other side, you know, we keep on talking about agents on this series, on this show, right?
Starting point is 00:17:41 And we keep saying the number one frustrating thing is that the humans need to kind of follow up with these agents. Oh yeah, that code is good. Oh, yeah, I'm okay with you buying this thing with my card. Just, you know, I approve it, right? And it gets kind of annoying. I just kind of wanted to do the thing and I wake up at the morning and the thing is done. And I feel good about myself, right? So I'm kind of like battling between these two things. And I think ultimately like there's going to be no black or white outcome. But these things are going to become really more ingrained in a day-to-day life as well as like kind of like physical competitive sport and other cultural aspects of humanity. And it kind of got me thinking, Josh, around, you know, one of the lead robot kind of creators that, you know, we're stepping aside from like this robot Olympic stuff. And I'm looking at like what other things robots can practically do. These things are getting.
Starting point is 00:18:33 really much more agile, right? So we've spoken about these guys before, but figure robots, they're kind of been like one of the leading consumer robotics companies. You've got Tesla, you've got Figure, you've got Apptronic, and a few other things, right?
Starting point is 00:18:47 What you're looking at right here kind of gives you a glimpse into what that ingrained robot human relationship looks like, right? You've got a robot in front of you that looks extremely human and is folding your clothes, but it's not just kind of like
Starting point is 00:19:02 slapping your clothes, modes around, it's being very delicate. Look at it, Josh. Can you see this? So it's pretty insane. Again, like, one of the things I really loved the most is that the games showed that robots can sprint, but they can also scrub. They have the perception and dexterity and like kind of the longer horizon planning where their goal isn't just to get to a finish line. It's to dynamically adjust to the thing at hand. And what we're seeing now with figure, which I would argue is probably tied for number one with Tesla right now in terms of just how impressive the humanoid robots are. we're getting this scrubbing. We're getting this the perception. You're seeing the dexterity. You're
Starting point is 00:19:36 seeing the long horizon planning where what we're looking at is a series of towels that are sitting on a table. And the robot's kind of picking a specific towel. It's placing it in front of it. It's folding it and then it's placing it inside of a basket. And that requires a lot of dynamic thinking and a lot of dexterity on the hands. And it appears to be doing it very gently. So if this was something fragile, it would actually be able to handle that too. And this is the part of humanoid robotics that to me feels really impressive. It's like, Like when, I mean, to draw again from another Tesla comparison, it's like when Tesla made the roadster, the first roadster, the idea was just to create this like hardcore smackdown on gas cars. It was to prove a point. And I think the Olympics is very much a same thing where it's to prove a point.
Starting point is 00:20:15 It's like, look how fast my humanoid robot can run. But I think, and then the best selling car in the world wound up being the Model Y. And that wasn't a hardcore smackdown. It was just really good at doing the basics right. And I think what we're seeing with humanoid robots is like, the Olympics are showing the headlines are like, look how fast this robot ran. like strong this robot was. But the reality is, is you just kind of want something that works, that's gentle, that's nice, that isn't too aggressive or intrusive, and it just does the things you want. And that's what we're seeing with figure and this robot right here. And as it's handing off the
Starting point is 00:20:46 basket to the employee and goes and walks away, does that seem right to you? Yeah. Actually, we're having a conversation with Aaron Tan tomorrow, who is building this robotic startup that is building this kind of like seamless set of robots that you don't notice in your home. They're kind of like furniture. And then when you're away or when you're out of the room, the furniture morphs into like robotic hands that can fold any clothes that are on your bed or do the ironing or stuff like that, kind of crazy, right? And I actually agree and disagree with you on the sense that it's tough to kind of measure
Starting point is 00:21:25 what these robots' impacts are going to be like on our lives, Josh. because I would say, going back to your earlier point, these benchmarks are much easier to see, right? I can see whether the robot's going to fall my clothes. I can see whether the robot can do the shopping because I'm going to come back home, open the fridge, and see whether my shopping is in there, right? It's much harder to gauge on the LLM level, right?
Starting point is 00:21:47 When it's more of a vibe, when it's more of a thing. But robots having such a physical capacity to change the world, I think is like a great metric, not just for me to observe, but the rest of society to observe, to really see whether the impact of AI and robots are actually a thing or not. Yeah, this is amazing benchmarking because the limit is physics. You are limited and constrained by the finality of physics, and you could see that visually right in front of your eyes. And there's something really special about that where it's very clear and obvious. I mean, a lot of large language models is very subjective, which one's best,
Starting point is 00:22:21 because they're like a brain, they're just good at different things and some are better at other aspects than some. But with the physical opportunities, I think it is very clear and obvious which one will win. And you're going to see these companies slowly get closer and closer to these limitations of physics. But the thing you mentioned with the different form factor of robotics, and the guests we're going to be having on later this week, is that when I saw the robot you're describing, which is basically, if anyone has seen a bed frame before that has these like pillars that stick up from the corners, it's like that, except those pillars are robotic arms. and the robotic arms can come down on the bed and it could fold clothes on it,
Starting point is 00:22:56 and it kind of shattered my perception of what the form factor of these robots can be. I think a lot of times we see humanoid robots, and in the case of these Olympics, we're getting the humanoid robot form factor. And why humanoids? Well, it's because most of the world is actually just built for humans. So we might as well emulate that form and allow them to walk around because we don't have to rebuild the world for these robots to interact with it. But there's the case, and here we'll, here we go, we'll share it on screen right now.
Starting point is 00:23:22 there's the case for these types of robots. There's the general purpose, but then there's also the narrow band robots that are good for one thing. And in this case, I mean, this, watching this video really gets real excited. This is doing something to me. Because this, it gets back to the earlier point
Starting point is 00:23:36 where it's just doing the thing that I wanted to do very simply, very elegantly, and very nicely. And when I get home, it's just folded. And it's not obtrusive. They look nice. They're multifunctional. They're like a little light post. And then when I leave, they come down
Starting point is 00:23:48 and they fold my clothes. And it just seems like a really wonderful application. So one of the things I'm going to look forward to for future Olympics, but also just future robots in general, is removing the word humanoid from robots and just seeing what we can do with robots, period. Because I think that the form of robots, the form that robots will take over time will probably vary much differently than what we perceive. And like this, this loom product is example number one of that happening. What do you, what do you take on on robots? Do you have any other form factors that you think are cool that we're not thinking of? No, I actually think that humanoid-looking robots haven't been designed from first principles.
Starting point is 00:24:29 And that might be like a hot take, but I think purely making a robot to look human because that's what humans look like is kind of like a weird design choice. And maybe there's much more in-depth thinking. Maybe bipedal type robots are the best ones to kind of move around and navigate the world because that's the world that humans have built. and that's fair and game. However, I think that the best robots that will do the best to start off with are going to be unilateral. I don't know if that's the right word,
Starting point is 00:25:02 but what I mean is singular function. So they're going to do one thing really, really well. And this isn't too indifferent to what we've seen with LLMs or AI models, right? Like general AI models are really good at giving you general feedback and information. But where do they excel, Josh? it's at things like coding
Starting point is 00:25:22 and specific mathematical formulas and scientific kind of processes and a number of other things, right? They excel at those things, whereas getting it to kind of like the same standard at generalistic kind of intelligence is super hard. That's why the race to AGI is anything but a race, right?
Starting point is 00:25:40 It's like a slow moving thing where every update we get is incrementally better but not as bigger leap as we saw six months ago and that might still change. And I think the same applies with robots right now. I think we are kind of in the, not GBT2, but GPT3 kind of phase. We're at the point where like these things are kind of blowing our minds and we're like,
Starting point is 00:26:02 well, hang on a second, these robots can move quickly. They can fall my laundry. I can order one and it's going to arrive in six months, by the way. This is not a shell for Aaron Tan's product, but Josh and I have put our names out on the waitlist and we're hoping to get one soon. But, you know, it's starting to kind of like appear in our lives more. But as you said, they're V1. And I have a feeling that we have a few more versions to go until they get really, really good. The first principles take is it's a pretty hot take because you're at war with Darwinism there for millions of years of evolution designing humans and we're trying to emulate it.
Starting point is 00:26:31 But I think that makes sense for the narrow use cases because humans were designed to survive, right? Now, these are designed to fold my clothes. And a human is far too large, far too energy needy in order to just fold clothes. But there was a little bit of drama associated with this post, right? Because it wasn't just the humanoid robot. someone clapped back. Noah Brown. Noah Brown works for Open AI, right?
Starting point is 00:26:53 Oh yeah, he's not just someone. He's the head of reasoning. He's like... The head of reasoning at Open AI. He had something to say. Yeah, he had something to say. So basically, a lot of robot skeptics thinks a lot of these startups, these robotic startups, are kind of fake.
Starting point is 00:27:08 They're kind of performative. So, you know, you're looking at this robot folding these towels, and he kind of commented sarcastically in response to this video, does it still work if you raise the tape? table six inches. And so the founder of Figer decided to respond and said, anything else, Noam? And it's this video of the same robot folding these towels, but midway, one of the operators comes to the table and raises it six inches. You're watching that on the screen right now. And as you can see, even whilst he's raising the table, the robot just simply reacts to it.
Starting point is 00:27:46 It lifts its elbows up. It doesn't even hesitate, which is super cool. And what I found super funny. I love that. Dunk on the pessimist. But what I found funny is he took it a step further and absolutely roasted no one. That's a bad beat. And he pulls up a chart which OpenAI shared during their live stream announcing GPD5
Starting point is 00:28:09 where they basically committed something called chart crime, where they showed a statistic which implied that GPD5 was. so much better than previous models, but actually the bar charts were just drawn up incorrectly and it was actually not that incremental. And he captioned this with, should we raise the middle chart by six inches, question mark? Which I thought was hilarious. Yikes. Yeah, not a good look. Listen, everyone's building towards the same future. The open AI guys should know better than anyone else who are building towards an optimistic future support the people doing hard work, support the people who are doing cool stuff. Like, this is an impressive feat. They're folding laundry with
Starting point is 00:28:46 the humanoid robot for the first time. He got what he deserved. Play stupid games, win stupid prizes. I think that's that about that. So that, I think that's a wrap. That's the, that is the robotic Olympics, right? We just, I mean, it was funny because I had no idea this was even happening. And then out of nowhere, I just started seeing pictures of robots fighting each other and boxed. And I was like, oh, all right, that's kind of cool. Shout out to China. China is really asserting their dominance in the, the robot race. They, they frequently do these mega drone light shows. I'm not sure. Have you ever seen these? The drone shows where there's like millions of insane and also scary. I'm going to be honest. Yeah, it's like a subtle assertion of dominance, right?
Starting point is 00:29:22 Because, I mean, the implied future of warfare is with drones. They're showing off their ability to use drones. Now they're hosting the place, hosting the event where the future of humanoid robots will be spawned out of. And I mean, hey, China's doing their thing. And they're letting the world, they're letting the world know. But I think that that covers it. That is our humanoid robot 101. What I've learned, yeah, what I've learned, Josh, is that we need to get up. more on a Friday and Saturday night. Well, at least I do. And not stare at like robot live streams from China. Yeah, but think about it. I mean, five years from now where we're sitting here, you'll be able to say, yeah, I watched the Olympics where those robots, they were actually
Starting point is 00:29:59 slower than humans. Yeah, they were slower than humans and humans still controlled them instead of the robots controlling us. All right, five years from now, they're going to be teleoperating us. So, watch out for that. Thanks for making me feel better about myself. Yeah, there you go. It was time we'll spend, and we got a great episode about it. So thank you for informing all listeners. Thank you listeners for listening and sticking with us to the very end of the episode. We appreciate it. If you enjoy the episode, please don't forget to share it. Like, just get the word out about Limitless. We really, we enjoy making these episodes. A lot of new AI contents come, including an interview with that really cool robot maker with those crazy looking arms that'll
Starting point is 00:30:31 fold your laundry. So there's a lot to look forward to. So thank you for listening. We will see you guys in the next one. And I hope you enjoyed. See you guys.

There aren't comments yet for this episode. Click on any sentence in the transcript to leave a comment.