Pod Insights Moonshots with Peter Diamandis Robotics CEO: The Humanoid Robot Revolution Is Real & It Starts Now w/ Bernt Bornich & David Blundin | EP #188 About this episode Get access to metatrends 10+ years before anyone else - https://qr.diamandis.com/metatrends
David Blundin is the founder & GP of Link Ventures
Bernt Bornich is the Founder and CEO at 1X Robotics, a competitor to Figure and Optimus.
–
My companies:
Test what’s going on inside your body at https://qr.diamandis.com/fountainlifepodcast
Reverse the age of my skin using the same cream at https://qr.diamandis.com/oneskinpod
Apply to David's and my new fund: https://qr.diamandis.com/linkventures
–-
Connect with Bernt:
X
Linkedin
Learn more about 1X Tech: https://www.1x.tech
Connect with David
X
LinkedIn
Connect with Peter:
X
Instagram
Listen to MOONSHOTS:
Apple
YouTube
–
*Recorded on July 28th, 2025
*Views are my own thoughts; not Financial, Medical, or Legal Advice.
Learn more about your ad choices. Visit megaphone.fm/adchoices
Transcript 19,422 words · 4 speakers
Speaker names marked ? are inferred from speaking time and the show’s known hosts — they are not voice-verified.
0:00 Peter Diamandis? You think about robots in the world probably more than anybody else. What's your vision 10 years from now?
0:07 Speaker 01 First of all, what will happen is...
0:10 Peter Diamandis? Everybody, we're here at One X Technologies in Palo Alto. Bernd Bornich, the CEO, founder. NeoGamma 1 and NeoGamma 2 over here. I imagine we're going to have the same level of AI eventually in the robot, where I feel like I'm talking to a fully intelligent being.
0:26 Speaker 01 An humanoid is grounded, right? That actually understands what this existence is.
0:30 Peter Diamandis? I'm shocked by that.
0:31 Speaker 02 I'm shocked by that, too.
0:33 Speaker 01 How do we solve the remaining really hard problems in science? This is not going to happen without humanoid. It's almost existential to us for human happiness.
0:40 Peter Diamandis? So Saleem is constantly saying...
0:42 Speaker 02 Have it look like an octopus and let it operate in the all the elegance that an octopus can, rather than trying to constrain it into five fingers on this hand that do certain things and manipulate objects the way we're supposed to manipulate them.
0:54 Peter Diamandis? So what's the definitive answer to him?
0:56 Speaker 01 Let's just say humanoid is a face.
1:00 Peter Diamandis? Now that's a moonshot, ladies and gentlemen.
1:05 Peter Diamandis? That's me. Dave London, my moonshot mate. Hello, all. And NeoGamma, NeoGamma 1 and NeoGamma 2 over here. And we just did a tour of the facility and it's pretty extraordinary. We saw, you know, probably dozens of NeoGammas in different stages of development. They literally manufacture everything from head to toe. And how many components inside NeoGamma, roughly?
1:29 Speaker 01 Oh, top secret.
1:30 Peter Diamandis? Top secret, okay.
1:31 Speaker 01 But it's in the hundreds, not the thousands.
1:33 Peter Diamandis? Okay.
1:33 Speaker 01 We're pretty proud of.
1:35 Peter Diamandis? I just secured my first NeoGamma at my home by the end of the year, is that right?
1:40 Speaker 01 Oh, yeah.
1:41 Peter Diamandis? Okay, fantastic, great. So we're about to do a podcast, either with Bernd or with NeoGamma, depending upon what you want. And let's go ahead, we'll go over to the podcast area. Would you lead the way and maybe clear the way for me?
1:56 Peter Diamandis? Awesome.
2:00 Peter Diamandis? By the way, those bags over there, Neo Gamma can carry those.
2:08 Peter Diamandis? Yeah.
2:10 Peter Diamandis? Okay, I'm not the Neo Gamma. Can I give this to you to carry?
2:16 Speaker 01 You can try. It might hit some safety limits, but it usually works.
2:19 Peter Diamandis? All right, arms up.
2:21 Speaker 01 Figure it out properly. There you go. You can let it go and it can take a few steps. There you go. It might, after a few steps, decide that like, this is a bit unsafe for me. Thank you, Neo.
2:32 Peter Diamandis? Incredibly strong. All right.
2:37 Peter Diamandis? And it's nice to know that Neo Gamma will clean up the house around you.
2:44 Peter Diamandis? Yeah.
2:47 Peter Diamandis? Well, listen, I'm not sure what number you are, but I want to say thank you so much. Thanks for cleaning up.
2:52 Speaker 02 Of course. Thank you for your time.
2:53 Peter Diamandis? A pleasure. A pleasure. And thank you very much as well. I want to be polite. You know, you never know when the robot overlords are gonna like come after us. I want you to remember, I was really polite. I was really, really polite. Okay, I'm safe. Great. Have you ever been in love? I mean, you meet all these other robots. I mean, some of them gotta be turning you on. No?
3:13 Speaker 02 You should take a look at 41.
3:14 Peter Diamandis? 41. Okay, Gamma 41 is your gig. Okay, got it. Thank you. Okay. Oh, listen, Burnt's here. Let's stop this conversation.
3:22 Speaker 02 Sorry, sorry.
3:23 Speaker 01 Do behave.
3:24 Peter Diamandis? Everybody, welcome to Moonshots. I'm here with my Moonshot mate, Dave Blundin. Salim Ismail is offline with his kids this weekend. His son this weekend. But I'm here in particular with the CEO and founder of One X Technologies, Bert Bornek. A pleasure. Burnt? Awesome. Looking forward to this one. Thank you. Yeah, I mean, we just finished this tour and it's pretty extraordinary. When did you move into these facilities here recently? It's like one and a half months ago. Nice. Well, I mean, just many levels of people building robots. No robots building robots yet. We're getting there, but not yet. Yeah. So, I mean, when I, you know, I'm very familiar with the robotics, the humanoid robot space.
4:05 Peter Diamandis? And while companies like Figure and Tesla are focused initially going into factories, automotive factories in particular, you made a commitment to the home.
4:18 Speaker 01 Yes.
4:18 Peter Diamandis? And personally, I'm excited about that, but I'd like to start with why the home?
4:24 Speaker 01 There's like, to me, there's two, like, there's a lot of reasons, but there's like two main reasons. Uh, now, the first one is kind of obvious, which is just like, I mean, consumer hardware just scales at a different pace than everything else, right?
4:38 Peter Diamandis? Yeah.
4:38 Speaker 01 Like, uh, we got to more than a billion devices of the iPhone in like a bit more than a decade. Um, and to me humanoid robots does not make sense unless it's at scale, right? There's always a better automation system that you can use for one specific problem. You need scale so that you really get this incredible reliability, incredibly low cost, incredible ecosystem and intelligence. Now, the slightly deeper one is also that intelligence comes from diversity. And this has been very clear, actually, from all the way in the beginning, really, uh, in, uh, all kinds of AI research and also now more practical applications of AI across all different domains, where it's like a language model or an image model or a video model, or in this case, a robotics model.
5:24 Speaker 01 You don't really need data of the same thing over and over. Like, if you think, think about it, it's very logic, right?
5:29 Peter Diamandis? So if you're in the, if you're in an automotive factory, you're basically doing the same thing over and over again. You're not learning new stuff.
5:35 Speaker 01 Yeah, and we actually have some data on this. We have some real data because we, our previous generation humanoid Eve, we deployed that into both guarding and logistics back in, uh, twenty twenty-two, twenty twenty-three. And about twenty to forty hours, our robots kind of plateau and stop learning for that specific task. Depends on how complex it is, like if you're guarding a facility and you're kind of driving around, because that had wheels, but it was a humanoid wheels, opening the doors and like, there's some diversity to that. So then you're more on like the forty plus hours. And if you're just like moving this cup from here to over here all day, right? Then like you're in the lower end of twenty. Um, and.
6:13 Speaker 01 There's just no path from there to like general intelligence. And we are maybe kind of a bit different than the rest of the humanoid space in this. That I see as more as a company really running towards AGI. And how can we come there as fast as possible? Versus how can we apply labor in industrial or similar settings?
6:34 Peter Diamandis? So it's robotics in service of building true AGI models and getting enough new rich data to train up these models.
6:45 Speaker 02 You said 20 hours, 40 hours for security guard robot. What's the equivalent for all the variety of things you can do in the home? How many hours?
6:53 Speaker 01 We don't know yet.
6:54 Speaker 02 So tens of thousands.
6:56 Speaker 01 We don't, we do, yeah. So like at our current scale, we don't really see any kind of cap on diversity. It'll get there and we'll need to diversify. But I think like you ask a very important question, right? Because we want to talk about like, what is the goal? And like, to me, it's not just AI or robotics. It's a combination. Because if you think about what this is like, what is abundance, right? It's an abundance of knowledge or intelligence kind of multiplied by an abundance of labor or goods and services. You kind of need both. And they follow hand in hand. And we could talk more about that. But like, the constraints we have in society aren't always only on the intelligence or data layer.
7:41 Speaker 01 They also are on the substrate that we're building on, right?
7:43 Peter Diamandis? Every week, my team and I study the top 10 technology meta trends that will transform industries over the decade ahead. I cover trends ranging from humanoid robotics, AGI and quantum computing to transport, energy, longevity and more. There's no fluff, only the most important stuff that matters, that impacts our lives, our companies and our careers. If you want me to share these meta trends with you, I write a newsletter twice a week, sending it out as a short two minute read via email. And if you want to discover the most important meta trends 10 years before anyone else, this report's for you. Readers include founders and CEOs from the world's most disruptive companies and entrepreneurs building the world's most disruptive tech. It's not for you. If you don't want to be informed about what's coming, why it matters and how you can benefit
8:28 Peter Diamandis? from it, to subscribe for free, go to diamandis.com/metatrains to gain access to the trends 10 years before anyone else. All right, now back to this episode. So when I think about it, I imagine this is like why a toddler crawling around, playing, investigating the physics universe, it's in interacting with different people and different things is learning and building a model in its in its neocortex. And so it's is that basically the same your neo gamma is a is a infant learning in a diverse environment?
9:04 Speaker 01 It is. Yeah. And I think just like
9:10 Speaker 01 to some extent for humans through right also, but it's it's more pronounced in older animals, like how much of this kind of intelligence is innate and part of your instincts. You don't want your robot to just go around randomly doing anything. You wanted to try to do things that might succeed. So there there is room here for like the more kind of like call like classical AI models where we're training based on internet data, simulation data, synthetic data, everything that everyone else is doing that's useful to get you off ground. But it doesn't fully get you there. It gets you to something that does something seemingly kind of maybe useful. And then you can experiment, you can have the robot really have this interactive
9:52 Speaker 01 learning loop, where it's learning in the real world. And that can get you, we don't know how far it can get you, right? We don't, we don't know.
9:57 Speaker 02 And this whole topic of data gathering, you know, it's amazing watching them walk around the building here and walk around the kitchen. They're so unintimidating, you walk right up to it intuitively. You don't feel like it's ever gonna do anything awkward, hit you or anything like that. So that's got to be incredibly important.
10:14 Peter Diamandis? You're saying it's cozy, it's cozy.
10:16 Speaker 02 It's cozy and it's, you know, it doesn't, it doesn't seem to break the glasses or anything. So I mean, that's got to be really core to the data gathering mission, right? Because you have to, like you said, let it experiment, otherwise how's it gonna learn?
10:29 Peter Diamandis? So along that lines, what what design elements did you? Build into NeoGamma to make it for the home?
10:36 Speaker 01 Sure, this actually goes all the way back to the founding of the company a decade ago now.
10:39 Peter Diamandis? Yeah.
10:40 Speaker 01 Um, really, I've been in the field for a long time, so I kind of like started the company.
10:44 Peter Diamandis? How long?
10:44 Speaker 01 Um, since I was a kid. Like I, I.
10:46 Peter Diamandis? Building robots at age what?
10:47 Speaker 01 Um, I was eleven when I decided that I was gonna like do humanoid.
10:51 Peter Diamandis? And what was, what was your, what was your humanoid robot, uh, that you, that you modeled? Was it Star Wars? Was it Star Trek? Was it, what was it? Lost in Space?
11:02 Speaker 01 Honda Asimo.
11:03 Peter Diamandis? Which is it?
11:04 Speaker 01 Honda, Honda, Honda's Asimo.
11:06 Peter Diamandis? Asimov?
11:07 Speaker 01 Asimo, yeah. It's a beautiful robot, right? Um, they, they started very early and you can, you can check out like the Honda.
11:14 Peter Diamandis? Oh, yeah, yeah, yeah, sure, of course.
11:16 Speaker 01 Like, there's more modern ones, but like the Honda Asimo P6 was like end of the nineties.
11:20 Peter Diamandis? Yeah.
11:21 Speaker 01 And that was walking upstairs.
11:23 Peter Diamandis? Yeah.
11:24 Speaker 01 Running around the stage, giving someone a ball, like.
11:27 Peter Diamandis? It greeted President Obama, I think, at one point.
11:30 Speaker 01 That, that was a bit later, but yes.
11:31 Peter Diamandis? Okay.
11:31 Speaker 01 They, and like, it was so ahead of its time, right?
11:34 Peter Diamandis? Yeah.
11:34 Speaker 01 Uh, but I built a lot of stuff up through the, through the years, but I think importantly, when I started the company, I sat down and I thought really deep about this, like, okay, there's all these amazing robots that we worked on.
11:46 Speaker 01 And it didn't really work. Uh, why didn't it work? Right? And then it comes down to these like fundamental principles of, first of all, if you actually want to make something that's scalable with respect to intelligence, it needs to be able to live and learn among us. And there's just so many nuances to this through everyday life, right? Um, everything we do is social, like work is social, every task is social, and we navigate these social situations all the time while we do the things we do. And then most of the world's labor also happens in a social context in that there are other people around you when you do it. Objects have social context, right? The coffee cup is empty. Do you need a new, do you need a new, is it dirty?
12:28 Speaker 01 Or do you want to refill? Or do you keep your cup out for the day? And like, there's this like,
12:32 Peter Diamandis? Right.
12:33 Speaker 01 role of diversity that you want to access. So, uh, kind of if you're a big believer in that. Then it boils down to, okay, the robot needs to be safe. From my first principles point of view, not able to harm people. It still needs to be very capable, needs to be as strong as a human. Then it just needs to be incredibly affordable. Like, you need to find this beautiful combination where you can simplify, simplify, simplify, and still get a very capable system, so that you can manufacture this at scale and really drive quality up and cost down, right? So that was the founding principle of the company a decade ago. They said, like, we're gonna make robots that are safe, capable, and affordable. And by affordable, I mean, it's gonna be like, first principles, manufacturable and affordable.
13:14 Speaker 01 Very lightweight, very energy efficient, so you can have a small battery, very few parts, designing a model that doesn't require tight tolerances, no special alloys or materials, and just like incredibly simple but performant. That's really what I set out to do, and that's also why it took a decade, right?
13:33 Peter Diamandis? Yeah.
13:33 Speaker 01 Because like, there's so much novel research that's been done in the company to get to where we have these tandem-driven robots.
13:40 Speaker 02 So what's the vision there relative to the car, say? Like, one in every household, two? Like, and you mentioned the iPhone, you know, go direct to consumer, iPhone sales get to, you know, a billion, but it's exactly one per person, it's pretty obvious, right? But robots, it could be two, it could be four, it could be rare.
13:59 Peter Diamandis? I've done that poll, and everybody routinely says I would have at least two, depending on the price point, right? So, price point-wise, you know, when I think about this, what I've heard is, you know, thirty K, twenty K. We've seen Chinese robots in much cheaper price points, but not as capable as NeoGamma. Do you have a price point that you're thinking about?
14:21 Speaker 01 You're not far off. It's cheaper than what people think.
14:24 Peter Diamandis? Okay.
14:25 Speaker 01 And it's quite interesting, because like, I think this is very important.
14:31 Speaker 01 I wanna make sure that we are not only making the best product, we wanna be price competitive. I think that's gonna be incredibly important. And...
14:41 Speaker 01 We are actually still price competitive with the Chinese ones. But you have to count, like you said, it's not the same, right? So if you think about the number of degrees of freedom that the robot has, like how much capability, basically how many joints, then we actually have a significantly lower cost. So I think we've done a really good job on reducing complexity to get there.
15:01 Peter Diamandis? The numbers burnt that I keep in my mind is like thirty K purchase or three hundred bucks a month to lease, ten bucks a day, forty cents an hour. Am I in the right range there?
15:15 Speaker 01 I think we could do better, but yes.
15:16 Peter Diamandis? Okay, that's fantastic.
15:18 Speaker 01 And but I mean, do you need to do better?
15:20 Peter Diamandis? No, I mean, in a heartbeat, in a heartbeat, that's good enough.
15:23 Speaker 01 Yeah.
15:24 Peter Diamandis? But in that case, I think people could imagine having owning a couple of those robots.
15:29 Speaker 01 So I think it really depends on the lens you see this through. So I think clearly everyone's gonna want a robot. And I think this is a beautiful thing about the companion aspect of this, which is so underrated, right? Because the humanoid is just such a beautiful interface for AI. And when you talk to it and you like see the body language, it can look at you, it sees who's talking to it, direction, all these things. Like all my eleven year old daughter can do, if she has a robot, is like she just wants to sit next to it on the couch and talk about things, right? And that is clearly gonna be such a big aspect of it. And I see it as like, it's not, how do you say, it's not another pet. But it's not another humanoid. It's something kind of in between.
16:12 Peter Diamandis? Yeah.
16:12 Speaker 01 And like I said, like, it's kind of like the my Hobbes. Like, if you've ever read Calvin and Hobbes, it's the Hobbes.
16:19 Peter Diamandis? Yeah.
16:19 Speaker 01 Uh, and I think it's gonna be incredibly exciting to see how these relationships develop. Because it's the thing that will be like around you all your life, right? It will remember everything about you and it's gonna like.
16:31 Speaker 02 Things that are like, really, if you compare C three PO and that vision of a of an assistant robot, and you compare it to what you've actually built, the two things that jump out at me right away. One, it's soft, it's not like a metal outside. And two, the voice is perfect.
16:48 Speaker 02 Like when you're speaking to it, you immediately are disarmed and you just talk to it because it doesn't have a C3PO robotic voice. It has just a perfectly soothing, normal voice and it's very responsive to anything you say, any gesture or anything.
17:01 Peter Diamandis? So I imagine that these robots will all have advanced AIs at the level of, you know, a GPT-5 or a, you know, a Gemini 3. And in so being, those robots will be hyperintelligent and able to understand fully and answer what you need. And once they've learned the physics models fully, do whatever you need. You've made a decision to build your AI systems in-house. And I find that fascinating. And in fact, a number of the other robot companies, humanoid robot companies, not going to put you into a comparison mode here, but have made that same decision versus partnering with the large hyperscalers. Can you speak to that?
17:47 Speaker 01 Well, we're not doing the same thing. I mean, to me, intelligence does not begin with language. Like, language is this generative artificial construct that we have come up with. And it's incredible. I mean, it's such an efficient, compressed way of conveying meaning and instruction. So, language is very useful, but it's not the core of your intelligence. The core of your intelligence is spatial and temporal, and it has to do with how you perceive the world around you, both with respect to how you see the world, but also how you feel the world, right?
18:19 Speaker 01 And we're getting to where we're seeing that like models that are native to that modality, and then you add text, will be more intelligent and more powerful than the language first.
18:32 Peter Diamandis? I mean, I've read about, I've read about intelligence and the belief is that you needed embodiment for intelligence to exist and language for intelligence to scale.
18:44 Speaker 01 I don't feel that I can prove, I don't, I don't have rigorous proof that embodiment is needed.
18:51 Speaker 01 I do have very, very strong proof that from an engineering perspective, it's just a way easier path, right? So if you think about like the information in the world and can you access this?
19:07 Speaker 01 You could train a world model that can predict video and tell you like, hey, here's a new video frame, right? Render this for you. You could, in theory, you could probably train that only on text. Like if you have enough text descriptions of things, maybe at some point you could like get a high enough signal noise that you actually can get something useful out. At least if you kind of like have some feedback loop with like some RLHF or something where you're like, am I happy with this frame?
19:35 Speaker 02 Hmm.
19:35 Speaker 01 But I mean, why would you do that? That's just like such an inefficient way of doing that. You, of course, you train on video because you're gonna like output video, right? So from, from that perspective, I think it's just obvious that like you need all the modalities that we experience if you want to get to first and foremost, like human level intelligence and hopefully pass that again. But then I think there's one other thing about robot. There's two other things actually that are quite important when it comes to learning. And the first one is quite obvious and I think we all kind of identify this, which is like robots can do interactive learning, right? So you, you interact with the world and therefore you can learn.
20:14 Speaker 01 But if you think about it more from a kind of academic point of view of like how those intelligence kind of evolve, uh, how do you get reasoning all these things? Then
20:25 Speaker 01 what we generally do is that we have some observation of the world. Like we, we kind of know how the world works. So I know that if I do this, I know what is going to happen, right? I've seen this before. Yep. So, um, I actually start with that and I have a goal. I want to pick up the cup. So now I have a model of the world. I have a goal of picking up the cup. I take an action, I know which action I took. I know the action I took was to like reach for and grasp the cup.
20:53 Speaker 02 Yeah.
20:54 Speaker 01 And then I observe the result. If you look at the internet, or in general, you can look at YouTube, right? All you have is just the observations.
21:01 Speaker 02 Right.
21:01 Speaker 01 You don't have any of the mental model of like the person in that video. You don't know which actions they took. You don't know what they tried to achieve. You only have the observation.
21:09 Speaker 02 Yeah.
21:09 Speaker 01 This is not how we learn. You can actually bring it all the way back to the scientific method. It's like you should have like, you, you, you should have a theory. Come up with a hypothesis. You test your hypothesis, you observe the result. And then you do it again and you learn. And that is just not possible with internet data. So there's.
21:26 Speaker 02 Definitely, definitely impossible with the next token, you know, raw internet scrape and with all the video scrape. So then in these limited domains like coding and physics experiments, you can actually have that same experience, but it's only within that domain. Like coding is a good example, like, oh, let me try writing it this way. It didn't work. Let me try it that way. It didn't work. So you get very, very good at that narrow domain, still have no intuition about how the world works.
21:51 Speaker 01 No, you can, you can do simulation though. So again, like it's hard to prove that this won't work, right? So sure, if you have a really good simulator and you really scale simulation and learning and simulation with agents.
22:01 Speaker 02 Yeah.
22:01 Speaker 01 Maybe you can get something similar. Um, but I mean, the fidelity of your simulator is nowhere near the real world. And it's just like so incredibly hard to get there and close that gap.
22:11 Speaker 02 Hmm.
22:12 Speaker 01 And it's also so compute inefficient compared to just being in the real world. But I think for me, it boils down to not this academic exercise of like proving who's right and wrong. It's more, what's the engineering approach that makes sense here?
22:23 Speaker 02 Yeah.
22:24 Speaker 01 And it's just a way shorter path.
22:26 Peter Diamandis? You mentioned before in our conversation, the amount of data that's being collected relative to Google or YouTube or Tesla, can you speak to that? I mean, your, your mission is get as much possible data during the day of an interaction of these robots in the home.
22:43 Speaker 01 Yeah, I mean, you can do some napkin math, right? And of course, we don't know exactly like what is the most useful data from which modus etcetera yet. But if you think about it, if you have 10,000 robots out there and they gather data most of the day, then that is more data than the like non-duplicated useful data that gets uploaded to YouTube each day.
23:05 Speaker 02 Yeah.
23:06 Speaker 01 So already at that scale. You actually have like your fleet of robots generating more useful data than YouTube. So that's just a 10,000. And then if you think about like how we scale manufacturing here as this starts deploying into society, you actually very quickly come to the conclusion that like, you know what, the internet isn't actually that big. Like you're gonna have way more data from robots than you're gonna have from the internet.
23:28 Peter Diamandis? So I want to hit, I want to hit some numbers here just to set them as foundations. You built hundreds of the Neo Gamma at roughly, but you're about, you got a new manufacturing plant that you're about to open. Can you give us a sense of, and then another one that's in plans, right? Without disclosing anything you're not willing to, but can you give me a sense of by the end of 26, how many you're manufacturing on an annual run rate? And then in 27, 28, what's, what's the growth path you imagine?
24:02 Speaker 01 Yeah, first of all, just small correction. We didn't, we haven't built more than 100 of the gammas. But we've built, we've built more than 100 of the robots.
24:09 Peter Diamandis? Right.
24:10 Speaker 01 Um, there's been multiple versions, but um, the factory run rate end of 2026 is north of 20K.
24:19 Peter Diamandis? 20,000.
24:19 Speaker 01 Yeah.
24:19 Speaker 02 Annualized.
24:20 Speaker 01 Uh, annual, yep. Uh, of course, there's a ramp to get there. So you don't reach quite that number.
24:24 Speaker 02 So a couple thousand a month.
24:26 Speaker 01 Yep. Uh, now the factory after that is kind of like, we're trying to follow an order of magnitude, right? We're not going to quite be able to do that. Um, I think the iPhone ramp is a very good comparison here where you see like, they almost double, but like you have a few plateaus as you reach certain scales where you run into problems. And there are some quite interesting problems here if you're going to scale the manufacturing of humanoids to the iPhone level, right? Because you, you run out of some basic stuff like aluminium, um, for example. Uh, you, you don't use all the aluminium on the planet. That's not what I mean, but like, there's a certain amount of percentage of current refinement of aluminium you can use before you start to really struggle sourcing aluminium. Uh, and that might be a challenge, I think.
25:11 Speaker 02 The iPhone ramp was about doubling. This is an interesting stat I hadn't even thought of. I mean, you get to a billion.
25:16 Speaker 01 It was more like 1.7, but.
25:18 Speaker 02 1.7 annualized over, wow, that's not as fast as.
25:21 Peter Diamandis? So you can imagine a hundred.
25:22 Speaker 01 Well, exponentials are quite.
25:23 Speaker 02 Yeah, no, I, no.
25:24 Peter Diamandis? We've heard.
25:27 Peter Diamandis? So, but you can imagine a run rate before the end of this decade of hundreds of thousands per year.
25:34 Speaker 01 End of this decade, way more.
25:36 Speaker 02 Way, way more at that.
25:37 Speaker 01 Yeah, yeah. Now, at that point, you need to really think about like, what are the things that will slow you down, right? And it comes down to refining, like mining and refinement, of course. But increasingly, it actually comes down to labor. Like, you're not gonna get there without really using robots for labor. If you think about the iPhone ramp, then Apple kind of displaced large part of like the Chinese population across the country for labor, and they still, they still ran out of labor and had to expand into neighboring countries.
26:12 Speaker 02 That's wild.
26:13 Speaker 01 Um, now, I think we've done an incredible job in the design. So it's very few parts, it's very simple to assemble. But it's still, it's still more complicated to assemble than an iPhone.
26:23 Speaker 02 It's what I was gonna say, it certainly looks more complicated.
26:25 Speaker 01 Yeah, it is, it is more complicated than iPhone, right? So let's say it takes five times as long, so we need five times as much labor as the iPhone.
26:32 Speaker 02 Okay.
26:33 Speaker 01 Then you're in trouble.
26:33 Speaker 02 Good metric.
26:34 Speaker 01 Then you're in trouble.
26:35 Speaker 02 So it's gotta be solved.
26:35 Speaker 01 So like, you have to automate, right?
26:37 Speaker 02 Yeah.
26:37 Speaker 01 And, of course, that's the goal anyway. Like, we wanna get as quick as possible to what I call like this hard takeoff moment, where you have robots building robots, robots building out the data centers, the cheap fab, the energy infrastructure, and...
26:53 Speaker 02 So what can we learn from the car, actually? So you've got the iPhone, fewer parts, one-fifth the labor per unit. Then over here, you have a car. How does the part count compare to a car?
27:04 Speaker 01 So we have a few hundred parts. A car has fifty thousand, roughly.
27:08 Speaker 02 Fifty thousand, so it's much simpler.
27:09 Speaker 01 And I mean, a car weighs four thousand pounds.
27:12 Speaker 02 Yeah, a lot of material.
27:14 Speaker 01 We, our robot weighs sixty-six.
27:16 Speaker 02 Okay.
27:17 Speaker 01 So I think like, i- it's not really comparable to a car. I've seen a lot of like the space compare humanoids to cars. But I think then you should go back to the drawing board, to be honest. Like, it's not a car. If you do a really good job here, it's closer to a refrigerator. It's a very complicated refrigerator, but it's closer to a refrigerator than a car.
27:33 Peter Diamandis? Let's dive into a little bit of the, let's shape the understanding of the robot for our, our viewers and listeners. Uh, sixty-six pounds. Uh, let's talk about... Battery life, its abilities, you know, describe it from a specific stats point of view, if you would.
27:51 Speaker 01 Oh, yeah, sure. So I think first of all, I think the most important stat is huggable.
27:55 Peter Diamandis? It's huggable. Yes, it is huggable. I have hugged a robot, yes.
27:58 Speaker 01 Um, and like this is a safety and how it is to feel safe in its space, soft.
28:05 Speaker 01 But from a pure stats point of view, it's sixty six pounds. It can lift about a hundred and fifty pounds.
28:11 Peter Diamandis? Which is amazing. I mean, in terms of the the weight strength ratio.
28:15 Speaker 01 It it is the weight to strength ratio of an athletic human.
28:19 Peter Diamandis? Yeah.
28:19 Speaker 01 Um, and then it can carry like about fifty pounds around. That's what you hopefully saw earlier here. Um, and battery life is about four hours.
28:30 Peter Diamandis? Rechargeable in half an hour, half an hour?
28:33 Speaker 01 Half hour, so like two hours if you use a full battery. Uh, now, actually, interestingly enough, I have one in my house, right? So I I'm starting to get some get some data on this now. And.
28:43 Peter Diamandis? It's it's five foot four, five foot five, what is it?
28:47 Speaker 01 Yeah, five foot four.
28:47 Peter Diamandis? Five foot four, okay.
28:49 Speaker 01 I think so.
28:49 Peter Diamandis? It's a perfect height, by the way, just in case you're wondering.
28:51 Speaker 01 Yeah, it's also the height of my wife. It's also the height of my wife. So it's like, I I agree with you.
28:57 Peter Diamandis? It's it's mine, so that's good.
28:58 Speaker 01 Yeah, so it's, but I mean, it's, um, what's very interesting, like once you start actually using the product, right? You you notice a lot of things that you don't usually show up on a spec sheet, like the robot is completely quiet. And that that's not a coincidence. That's something we worked so hard on. And the first time you put this in your home and you think like the robot is very quiet, it's fine. You put it in your home and you're like, first day it's fine. Second day it's a bit annoying, like third day you're like, oh, man, is it gonna leave my living room soon? Because like this sound, right? It's such a requirement for like just dead quiet, right? We're gonna have this in your space. Charging wise?
29:34 Speaker 01 Don't really ever run into the problem. Because the robot just takes like these micro breaks every now and then where it's not doing something. And like, I actually don't care that much about how many hours it can run. I care about that it charges fast enough that it can just always do whatever I want to do.
29:46 Peter Diamandis? Nice.
29:47 Speaker 01 Um, yeah. Well, and I want to talk about quickly, since you said that, specifications. Uh, like the number of degrees of freedom, right? Which basically is how many joints does the robot have, right?
29:57 Peter Diamandis? Yes.
29:58 Speaker 01 So like, humans have like six joints in their each leg, that's 12. You have seven in each arm, so that's 14 more. So now you're like 12 plus 14, that's 26. You see all our robots today that have 26, that's quite common. Usually they all have their wrists, they have the, they actually have the neck instead. So two here and then you're like at 26. Uh, we have three here. So you have proper expression with your head, it's quite important. Um, we have all the seven from here, we have three in the spine. And then of course, we have 22 in each hand.
30:28 Peter Diamandis? I mean, the, what I saw in the arm design was incredible.
30:31 Speaker 01 So. How many, how many do humans have in their hand? 22. So you matched it. Well, okay, depending on how you count your capillary bones. So like the, the small bones that you have here that allow you to cup your hand.
30:42 Peter Diamandis? Yep.
30:42 Speaker 01 Um, you could to some extent see that that likes, that that's more like four or five degrees of freedom. Not, not really two. Uh, so then the humans have a bit more, but um, functionally, it's quite similar. Um, and this again, just is incredibly important to be able to do all those tasks in a home. But also from an AI perspective, there are like, we talked about diversity initially, right? It is the one metric for intelligence. And the diversity.
31:08 Peter Diamandis? Diversity of environment and data.
31:10 Speaker 01 Well, diversity of your data and your diversity comes from two things. So the limit to the diversity you can achieve. It comes from the environment you're deploying in. So right, if you're in a factory doing the same thing every day, it doesn't matter how good your robot is, it's not going to be diverse. And then, how capable is your robot? How many things can it do, right? Because if it cannot do any kind of like in-hand manipulation or handling like soft deformables, all these kinds of things, or delicate objects or whatever, then you get no data of that. So like, you really have to kind of go max, max on both, right? If you want to maximize your diversity.
31:43 Peter Diamandis? It was about 18 months ago that I partnered with one of my closest and most brilliant friends, Dave London, to start Link Exponential Ventures. At Link, we manage about a billion dollars of seed stage money, based out of Kendall Square in Cambridge, right between MIT and Harvard. When Dave and I both graduated from MIT, each of us immediately started companies. But at that age, everything is working against you. You have an idea, you're challenged to raise money, and you can't afford rent. And even with all the accelerators out there, you're competing against thousands of other startups for the same pool of investors. Both Dave and I have spent a big chunk of our lives focusing on how do we inspire and support founders to knock down those barriers, to go big, to create wealth, to impact the world, to build and scale as fast as possible,
32:28 Peter Diamandis? especially in today's AI everything world. We're seeing so many companies reaching multi-billion dollar valuations in just two to three years faster than ever before. Some companies are adding millions or tens of millions of dollars of value in just weeks. So we started asking ourselves, how do we help these founders go faster and not skip a beat? As an example, a couple of months ago, we bought an apartment building adjacent to MIT where a graduating entrepreneur can move in immediately without slowing down their tech build while they search for a place to live. And so we're doing everything we can to accelerate builders and their super smart teams. Of course, funding is part of it. Mentoring is part of it. Connecting them with my personal network of abundance minded CEOs and investors is part of it.
33:12 Peter Diamandis? We house 66,000 square feet of purpose-built incubator space and 26 AI startups call Link XPV their home. And the returns have been amazing. I have nothing to ask, but if you are building a company in the AI era, check us out at linkventures.com. Now, back to the episode.
33:31 Speaker 02 Yeah, a geeky question for you, but really, really curious to know. Because when you build something physical and then you attach a neural net to it, it's actually very hard to tell whether the constraint in what it can and can't do is in the neural net or in the physical construction of the hardware. Is there any way to decouple that and debug? Is it two different sizes? Or is it just incredibly impossible? I mean, once it's meshed together, you just can't.
33:54 Speaker 01 Well, we have a pretty good neural net here.
33:56 Speaker 02 Yeah.
33:57 Speaker 01 So usually the way I approach this is, can we do it in teleop? And if we can, the right neural net can do it with enough data.
34:04 Speaker 02 Interesting.
34:04 Speaker 01 And that's generally been proven to be true. Like, if we manage to do something teleop, it's just like. Okay, now we need a lot of diverse data of similar tasks. So we get some transfer learning and we need a lot of that specific task. And then almost irrespective of like how complicated that task is, you can get it to work.
34:21 Speaker 02 Yeah.
34:21 Speaker 01 Um, now, of course, that doesn't mean you can get everything to work with generalization across. We're not there yet.
34:26 Speaker 02 Yeah.
34:26 Speaker 01 But you can see that like, okay, you can get the neural network to do this. Now we need to scale it. So we kind of get this beautiful transfer of knowledge between tasks and like auto distribution, generalization, and all these things that we currently see in L L M's that we don't see that much in robotics yet. We have some pretty cool stuff internally where we see some.
34:43 Speaker 02 The one I'm picturing though is you, you ask it to make crepes Suzette or you, you ask it to do microsurgery and it can't quite do it. And then you say, well, look, the hardware guy is claiming the hardware is good enough. It must be the software guy.
34:57 Speaker 01 Hmm.
34:58 Speaker 02 And then the software guy is saying, no, no, no, the software, the neural net is fine. The hardware just can't do it. And then they fight it out.
35:05 Speaker 01 And then we just say like, well, bring it, we bring in our best tell operator and we say like, he can do it. Then the hardware can do it clearly.
35:13 Speaker 02 Clearly.
35:13 Speaker 01 Proof of existence.
35:14 Speaker 02 Yeah, okay, so that's, that's where I was going. So you have, you have a remote operator option.
35:17 Speaker 01 Yep.
35:18 Speaker 02 Who can control the hardware. Oh, that's really interesting.
35:21 Speaker 01 So then you again like, well, but we're getting to where this gets hard. Where we can kind of no longer do this.
35:26 Speaker 02 Because?
35:27 Speaker 01 Because the hands are just so good.
35:30 Speaker 02 Yeah.
35:30 Speaker 01 And they have very high fidelity tactile feedback.
35:34 Speaker 02 The human hands are so good.
35:35 Speaker 01 No, the robot hands.
35:36 Speaker 02 Okay.
35:37 Speaker 01 So they have, the humans' hands are still even better, but like the problem there is the robot hands are really, really good. And they have really fast, highly detailed tactile.
35:44 Speaker 02 Yeah.
35:45 Speaker 01 And we can't really transfer this efficiently enough from the human.
35:48 Speaker 02 Yeah, because the tell operator has, is using some kind of clumsy.
35:51 Speaker 01 I mean, uh, back to the X price, right?
35:53 Speaker 02 Yeah.
35:54 Speaker 01 The avatar challenge.
35:55 Speaker 02 Yeah.
35:55 Speaker 01 So it's a really hard problem to transfer that fast enough.
35:57 Speaker 02 Yeah.
35:58 Speaker 01 So now we start to see that the robot actually learns how to do manipulation way better from reinforcement learning in real.
36:03 Speaker 02 Yeah.
36:03 Speaker 01 So you're like, actually have the robot like interactively learning in real how to handle objects. And it can do things that the operator could just dream of.
36:09 Speaker 02 Ah.
36:10 Speaker 01 So that, so now we're kind of screwed. No, we don't, no, we can't do that anymore.
36:14 Speaker 02 All right, let's, I want to talk about, uh, three things in, in, in sequence. Uh, teleoperations versus full automation.
36:21 Speaker 01 Uh huh.
36:22 Peter Diamandis? Safety in the home and privacy in the home.
36:27 Speaker 01 Yes.
36:27 Peter Diamandis? Because those have got to be critically important as you're entering the home. So the robots, we saw the Neo Gamma out here operating in tele operator mode, but also in AI, full AI mode, right? And it was able to do both and its AI systems are going to increasingly get better and better and more capable. Again, as we're talking, as I'm talking to Gemini three or Grok four or, you know, GPT five soon.
36:56 Peter Diamandis? I'm talking to a highly intelligent human and getting a feeling that it understands what I want. And it's able to, you know, sort of like take action on my requests. I imagine we're going to have the same level of AI eventually in the robot, where I feel like I'm talking to a fully intelligent being in one sense.
37:17 Speaker 01 Oh, yeah, clearly. Very close already. It is. And all that is grounded, right? That actually understands to some extent what this existence is. Which today's L L M's are kind of like, they have this kind of abstract notion of it, but it kind of like, it's a facade that kind of quickly falls away if you start to probe at it. But that will get there, I think.
37:39 Peter Diamandis? In the teleoperations mode, you've got humans wearing VR headsets and using haptic controls?
37:47 Speaker 01 No.
37:48 Peter Diamandis? What are the humans doing?
37:50 Speaker 01 They're giving slightly more like high level commands. So just guiding like, hey, put your hands over here, like grasp this thing.
37:57 Speaker 01 You don't want like over constrained a system. You want to give it some opportunity to kind of like solve for how to do the task.
38:02 Peter Diamandis? Okay.
38:03 Speaker 01 So we, we kind of have like the, the learning coming up from the bottom and kind of like enabling a more and more abstract interface for the operator. And then we have the learning of like the, all the large amounts of data we have coming kind of from top and getting more and more like the general behavior that you want the robot to do. And they kind of like meet in the middle, right? Where the operator goes away.
38:23 Peter Diamandis? You're using full, you're using automation and teleoperations always together in that regard.
38:28 Speaker 01 Yes. So everything that enables the robot to do anything that the operator does, teleoperator, is fully learned end to end. Like the, the network outputs torques to the motors. That's.
38:40 Speaker 02 That's that's very similar to what Tesla and Elon Musk were saying where the self-driving car was originally all C++ code with a little bit of neural net, maybe 80% C++ 20% neural net. Then every year that went by became more neural net and now there's no-
38:53 Peter Diamandis? 300,000 lines of C++ were eliminated.
38:56 Speaker 02 Yeah, just a few guardrails left and the rest is just one neural net. So same thing here.
39:02 Speaker 01 It's all weights, right? Like the code is just a few hundred lines.
39:07 Speaker 02 Is it really?
39:08 Speaker 01 Yeah.
39:09 Speaker 02 What's the parameter count or is that all super secret?
39:13 Speaker 01 It's kind of secret, but it's it will be small if you compare it like today's neural networks because it's running on the robot very fast.
39:21 Speaker 02 Yeah.
39:22 Speaker 01 It's kind of like your muscle neural system, but it does take in vision, so it's not very small.
39:27 Speaker 02 Well, that begs a question I'm dying to ask, which is, you know, you've seen X Machina, right? When I saw that movie, I'm like, why?
39:35 Peter Diamandis? Don't go dystopian on this here, okay?
39:37 Speaker 02 Why is the the brain, the blue blob, in the head?
39:41 Speaker 01 Yeah.
39:42 Speaker 02 Why isn't it in the server room?
39:44 Peter Diamandis? So learning is shared between all robots.
39:47 Speaker 02 Well, yeah, and it can be much bigger and, you know, like if half the power of the robot is going into the thinking, you could say you could run twice as long on a battery charge if you move it over to the server room and have it just communicate remote. So why did you choose to put it in the head? Aside from being anthropomorphic and cool?
40:03 Speaker 01 No, no, it has nothing to do with that. So there's some simple answers to that, which is, I mean, the head is where nothing else is unless you put the brain there. Like the rest is, yeah, like it's like everything else is pretty freaking full. Like it's building a humanoid with this kind of like power level in such a miniaturized form and still having like enough space to make it like completely soft and all this. It's really hard engineering problem. So it's like, where are we going to put this if we don't put it in the head? If you're going to put it on the physical robot? Now, there's some other arguments. The very high bandwidth thing that happens in your brain is vision and to some extent audio, smell, right? Tactile, but vision just dominates.
40:47 Speaker 02 Yeah.
40:48 Speaker 01 And you just want to minimize the distance between your eyes and the compute.
40:51 Speaker 02 For real? So the bandwidth between the sensors, the eyes mostly wouldn't make it over the home wifi?
40:58 Speaker 01 Well, it wouldn't even make it down to the stomach of the robot.
41:01 Peter Diamandis? Really?
41:02 Speaker 01 Um, without getting overly complicated on like which physical interfaces you would choose for this transfer. It's very high bandwidth.
41:09 Peter Diamandis? I'm shocked by that.
41:11 Speaker 01 Shocked by that too. But I mean, we're running like no LiDAR, no structured light, no wrist cameras, no nothing. We're running pure like emulation of human vision, right?
41:20 Peter Diamandis? Yeah.
41:20 Speaker 01 So we're relying so heavy on that. So it's a very, very high resolution, very high bandwidth, very high frequency.
41:27 Peter Diamandis? That's fine, because that's exactly where the human brain is very close to the eyes too.
41:30 Speaker 01 It is. Now, that doesn't mean that you can't do things in the cloud, and we do things in the cloud, but it kind of becomes hierarchical from an intelligence point of view. Just like if you think about like your, um, your, uh, kind of like your muscle nervous system, this runs quite fast, right? It usually runs at like 25 hertz. Um, and it doesn't necessarily, it doesn't go up to your brain. There are neurons distributed out through your system that makes decisions.
41:55 Peter Diamandis? Right.
41:56 Speaker 01 Uh, we have this in the robot. Uh, we have some of our stuff pushed to the, um, power electronics that controls.
42:03 Peter Diamandis? Just for latency, just for speed.
42:05 Speaker 01 Yeah. And then you have the brain itself, which actually runs pretty fast, right? It's usually like between five and 10 hertz and very, even though it's five to 10 hertz, very low latency. And this runs on the robot. Now, if you're running like more like a one hertz streaming thing than typically in L L M, first time to token land, right?
42:23 Peter Diamandis? Yeah.
42:23 Speaker 01 That runs off board, but that, that can't solve the like high frequency tactile feedback manipulation tasks. That's too slow.
42:30 Peter Diamandis? Okay. The first time my Neogama learns to crack open an egg to make an omelet. The question is, do all Neogamas then learn that? Are you shared learning?
42:40 Speaker 01 They do. Now, there's a shared learning in the sense that you can say like this data goes to the cloud model that is doing this for all Neogamas, but there's also the distributed models. So of course there would be like a nightly checkpoint where like, hey, this model is better. We have more data. We validated this. We get it to safety, which I'll talk about later when it comes to how do you validate the models? Um, and then we deploy that to all the robots. So even though it's distributed on the robots, they can still learn from each other, of course. It's just you need to do like one hop through the server layer and like do the training and propagate this out. Um.
43:13 Speaker 01 There is a future not so far away where I'm pretty bullish on there being a lot of federated learning happening on device. And this has to do with how do we have your companion, really, throughout life, learn from all of the experiences that are there to you but private.
43:31 Peter Diamandis? Yes.
43:31 Speaker 01 So all robots will not be the same, but they will share an intelligence backbone.
43:35 Peter Diamandis? Let's go into the conversation of privacy and safety. So you're inviting these robots into your home and where there will be activities that you may not want shared with the world. And then, of course, you're asleep and the robot is running tasks at night. You don't want to wake up in the morning and find your safe has been opened and the robot's gone, you know, to talk about. Or you don't want the robot to be, you know, taking care of your aging mother and find out that it's, you know, giving her, giving her shots of scotch at night when she asks for them. I mean, so how do you deal with safety and privacy?
44:14 Speaker 01 The last one is the hardest one, by the way. We can get back to that.
44:17 Peter Diamandis? Okay.
44:17 Speaker 01 Multi-gramma.
44:18 Peter Diamandis? Scotch.
44:19 Speaker 01 Scotch. Because, you know, generally, models are, they're always kind of tuned to be kind of sycophants and they end up doing whatever you ask them to do. But so if we start with the privacy side, I think first of all, it's just, it's a lot about transparency. Like, if you're one of the first people like you, Peter, that will have a Neo Gamma in your house.
44:39 Peter Diamandis? Can't wait.
44:39 Speaker 01 We are kind of trading a bit on privacy versus being an early adopter, because without the data, we can't make the product better. Of course, we're going to do everything we can to make sure this is privacy on your terms and that you are in control. But we do need your data if we're going to make the product better.
44:54 Peter Diamandis? Sure.
44:55 Speaker 01 So.
44:56 Peter Diamandis? I mean, I, listen, I give my data to Google, to Amazon, to X all the time. And I mean, people don't realize that you're sitting in the home having a discussion with your spouse and Amazon's, you know, Alexa's listening, right? Siri's listening.
45:11 Speaker 01 But they're doing something very important, which we also do, which is no human in our company can hear or see that data.
45:17 Peter Diamandis? Yes.
45:18 Speaker 01 That is. Going into the training mode, yes. But it doesn't go by a human.
45:25 Peter Diamandis? Right.
45:25 Speaker 01 Now, if we want to look at that data. And sometimes you might need to, right? Might be like, let's figure out what happens here because something clearly is happening across multiple robots that we want to figure out what is. Then we'll send you a notification on your phone or say like, hey, this specific window. We want to review the data and you'll get a video of like what that data is.
45:44 Peter Diamandis? Yeah.
45:44 Speaker 01 And then if you say yes, then we get the decryption key and we can look at the data. If we don't, if you say no, then we can't.
45:52 Peter Diamandis? Yeah.
45:52 Speaker 01 So that you're in the control of that. And actually, even with respect to like going into the training data, we always run like a 24 hour delay on training. So if there is something that you really don't want, even in the training data, like this never happened. Erase it from existence. You can go in and delete it before it gets into the training.
46:11 Peter Diamandis? I just want everybody here, there is a, there are policies and plans that make this acceptable and are are used by technology companies and you're going to be implementing the best of those.
46:23 Speaker 01 It's pretty cool. A lot of them. But there is like, so the mode I talked about now is when the robot is what we call best effort autonomy, right? Which is most of the time. Which we saw earlier today, where you talk to it, you ask it to do something, hopefully it does the right thing. If it doesn't do the right thing, then you can say bad robot and hopefully it's better next time. But it's very, it's actually, this is learning in real, right? Uh, this is really like interactive learning and the robot, interesting enough, actually progresses faster on tasks when it fails than when it succeeds.
46:51 Peter Diamandis? Sure.
46:52 Speaker 01 It learns more from failures, just as we do. But in this mode, that's the privacy. Now, when it comes to teleop, then of course, there's no way you can do this task without seeing the glass. So we do some abstractions so that you actually don't see people, people. You can like just see blobs and like you just see the object you're interacting with. And we can do a lot on interaction or on like the filtering side to ensure privacy. But the most important thing we do here is that no one goes into teleop in your robot unless you approve it, right? And it's very visible on the robot, like the lighting changes and like it's like someone is in your robot and it's one of the pre-selected operators that you have approved from like a large set of operators, like here are the four that services you. So that's kind of like inviting your cleaner or whatever into your house.
47:37 Peter Diamandis? Another human.
47:38 Speaker 01 Another human into your house and you just need to make sure that they're actually invited.
47:42 Peter Diamandis? So to to actually take a second and spell this out in more detail, in the early days when I have NeoGAM in my home, it'll be baseline autonomous, but there will be times where it needs to bring in teleoperator and so you'll have teleoperators in headquarters that if it needs help or doing something complicated or it gets something wrong, the teleoperator can step in and actually make the task happen.
48:10 Speaker 01 Yeah, it's in the beginning, it's actually there's two different modes. So so you have the mode which I call like the best effort autonomy that we just talked about.
48:17 Peter Diamandis? Yeah.
48:18 Speaker 01 And then you have task scheduling, which like my robot at home now is doing that. So I take my phone and I schedule and say like, hey, between these hours, here are the tasks I want you to do for me today is like do my white laundry. And then there's a package coming from me in Stuttgart, I can receive it at the door and unpack it in the fridge. And it's just like generally tidy. And I've given it when I'm not home, like these hours I'm at work, just get it done, right? Now, I don't care if that happens autonomously or to a teleoperator, right? So a lot of that happens for a teleoperator because some of these tasks are quite complicated and we don't know how to automate them well enough yet. Now, of course, that teleoperator uses autonomy to help improve the efficiency. So it's not all teleoperation, but I don't really care about the mix. The task gets done.
49:04 Peter Diamandis? Yes.
49:04 Speaker 01 Um, so we kind of split it like that. And then there's a gray zone kind of in the middle. If it's like you wanna like, I don't know. Having your friends over for a party and you want the robot to be the bartender? And we don't have like a bartender mode yet. Uh, then you can approve a teleoperator to do that. So you can.
49:20 Speaker 02 And all of the, in most of all the videos we see of Optimus uh, at, you know, Tesla's diner or at their events are teleoperation ones.
49:30 Speaker 01 They are, but I think teleoperation has gotten this kind of like underserved, bad reputation or name.
49:36 Speaker 02 Why?
49:36 Speaker 01 Um, I think this because people don't have enough like clarity and like, hey, is this teleopt or is it autonomous? Um, but it is just labeled data. It's expert demonstrations, right? If you look at any of the big AI models that were trained, there was an enormous amount of people that sat down and hand labeled data and like looked at examples, wrote it out like question answers and like bootstrapped kind of like this highly uh, high, very high quality data set for this to work, right? So you pre-chain on like general information. We also do that, just everything that's happened with the robot. Then you have a fine-tuned data set that is very high quality.
50:14 Speaker 02 Yeah.
50:14 Speaker 01 And in robotics, that is teleoperation because it's the expert demonstrations. It's the hand labeled data.
50:19 Speaker 02 Yeah.
50:19 Speaker 01 It's not different. It's just, I think there, there's some lack of transparency in what's going on.
50:24 Speaker 02 Well, I think the, the objection is if you have a demo, like a video, it makes it look like it can do something and it actually can't because you, you hand coded it.
50:32 Speaker 01 Well, it clearly can, but it can't do it autonomous.
50:35 Speaker 02 Yeah, I can't do it, but yeah, but I think you're, you're dead on that the, if it can physically do that, if the mechanism can do that, the neural net will fill in that blind spot instantly anyway. You know, once you've trained it, so I think it's, it's perfectly legit.
50:47 Peter Diamandis? And now it's time for probably the most important segment, the health tech segment of moonshots. It was about a decade ago where a dear friend of mine, who has incredible health, goes to the hospital with a pain in the side, only to find out he's got stage four cancer. A few years later, fraternity brother of mine dies in his sleep. He was young. He dies in his sleep from a heart attack. And that's when I realized people truly have no idea what's going on inside their bodies, unless they look. We're all optimists about our health, but did you know that 70% of heart attacks happen without any precedent, no shortness of breath, no pain? Most cancers are detected way too late at stage three or stage four. And the sad fact is that we have all the technology we need to detect and prevent
51:32 Peter Diamandis? these diseases at scale. And that's when I knew I had to do something. I figured everyone should have access to this tech to find and prevent disease before it's too late. So I partnered with a group of incredible entrepreneurs and friends, Tony Robbins, Bob Hury, Bill Cap, to pull together all the key tech and the best physicians and scientists to start something called Fountain Life. Annually, I go to Fountain Life to get a digital upload, 200 gigabytes of data about my body head to toe, collected in four hours. To understand what's going on, all that data is fed to our AIs, our medical team. Every year, it's a non-negotiable for me. I have nothing to ask of you other than please become the CEO of your own health,
52:12 Peter Diamandis? understand how good your body is at hiding disease, and have an understanding of what's going on. You can go to fountainlife.com to talk to one of my team members there. That's fountainlife.com. I want to jump into another fun subject, which is the uncanny valley and the face. So, I mean, you've probably had endless conversations internally about how do you make a face look? How human do you make it? How skin-like do you make it? How do you represent it? Can you, can you tell us sort of philosophically, you know, what you and Dar, who on your design team, how do you think about that? Where do you, where do you make it human enough?
52:57 Speaker 02 Yeah.
52:58 Speaker 01 It is this very delicate line where you wanna make sure like body language comes across crystal clear. Because that's like, that's the magic of the device, right, of the companion.
53:06 Speaker 02 Yes.
53:07 Speaker 01 But at the same time, you don't want it to get like to where your kinda instincts tell you, hey, something's wrong. Like this is a human, but there's something wrong with it. You you don't want it to be a human. And it is, it's actually pretty surprising that there's this, there is this gap where people clearly identify this as like, hey, this is a being I identify with, I understand his body language and everything, but he's also clearly not a human.
53:33 Speaker 02 Yeah, yeah.
53:34 Speaker 01 And you wanna be in that space. And then where you are in that space, kinda like depends a bit on like who you ask, right? People have like a different threshold here. Uh, so we're trying to hit kinda like in the middle of that and ensure that for as many people as possible, this is just like an incredibly easy to understand product. But at the same time, that is not creepy. And I think adoption here, by the way, talking about scale. Adoption is so important. And adoption of new technology usually takes some time because there's just like this knowledge barrier, right? There's a barrier to entry, even using a phone, there's a barrier to entry.
54:14 Speaker 02 Yeah.
54:15 Speaker 01 This interface is just so natural. Like there is no barrier to entry.
54:20 Speaker 02 Yeah.
54:20 Speaker 01 It's something you just talk to, like a person.
54:22 Speaker 02 You know, what's incredibly cool to me is that there are, there are like fifty things around the house that I don't know how to do.
54:29 Speaker 01 Hmm.
54:29 Speaker 02 Including the frickin', the other way to backwash the pool, like all this crap. The robot can in real time access the information, learn how to do it and just do it.
54:40 Speaker 01 Yeah.
54:40 Speaker 02 There is, I can't do that, it would take me an hour to study. And there's no laborer that's gonna come into the house and do it for under like four hundred bucks. And so it's like, like, there are so many things that are in that category where I'm not trying to replace a human being. I'm doing something that there literally was no other option for because the knowledge is obscure. And there's so many of those things around a house now, like resetting the, the, the water heater keeps going out and the reset process. But you, you can look it up, the robot can look it up.
55:10 Speaker 01 Yeah.
55:10 Speaker 02 And just go do it. And, you know, it's like.
55:13 Speaker 01 This is micro units of work essentially, right? Like you, you need five minutes of.
55:18 Speaker 02 But hyper, hyper spatial, hyper specialized micro units of work.
55:22 Speaker 01 Yeah. And you need like five minutes of it every now and then, and it's super high value to you.
55:26 Speaker 02 Yeah.
55:26 Speaker 01 Um, and it's just really hard to get to.
55:28 Speaker 02 Shop back, you know, the, the shop back, you can run it forward or backward. It's, there's a manual there, you could read the manual. I just want to get like this crap off the garage floor. The robot will know how to shop backwards because somebody else's robot, one of the other 10,000 has already done it.
55:42 Peter Diamandis? Make my perfect teriyaki salmon on the grill.
55:45 Speaker 02 Yeah, some obscure mixing, some food.
55:47 Speaker 01 Which brings us to something that you talked about earlier.
55:49 Speaker 02 Scotch, grandma.
55:50 Speaker 01 Scotch for grandma and safety. So, um, I do hope to make you a perfect salmon teriyaki.
55:56 Speaker 02 Thank you.
55:56 Speaker 01 Um, but I have to do it myself. I'm not gonna let the robot do it. Uh, because that's one of the things we're actually not doing, uh, when we're launching now, and that is due to safety.
56:06 Speaker 02 Okay.
56:06 Speaker 01 Because what I worked so hard on, right, for this decade is to make robots that are safe intrinsically. And what I mean by that is just like if something goes really wrong and accidentally it hits you. That might be painful, but it's not gonna be likely to severely harm you.
56:22 Peter Diamandis? Injure us, right?
56:22 Speaker 01 Yep. And once you pick up a kettle of boiling water, there's no more guarantee that you are safe, right?
56:30 Speaker 02 Hmm.
56:31 Speaker 01 So we generally avoid any kind of dangerous objects so that we can ensure safety in the beginning. Now, of course, over time, as the AI improves and we get more and more certainty on all behaviors being safe. We will allow cooking and other things. So we're doing internal projects on this. But we're not gonna be rolling it out to the customers in the beginning, just due to, due to safety concerns.
56:52 Speaker 02 Yeah.
56:52 Speaker 01 Um.
56:53 Speaker 02 Yeah, cooking, cooking and safety is a.
56:56 Speaker 01 It, it's a real problem for humans.
56:57 Speaker 02 Fire is not an easy thing.
56:59 Speaker 01 I mean, it's a real problem for humans too. It is. Um, but there's the notion of intrinsic safety. This is incredibly important. And then it's the safety of the AI. And this is the reason, we have a white paper on this that I recommend if you guys are interested, read it. But why we have started very early betting extremely heavily on role models.
57:20 Speaker 02 And role models.
57:21 Speaker 01 World.
57:22 Speaker 02 A world models.
57:23 Speaker 01 World models. They are, of course, the currently best known path towards ATI. Um, but even more importantly for us, like short term, as we progress here on like the data collection and model training. They give us this incredible opportunity to automate evaluation of models, including safety and red teaming and all these things.
57:43 Speaker 02 Hmm.
57:43 Speaker 01 So you can think about like, if you train a new model, and now you want to know if it's better than the previous one. And you can deploy it to all your customers and you can get some vibe check a few days later, like, hey, are people more happy now? Generally how it's done.
57:56 Speaker 02 Yeah.
57:57 Speaker 01 You don't want to do that with a physical system, right? You can't do that with autonomous car either.
58:01 Speaker 02 Yeah.
58:01 Speaker 01 What the world model actually is, it is a model that is able to generate what will happen if you take specific actions. So you can think of it like a video model where you ask it to do something. And then it actually gets not only like the question of what to do, gets the actions to do so. And it gives you back not just a video, but how the world feels, like the forces, everything for the robot. So it's essentially like the robots in the Matrix. We take the robot, we put it in a world model.
58:34 Speaker 02 Yeah.
58:34 Speaker 01 And it doesn't know that it's in a world model. It thinks it's in a real world. And it does its things and we ask it to do the things we're usually doing around the house. And we see what it does. And we can put in lots of automated checks to ensure that like both that is performing better, but also that is not doing anything that can be deemed unsafe. So it's really like this incredibly important and powerful evaluation tool.
58:56 Speaker 02 Yeah.
58:56 Speaker 01 And that starts to solving the problem, right?
58:59 Speaker 02 Do you think that's why you guys and Figure and Tesla are getting monster evaluations? Cause is, is the valuation just purely, hey, we're going to sell 10,000, 20,000, then 200,000? Or is it, no, the world model is such a unique asset and so valuable in thousands of different ways. And that, that becomes a, you know, a very much a self-feeding barrier to entry. And, and that could also explain, like, do you, do you plan to productize that core capability?
59:29 Speaker 01 Yeah, so to me, it's this back to like, our mission is to create an abundance of artificial labor. And that that goes across both the digital and the physical. So yes, it will be productized, still a bit out, but like, yes, this will be productized.
59:43 Speaker 02 Yeah, because it seems like no matter how much factory capacity you build, it would be, it wouldn't be till like 2028, 2029, that you could diversify into all these like microsurgery and warehouses and, you know, drones, all that. But that same world model could apply to those much sooner, but you'd have to somehow get it into the hands of.
1:00:02 Speaker 01 Well, actually, I think revenue from the robots will dominate forever. Um, I I do think like the real physical world has way higher value than people think.
1:00:11 Peter Diamandis? I mean, just for folks to realize, right, we're at 110 trillion dollar global GDP. And labor is half of that, right? So the TAM, total but addressable market here is like 50 plus trillion dollars. Just, just if you keep doing what we already do.
1:00:25 Speaker 01 You will be cheap doing what we're already doing.
1:00:26 Peter Diamandis? Yeah.
1:00:27 Speaker 01 It's gonna be so much bigger.
1:00:29 Peter Diamandis? You've attracted some incredible early investors. Do you mind just sharing who who's come into your cap stack?
1:00:36 Speaker 01 I think maybe like we we have some big classical ventures like SoftBank, Target Global.
1:00:43 Peter Diamandis? Nvidia.
1:00:43 Speaker 01 EQT, Nvidia, OpenAI. Uh, so there's some good names in there, a lot more.
1:00:49 Peter Diamandis? That's damn good.
1:00:51 Speaker 01 I think it's becoming increasingly clear, right? That the bottleneck in society to superintelligence is not better algorithms or scraping the internet in a more thorough way.
1:01:05 Peter Diamandis? It's better data.
1:01:06 Speaker 01 It's, yeah, it's better data, and then you need the robots to generate this data. But even more importantly, it's the physical parts, right? You need more data centers, you need more power. You need like, and and like to do this, you need more labor. And like, and it's kind of like this, it's this bootstrapping problem. And if you just break down the pyramid and you say like, superintelligence consists of this incredible amount of data. And it consists of this substrate of compute and power. Then you see that humanoids is a solution to both of them. And if you just do the math, you'll see that. You're probably not going to get there without that. You're just running out of these like basic constraints.
1:01:48 Speaker 01 And I think humanoids will be surprisingly useful, surprisingly fast.
1:01:53 Peter Diamandis? Okay.
1:01:53 Speaker 01 Not perfect, but it's going to be surprisingly useful, surprisingly early.
1:01:58 Peter Diamandis? I have a question on behalf of Saleem Ismail, who's typically our third
1:02:03 Speaker 01 No.
1:02:03 Peter Diamandis? moonshot mate here, I have to ask, because I have to ask on his behalf.
1:02:06 Speaker 02 Exactly what it's going to be.
1:02:07 Peter Diamandis? So, so Saleem is constantly saying, why two arms, why two legs, why not six arms? Why, you know, why we need to have a humanoid form? I mean, in the kitchen, wouldn't it be better to have an extra pair of arms? So what's the definitive answer to him?
1:02:23 Speaker 01 Well, I think first of all, it's kind of right, like humanoid isn't the only thing that will work. Um, I do think that, and we've looked a lot, like, I don't know of any form factor that is as general as a human in doing kind of like any kind of labor in any kind of environment. Uh, and we've tried to simplify, we've tried to increase complexity. Like, this is a pretty good machine. Um, so if your goal is to just be as general as possible, then you need a humanoid. Now, if your goal is to transfer knowledge from humans, it's a lot easier if you have a humanoid.
1:02:55 Peter Diamandis? I think that's the most important part of the equation there.
1:02:58 Speaker 01 It's very important.
1:02:59 Peter Diamandis? And you're not going to transfer to a, to a six armed robot, learnings from a human.
1:03:04 Speaker 01 It is at least harder. And then, I mean, the world is made for humans. It's like Jensen says, right? Like it's brownfield deployment, it's very true.
1:03:12 Peter Diamandis? Yeah.
1:03:12 Speaker 01 And then, I think, lastly, it's just, do you want to live with a six legged robot in your kitchen?
1:03:20 Peter Diamandis? Yeah.
1:03:21 Speaker 01 But I, I view humanoids as kind of like the pinnacle of general technology. But there is this kind of repeat pattern through history of this happening with like zero to one novel products. So if you think about the, say, the computer, it started with big mainframe computers solving very specialized tasks. Uh, the equivalent in robotics would be industrial robotics, right? Now comes the PC or even before the PC, like the kind of like VIX or Atari or whatever, like more general computers. And this gets produced at such a scale that it just becomes generally available. And now it's super high quality and it's incredibly reliable.
1:04:06 Speaker 01 It's because it's huge ecosystem and it just becomes the best way to solve any problem. Even though, and here's the argument against humanoids, right? It's overly complicated for the task. Even though it's overly complicated for the task, when you take your beautiful apple here and you write a word document. I mean, that's the most complicated typewriter I can think of. Like humanity, like mastered nano scale chip manufacturing for you to have a typewriter. Like, but it's still actually the cheapest, most reliable typewriter because it's just like made at such a scale. Humanoids, exactly the same. Now, if you see what happens to computers now, because the market has become so big, it starts to actually become segmented again. And now you see you can carve out niches in computing and they're still so large that
1:04:51 Speaker 01 it has scale. So now you get specialized compute for AI, specialized compute for physics, specialized for all kinds of things, right?
1:04:57 Peter Diamandis? Yeah.
1:04:58 Speaker 01 And this is because it's become so big. Now, the same will happen in robotics. So we will get to where we have Star Wars. There will be different drones doing different tasks and they will look kind of like more specialized to like my repair drone with like six arms and like scissor hands and I don't know. Like it'll get there, but you have to go through this humanoid phase first. So let's just say humanoid is a face.
1:05:18 Peter Diamandis? Yeah. My favorite robot is still Data from Star Trek.
1:05:22 Speaker 01 It's a great robot.
1:05:23 Peter Diamandis? Yeah. And it's kind of the closest thing I think of to what you're, what you're building. You know, lovable, happy robot that you can give a hug to. Do you have a favorite robot?
1:05:36 Speaker 02 Uh, well, I wouldn't have thought of Data, but now that you said Data, it's a, that's top of the food chain. Uh, everybody loves R two D two because for some reason R two D two has no voice, even though they have voice technology everywhere.
1:05:48 Peter Diamandis? It squeaks.
1:05:50 Speaker 02 All those visions that are built around what Hollywood can easily get on a set.
1:05:53 Peter Diamandis? Yeah.
1:05:55 Speaker 02 I think I think the humanoid form factor, though, has there's another aspect that you you kind of touched on, but when I bring it into my house, I have a vision of what it can do and what it can't do based on humans.
1:06:07 Peter Diamandis? Right.
1:06:07 Speaker 02 And so I ask it to do things that are rational and not irrational, because I know what a person could do. If I had a six legged thing that Saleem came up with.
1:06:14 Peter Diamandis? Yeah.
1:06:15 Speaker 02 I'm not quite sure, like, should it be able to climb on the roof and fix the shingles or not? I don't know, like, what this thing's capable of.
1:06:21 Peter Diamandis? Interesting.
1:06:22 Speaker 02 So it breaks the whole, kind of like, comfort zone of users.
1:06:25 Peter Diamandis? Expectations, yeah.
1:06:26 Speaker 02 The thing that does surprise me, though, about the, uh, the robots are unbelievably coordinated between themselves. And there's some good demos of this at MIT, they're just mind blowing. But when you have two movers trying to take a couch up the stairs, and they're like, it's like the Stooges, right? They're like, oh, move a little, move a little. When you see the equivalent act with two robots, they're just in phase, and they just do it seamlessly. So I think, I think there's a very high probability that the standard in the home is going to be like four or six, which gets the price point down a lot. But they work so well in concert with each other, it's almost, it's almost a crime not to have that, that teamwork synergy.
1:07:03 Peter Diamandis? Huh. It seems like a bit much to me, but in terms of, like, maybe I could, like, if I need to have movers, like, you know, I can ask my Neo Gamma and he'll invite some friends over.
1:07:14 Speaker 01 But you, you're not taking everything into account, Peter. Because you have to remember that by the time you have this many robots in your home, everyone's homes are really freaking big.
1:07:22 Speaker 02 Yeah.
1:07:23 Speaker 01 We have an abundance of labor.
1:07:24 Peter Diamandis? Hmm.
1:07:25 Speaker 01 I have a mental map on this. Like, you, I mean, like, your, your, your house is not going to be this small.
1:07:31 Peter Diamandis? So labor.
1:07:32 Speaker 01 It's going to be huge.
1:07:33 Peter Diamandis? So labor is going to, is going to continue to demonetize and, and democratize. Everybody, there's not a week that goes by when I don't get the strangest of compliments. Someone will stop me and say, Peter, you've got such nice skin. Honestly, I never thought, especially at age 64, I'd be hearing anyone say that I have great skin. And honestly, I can't take any credit. I use an amazing product called OneSkin OS01 twice a day, every day. The company was built by four brilliant PhD women who have identified a 10 amino acid peptide that effectively reverses the age of your skin. I love it. And like I say, I use it every day, twice a day. There you have it. That's my secret.
1:08:14 Peter Diamandis? You go to onescan.co and write Peter at checkout for a discount on the same product I use. Okay, now back to the episode. Let's go someplace that I'd love your insight on, which is China. So when I think about the robot industry, you know, I'm tracking 50 plus well-funded humanoid robot companies in different stages around the world.
1:08:38 Peter Diamandis? Majority are US and China. There's some in Europe. You started in Norway. There are some in India, parts in Japan and Korea. But China by far, I think, is dominating. And what I see there with the Robot Olympics and Special Robot Villages is pretty extraordinary, where the Chinese government is really accelerating this for obvious reasons. You know, they need access to low-cost labor to continue the manufacturing boom. They need it for supporting their elderly population.
1:09:17 Peter Diamandis? How do you think about China? What do you think of the work coming out of China?
1:09:21 Speaker 01 Well, first of all, I think we need the same thing here. We don't kind of like realize it maybe as much, but of course, we need the same thing.
1:09:30 Speaker 01 I think the Chinese ecosystem is incredible. I mean, I don't know anywhere else in the world where you can go and like develop hardware as fast. It's just, right, you need something and you go over and get a machine on the corner here. Something broken, and you just like go over the street and like buy some new components. That is someone like, there's someone doing a reflow over on the street corner over there. And like, this is incredible ecosystem. And I think the bay, I said a bay now, but I know it's also like, I mean, the Silicon Valley, like the hardware bay in Chenzan area is also a bay. But Silicon Valley Bay, this bay, we have a long way to go. If we want to like really get to the same level of rapid iteration of hardware.
1:10:13 Speaker 01 So that's just incredible. I think the manufacturing part is incredible. It's just so much process knowledge. And I think this is highly underrated, like, you know.
1:10:27 Speaker 01 Think about magnets.
1:10:29 Peter Diamandis? Hmm, I do.
1:10:30 Speaker 01 We have, so do I a lot. Um, we have great material scientists that know how magnets work and can design very good magnets.
1:10:42 Speaker 01 But then we lack that guy that knows that, yeah, you do all of that stuff that they told you in the books. But you know, after two hours, you have to stir to the left, not the right. Um, uh, like there's like, there's just so much of that, right?
1:10:55 Peter Diamandis? Yeah.
1:10:55 Speaker 01 And this is just so disseminated in China. There, there's so much process.
1:10:59 Speaker 02 Why? How did that evolve in China and not here? What's the cause?
1:11:04 Peter Diamandis? Top down incentives.
1:11:06 Speaker 02 Just funding?
1:11:07 Peter Diamandis? I, I think it's, it's the government saying you're a robot city. You're a neodymium magnet city. You're in just capital and people and just, you know, communist directed. But then allowing companies to build on top of that. I mean, is that what you see as well or not?
1:11:26 Speaker 01 I'm not sure. I think like
1:11:30 Speaker 01 the, the Chinese startup community is very alive, right? And the capital is quite alive. And I felt it, it runs very similar to kind of like more of like what we like to think of as like the Bay here.
1:11:46 Peter Diamandis? I mean, I used to take a group of investors every year to China and we would go and visit Shenzhen and Shanghai and Hong Kong and Beijing and meet with Baidu and Tencent and Huawei and the leadership of all these. And there was a super vibrant entrepreneurial community, right? The, the mindset was nine nine six. You work nine AM to nine PM, six days a week, and that was a great lifestyle. And you considered the 1.3 billion people in China, your market and the 300 million in America, your market as well. But there was a fall off um after 2019 and there was a real dip in that, in that ecosystem.
1:12:26 Peter Diamandis? I think it's beginning to reemerge, but I think the government is really pushing hard on supporting AI and you know, humanoid robots is a embodiment of AI. It's, it, they're, they're, obviously you know this, they're cleanly meshed. Um, so I do think there's a lot more support that the US government needs to give to US.
1:12:50 Speaker 01 Hundred percent.
1:12:50 Peter Diamandis? Hardware companies.
1:12:51 Speaker 01 But I think like what I wanted to say was just like, I think the most likely, the most genius thing they did was the economic zones, like the free economic zones.
1:12:59 Peter Diamandis? Sure.
1:13:00 Speaker 01 Like, it's not that people here don't want to build stuff, and we want to build stuff.
1:13:03 Peter Diamandis? Yeah.
1:13:04 Speaker 01 It's just, it takes too long and costs too much and it's too convoluted, right?
1:13:08 Peter Diamandis? We do it in spite of the challenges.
1:13:10 Speaker 01 Yeah, I think the US should just like spin up some free economic zones, like here you have like expedited permitting and.
1:13:17 Speaker 02 Well, in California in particular, it'd be a no brainer to do that. That's the simplest, best idea ever, but I don't know what it would take to get it through.
1:13:24 Speaker 01 No, I mean, but this is some of like what people are working on these days, right? Like if you look at like Massa Stream for Project Crystal Land, for example, it's very similar to to this kind of like a free economic zone in the US.
1:13:36 Speaker 02 There's another problem you need to solve too, though, which is that, you know, the US just did software, software, software for, you know, forever. And we were not only not doing hardware, we just didn't do chips.
1:13:45 Speaker 01 Well, not forever. We're in Silicon Valley.
1:13:48 Speaker 02 Yeah.
1:13:49 Peter Diamandis? Yeah, we're Silicon.
1:13:52 Speaker 01 So there was a phase in between here where people kind of like they lost the way.
1:13:56 Speaker 02 Yeah.
1:13:57 Speaker 01 They lost the plot.
1:13:57 Speaker 02 Yeah.
1:13:58 Speaker 01 And now we have to find.
1:13:59 Speaker 02 Well, the venture community got all messed up too, because they wouldn't find, if you had a physical component in your business plan, they're like, well, I'm looking for the next Meta or Google.
1:14:07 Peter Diamandis? Yeah, hardware's.
1:14:07 Speaker 02 I don't really care. Like, yeah, hardware's hard.
1:14:08 Speaker 01 Well, all I've been doing is I've been keeping this company for 10 years, so I can, yeah, I can.
1:14:14 Peter Diamandis? You've got.
1:14:14 Speaker 01 I can go on and on.
1:14:15 Speaker 02 You, yeah.
1:14:16 Speaker 01 Like.
1:14:16 Speaker 02 Do go on and on, because it needs to get fixed.
1:14:18 Speaker 01 How much people are afraid of hardware.
1:14:20 Speaker 02 Yeah.
1:14:20 Speaker 01 Right?
1:14:21 Speaker 02 Yeah.
1:14:21 Speaker 01 But.
1:14:22 Speaker 02 It's going to kill us if we don't find a solution.
1:14:24 Speaker 01 I think also it's going to like kill VC, to be honest.
1:14:26 Speaker 02 Yeah.
1:14:26 Speaker 01 I think people, like, like, if you look at return on venture, it used to be incredible, right? If you got to be an LP in a venture, you're like, oh, man, I'm like set, right? I'm going to make the big bucks.
1:14:38 Peter Diamandis? And now it's 10 years and no return. It's dog.
1:14:40 Speaker 01 And now it's more like a philanthropy thing. You want to fund startup entrepreneurs because venture doesn't really make that much money. And I think it has a lot to do with.
1:14:50 Peter Diamandis? Except at Link, you guys are doing, you guys are doing amazing. We are doing amazing.
1:14:53 Speaker 01 That is good.
1:14:54 Speaker 02 Yeah, I know that they.
1:14:55 Speaker 01 But you, you, you guys touched the hard stuff. The point is, if you don't touch the hard stuff, what's your mode?
1:15:00 Peter Diamandis? We touched the early stuff, right? So it's first checks into companies that then are scaling rapidly versus companies that are doing.
1:15:09 Speaker 02 But we're doing first checks into hard stuff at the seed stage, but we're not doing hardware. So I'm as much part of the problem as we were talking, like, what if somebody came to me with a seed stage, hard, physical device problem, and we very rarely will fund that, and it's a dysfunction.
1:15:26 Speaker 01 Well, you should look at,
1:15:30 Speaker 01 what's the biggest companies? They all have hardware.
1:15:32 Peter Diamandis? Yeah, I mean, listen, Elon cracked the code on that. I mean, he's been able to just make hardware sexy, and has generated incredible returns.
1:15:41 Speaker 01 Yeah, I think Jensen says it really well, right? They want to work on the really hard problems that are super painful, that you are uniquely capable of, because you know that your competitors have to go through the same pain or more, and they're not going to be willing to take as much pain as you. This is how you win. And things here are actually defensible, right? The moat we have on hardware, that's years. The moat we have, I'm incredibly proud of our AI team, by the way, we've accomplished some things that are so amazing on such a budget. So we're way ahead of everyone else in what we're doing on world models. So let's say we're three months ahead. Right? Because like,
1:16:21 Peter Diamandis? Way ahead.
1:16:22 Speaker 01 And that is way ahead.
1:16:24 Speaker 02 You're incredibly rare. And all props to Elon, he's incredible, but Elon's pathway to getting to hardware was through PayPal.
1:16:33 Peter Diamandis? Yeah, sure.
1:16:34 Speaker 02 A couple hundred million dollars, burned it all himself, got down to near bankruptcy, was almost dead on both of his big, you know, Tesla and SpaceX, barely pulled it out, and then made them huge, but the VCs were not touching it.
1:16:48 Peter Diamandis? Yeah, he had to borrow money in 2008 in a divorce with SpaceX having its third failure.
1:16:55 Speaker 02 That is not a repeatable funding model for me.
1:16:58 Speaker 01 You know, I was very lucky, I had a very good, like, early founding investor. I said, like, the company didn't start in a garage because we're in Silicon Valley, we started in a barn because we were Norwegian. And at some point, two years later, he sold the farm, so we had to move. He sold the farm to fund the company.
1:17:15 Speaker 02 So you would, this company in Silicon Valley wouldn't exist without a Norwegian investor.
1:17:20 Speaker 01 No, and so we wouldn't exist because we wouldn't, like, we wouldn't have had the runway, right? Operating this in Norway was just incredibly cheap compared to operating.
1:17:28 Speaker 02 So your initial Norwegian investor, did he or she believe?
1:17:33 Speaker 02 That they were gonna make a huge amount of money? Or did they do it because they're passionate about your vision and your mission?
1:17:38 Peter Diamandis? Or are they believing in you?
1:17:39 Speaker 02 Yeah.
1:17:40 Speaker 01 I think it's all three, all three, yeah. It turned out pretty well.
1:17:43 Speaker 02 Yeah, well, yeah, but it wasn't here. That's the point of Meg.
1:17:47 Speaker 01 No, no, no, but like, I mean, uh, there are different phases, right? Uh, if you wanna scale something, you have to come here.
1:17:53 Speaker 02 Yeah.
1:17:54 Speaker 01 I think you can do deep research in other parts of the world. There's talent everywhere, but really kinda like hyperscaling that and like getting it across the finish line.
1:18:02 Speaker 02 Yeah.
1:18:03 Speaker 01 That's it.
1:18:03 Peter Diamandis? Did you consider LA, Austin, Florida versus here in Palo Alto?
1:18:10 Speaker 01 Yeah, we even had manufacturing for a little time in, uh, in Texas, in Dallas.
1:18:17 Speaker 01 There's just, there's, there's a comparison to the talent.
1:18:21 Peter Diamandis? Something in the water, something.
1:18:22 Speaker 01 Like the talent pool, it's a talent pool.
1:18:23 Peter Diamandis? Yeah.
1:18:24 Speaker 01 Like there's talents, talent everywhere, but like the density of talent. And you know, there's different types of talent. Because when you have like a zero to one field like this, in the beginning, you have a lot of like really passionate people. They've been working on this all their life and they're so good at, in this case, like humanoid robotics, right? And I remember back in the day, if you went to the humanoid's conference, like everyone could fit around one table. And those people are still the ones that are most of these companies, right? So, um, and those people, they don't know how to make a great product. They don't know how to, don't know how to scale that to a million or a billion devices.
1:19:01 Speaker 02 Yeah.
1:19:01 Speaker 01 They don't know how to write like the incredibly good APIs for the software to support the ecosystem. They know this thing and they do deep research. And now your field kinda comes of age and it's time to actually do this because the timing is right. And we purposefully stayed very small for the first seven years.
1:19:20 Speaker 02 Yeah.
1:19:20 Speaker 01 Just doing core technology. Now, suddenly you get access to this talent pool of people that just go from field to field. That is the hottest thing right now. And just do it again and again and again and again. And that's Silicon Valley, right?
1:19:37 Peter Diamandis? But there's been an incredible inflection point in humanoid robotics. I remember, you know, we had the Avatar X Prize, right? Our A and A Avatar X Prize that had teams build robotic avatars that you could tell a presence. And I remember the finals, we had good teams. I know some of your team members here were parts of those teams. But it's come, you know, a thousand X since then. And really, in the last five years, right? It's, is it been the, is it, is it been the, the AI models that have made that? What's, what's caused the inflection in the last five years?
1:20:15 Speaker 01 The AI is clearly part of it. There are things we do with AI now that we couldn't do five years ago. I do think we saw kind of like the, the breadcrumbs and we were like on the path already then, but it wasn't kind of working yet. Uh, and I think it's just, you hit this kind of like critical mass of like accumulation of like innovations that has happened in hardware.
1:20:35 Speaker 01 I do think it's important to note though, that
1:20:39 Speaker 01 it's hard to see what is like really innovation or not in any field and especially in humanoid robotics. So I want to just point out again that like you can go on YouTube and find things from the early 2000s that look better than most things you see today that humanoid robotics companies are doing. So you can't just make a beautiful robot that looks good. You have to actually make a robot that is safe, that you can actually manufacture at scale for a very affordable price. And it still is capable, right? And I think that that's been the main unlock and the challenge. Like you, you need to get those things right.
1:21:11 Peter Diamandis? Yeah.
1:21:12 Speaker 01 And that just takes, takes a lot of time.
1:21:14 Speaker 02 I think the, the neural nets are light years ahead of anything anyone would have predicted five years ago. And then the hardware, you know, the, the Nvidia chip that it runs on is getting pushed as fast as any innovation in history because the demand is through the roof. So that part is well understood. On the physical hardware side, what's something that you use today that you couldn't have used 10 years ago?
1:21:36 Peter Diamandis? Yeah. What, what's improving in the motors, in the, in the harnesses, in the electronics, batteries?
1:21:43 Speaker 01 Yeah. So I think mostly it's been on like the motors and material science side. So we make our own motors, including not only the IP for the motor, but also the manufacturing and automation for all this and everything that goes into it.
1:21:55 Speaker 02 The supply chain is so broken. Yeah, you literally make your own motors, like wind the wires.
1:22:00 Speaker 01 Yeah, so
1:22:00 Speaker 02 Holy crap.
1:22:01 Speaker 01 We do it kind of special. It's the one X version of this. So motors is one of the things we really innovated in. And this is actually how I started. Like, you know, when I sat down a decade ago, the first thing I did was design a different kind of motor.
1:22:13 Speaker 02 Okay.
1:22:13 Speaker 01 Um, and the motors we have now in Neo, they are five and a half times the world record in torque to weight.
1:22:18 Speaker 02 Wow.
1:22:19 Speaker 01 Uh, and that's why we have something that's so powerful.
1:22:24 Speaker 02 Hmm.
1:22:24 Speaker 01 That we don't need gears. We can just pull on these tendons to like loosely simulate, uh, loosely simulate, um, human muscles.
1:22:32 Speaker 02 That's why the wheel is so light.
1:22:33 Speaker 01 And that's why it's so light. It's also why it's so like drivable compliant.
1:22:37 Speaker 02 Yeah.
1:22:37 Speaker 01 It's why it's a cheap to manufacture. Um, like everything kind of comes from this now. Of course, when you have these motors.
1:22:44 Speaker 02 Yeah.
1:22:45 Speaker 01 Then you can start using tendons.
1:22:48 Speaker 01 But then you need to think a lot of time into figuring out how to use tendons. And then comes all the material science to have tendons that can last millions and millions and millions of cycles.
1:22:57 Speaker 02 Yeah.
1:22:57 Speaker 01 And these are really hard research problems, right? They're not even engineering problems. They're hard research problems.
1:23:03 Speaker 02 Hmm.
1:23:04 Speaker 01 And we spent so much time figuring all that out. Uh, you can't make the motors that we make without doing some pretty significant innovations in electronics and how you do, um, power amplification and in general just motor drives. Um, so that kind of like, there, there's a lot of things that come together. You couldn't have designed the motors we do today without some of the innovations that happened in magnetics. Uh, and of course, you couldn't have done it without AI either. The first thing I did back in the day when I sat down was to
1:23:37 Speaker 01 program a network to learn how to make motors.
1:23:40 Speaker 02 Oh, you're kidding. You made them, you designed the motors via AI. How long ago was that?
1:23:45 Speaker 01 Uh, it's a bit more than 10 years.
1:23:48 Speaker 02 Wow, okay.
1:23:49 Speaker 01 But I mean, it wa- it wa- it wasn't transformers, but it doesn't matter.
1:23:52 Speaker 02 Yeah, well, yeah, for that kind of use case, but it was neural net driven. Yeah, wow.
1:23:57 Peter Diamandis? Uh, Bernd, you think about robots in the world probably more than anybody else. What's your vision 10 years from now? What are we seeing? What does abundance in labor enable that?
1:24:13 Speaker 02 Goes beyond people's initial reaction to how I would use a robot.
1:24:17 Speaker 01 I think like first of all, what will happen is actual abundance means everyone can have whatever they want. But not only can you have whatever you want, you can have whatever you want in a sustainable manner. Because sustainability is something we lose when you cut corners to shape costs, right?
1:24:34 Speaker 02 Sure.
1:24:35 Speaker 01 If you actually have an abundance of energy and labor, why would you not do things sustainably?
1:24:41 Speaker 01 And then I think the next frontier that comes after just in general, like building out the infrastructure across the globe that allows everyone to have an incredible quality of life is how do we solve the remaining really hard problems in science? And I think this is not going to happen without humanoids, because you need to build particle accelerators. You need to build enormous biotech labs where they're doing experiments. Like you need to do all the experiments, right?
1:25:09 Speaker 02 Yeah.
1:25:10 Speaker 01 And also, I think like it's kind of like almost existential to us for human happiness.
1:25:15 Speaker 02 Yep.
1:25:15 Speaker 01 I don't want like the godlike AI in the sky to be directing all of the planet's inhabitants around with their glasses to do experiments for it to solve science.
1:25:24 Speaker 02 Yeah.
1:25:24 Speaker 01 That's not the future we're aiming for.
1:25:26 Speaker 02 Yeah.
1:25:26 Speaker 01 Uh, we want to have this beautiful symbiosis of like co-invention between man and machine. And.
1:25:33 Speaker 02 Yeah, that particular use case is so acute where, you know, Dennis Sados is working on the full cell simulator to try and close the loop, but you know that you're going to need people to mix a huge number of chemicals to truly unlock longevity and health and chemistry. And, you know, the humanoid robots can do the work because everything in the lab.
1:25:53 Speaker 01 Not only can they do the work, I think this is a common misconception. Humanoid robots will do a lot of the work initially, but once it gets to a certain scale, the humanoid robot will make the automation system that will do the work.
1:26:05 Speaker 02 Hmm.
1:26:06 Speaker 01 Because humanoid robots will not be machining new parts with a terminal, right? You will use the CNC machine.
1:26:12 Speaker 02 Yeah.
1:26:13 Speaker 01 Uh, humanoid robots will not be moving car chassis around by like carrying it with thirty humanoids. Clearly, this does not make sense, right? We have existing automation system and we will build more. What humanoids will do for you is to build all of these automation systems and get them up and running and uncover the remaining gaps that you currently today can't do with humans.
1:26:30 Speaker 02 Yep. Yep.
1:26:32 Peter Diamandis? How are you gonna do in a vacuum? I want my NeoGamma to help me set up my space station.
1:26:39 Speaker 01 Well, we can go on and on and on. I think, first of all, we have a huge advantage because the robot is so light.
1:26:46 Peter Diamandis? Yes.
1:26:46 Speaker 01 And kind of like, I guess Elon's working on this, but payload to orbit is still expensive. Secondly, most of the stuff we have actually works pretty well in space. We have to do some stuff with epoxy on the motors. That's not going to be very vacuum hard.
1:27:05 Peter Diamandis? If you want to train in zero G, one of my companies is a company called Zero Gravity Corporation. These parabolic flights.
1:27:11 Speaker 01 Yep.
1:27:12 Peter Diamandis? And we flew Stephen Hawking in zero G, maybe NeoGamma should come next.
1:27:16 Speaker 01 That would be great. And I actually do think it's like, there's a real use cases for this. And one thing is like building a base on Mars or whatever, right? But even before we get there, just in-orbit assembly.
1:27:26 Peter Diamandis? Yes.
1:27:27 Speaker 01 Is this extremely high-value task? And I think there, actually, we will use teleop. And the reason I'm saying that is just like, the cost of mistakes is so hard that you want to use like the smartest, most expert humans you have. And until we get to superintelligence, that will be a human. And you have people in orbit, you have robots outside, very low latency. And you can tell operate in a very natural manner, as if it was your own body, how to do all of these in-orbit assembly tasks. And they can be incredibly complex and you can still do them with very high accuracy. And you're not endangering people. And of course, when you've done this for a while, you have the data to automate all this, which is very interesting.
1:28:06 Speaker 02 Yeah. Most of where your weight advantage would be really amazing too, because you can take five, six, seven of these.
1:28:11 Speaker 01 And the energy efficiency.
1:28:12 Peter Diamandis? You must be.
1:28:13 Speaker 01 You're gonna have to somehow bleed off your heat, right?
1:28:16 Speaker 02 Right.
1:28:16 Speaker 01 It's really hard.
1:28:17 Speaker 02 Yeah, that's right.
1:28:19 Peter Diamandis? You must be looking to hire people.
1:28:22 Speaker 01 We are.
1:28:23 Peter Diamandis? What kind of, what kind of people watching are you interested in potentially hiring?
1:28:30 Speaker 01 People are just really mission driven, that really believe in the beauty that will be a world where we have an abundance of labor. And like to solve really hard problems. People are really also can demonstrate that they've solved incredibly hard problems. Because that's what we're doing here, right? Everything from material science all the way in the bottom, all the way up to the foundation models at the top. Uh, and I think what we offer is just this incredible place to work. Not with respect to work-life balance and any of this, we're not quite Chinese, but it's a hard problem and we're in it to win, but. Probably the place on the planet with the most experts across all different disciplines in science.
1:29:16 Speaker 01 So if you come here as a mechanical engineer, you will learn so much about AI, about electrical engineering, about batteries, about material science, everything else. And like, doesn't matter which discipline you come from, right? You will learn so much from the people around you. And I think also that's one of our biggest strengths, how we really always work in this multidisciplinary groups. And we find the good solutions between the disciplines, where it's like, hey, you don't need to do that, that's kind of costly in manufacturing, I can calibrate that away.
1:29:43 Speaker 02 Yeah.
1:29:43 Speaker 01 Or like, you don't need to calibrate, I can, this doesn't cost me more.
1:29:46 Speaker 02 Yeah, I can see that actually when you're walking around the building here, you know, Dean Kamen's lab. Uh, in New Hampshire is very, very similar where you use the Segway inventor and everybody's just happy. Like all the MIT people that we know that work with them, they're just happy. And the reason is because when you do software, you're largely behind a workstation all day, you're sitting whatever. When you're doing physical things, you're moving around a lot more and you're, you're building and making and it's just, it energizes you all day long. It's just such a fun work environment around. It's so obviously tangible, just walking around, talking to people. No, it's a good, it's a good lifestyle.
1:30:22 Speaker 01 And it helps when there's a lot of robots walking around with you.
1:30:25 Peter Diamandis? Yeah, for sure. And people go to One X Technologies as the website to
1:30:29 Speaker 01 That's true.
1:30:30 Peter Diamandis? To go and find out what positions are open.
1:30:32 Speaker 01 Yeah.
1:30:33 Peter Diamandis? Yeah.
1:30:33 Speaker 01 And follow us on X and you'll learn more about us. Pretty, pretty active there and
1:30:38 Peter Diamandis? For sure. And one thing I'm excited about to announce is you and Dar and the Neo Gammas are going to be at the Abundance Summit in March.
1:30:46 Speaker 01 Yeah, can't wait.
1:30:47 Peter Diamandis? Yeah.
1:30:47 Speaker 01 Need a lot of great people.
1:30:48 Peter Diamandis? Yeah, so our theme this year is the rise of, is digital superintelligence and the rise of humanoid robots, because the two are going together.
1:30:58 Speaker 01 Spot on.
1:30:59 Peter Diamandis? Yeah, I think so. I mean, it really is, it really is. And without making any promises, I'm hopeful we'll have a number of the Neo Gammas there, sort of like interacting and and sort of living and hanging out with the Abundance members, yeah.
1:31:15 Speaker 02 How do they get there? You fly in an airplane seat? They just walk on?
1:31:18 Peter Diamandis? Yeah.
1:31:19 Speaker 02 You don't box them up, do you?
1:31:21 Peter Diamandis? Yeah, we're down in LA.
1:31:22 Speaker 01 Yeah, we're probably gonna drive down to LA. It's easier than getting them on a plane.
1:31:27 Speaker 02 Do you put them in the seats and strap them in?
1:31:28 Speaker 01 Yeah, we do. Actually, at this point, they're starting to sit into the seat themselves. So it doesn't strap itself in yet, but that's coming. It's an interesting story, interesting, it's funny, funny story here though, because we put one of the first robots on a plane back in the day. We were rushing back home from China. It was a proper, proper startup story where we were like, this is way back in the day. We were running out of money and we hadn't kind of like gotten to where the product was good enough to raise more money. So I took the entire team and we went to China and we lived in a hotel for five weeks, designing and manufacturing, kind of like as we go, no, design until late in the night and in the morning you walk down to the machine shop, you help get them some
1:32:10 Speaker 01 information, you get some new parts back and we just kept like iterating on this electronics market. It was magical, right?
1:32:16 Peter Diamandis? Yeah.
1:32:16 Speaker 01 And then we have to go back and we're just like, okay, we have to rush back on the plane to meet some investors. So we check, we take the robot and we fold it up, right?
1:32:27 Peter Diamandis? Yeah.
1:32:28 Speaker 01 And we put it in a briefcase. And then when it goes through the scat
1:32:32 Peter Diamandis? Oh, no.
1:32:34 Speaker 01 And you can see the guy just goes all white and he's like shaking in his hands when he's opening the bag. And we're like, no, no, it's just a robot. And he's like, yeah, it's a robot.
1:32:44 Peter Diamandis? That's hilarious.
1:32:45 Speaker 02 That's awesome.
1:32:46 Peter Diamandis? Uh, well, yeah, I really, really thrilled. I loved your, your TED Talk and excited to have but NeoGamma there hanging out with all our Abundance members and hopefully you'll be ready to make some, uh, some sell some robots. So in the early days of, of making, getting them to home, uh, no promises, but you're going to have sort of an application process to get, to get the robots in and start to build data, data assets. When, when do you think you'll be ready to take pre-orders and orders for NeoGamma?
1:33:24 Speaker 01 I'm gonna be kind to my team and not say a specific date.
1:33:27 Peter Diamandis? Okay.
1:33:27 Speaker 01 But it is happening this year.
1:33:28 Peter Diamandis? Okay.
1:33:29 Speaker 01 It's this year.
1:33:29 Peter Diamandis? This year, 2025.
1:33:31 Speaker 01 2025.
1:33:31 Peter Diamandis? Yeah.
1:33:32 Speaker 01 Now, we're gonna talk a lot about this in the preorder, but the most important thing we do here is expect, is expectation manage.
1:33:39 Peter Diamandis? Yes.
1:33:40 Speaker 01 This is incredibly early, right?
1:33:41 Peter Diamandis? Yeah.
1:33:42 Speaker 01 And what you're buying here is kind of a ticket to be part of this transformation. Adopt an EO into your family, help us teach it. It's gonna be a lot of fun. It's gonna be useful.
1:33:52 Peter Diamandis? I love that framing. That's perfect.
1:33:53 Speaker 01 Uh, it's going to be useful, but it's not gonna be perfect. It's gonna be a lot of like rough edges and we're gonna treat you really well. We're gonna figure it out together and it's gonna be an incredibly fun journey. And that's kind of like the early adopter program that we're launching this year.
1:34:07 Speaker 02 Yeah, you're gonna have a long waiting list. You know, we need, we need millions and millions of these and we need to get the price point down and then, uh, I mean, when, when you think about the constraints to human happiness globally, uh, a lot of them are gonna be solved through regular AI, but another big chunk, most of them are related to houses and food and physical happiness.
1:34:30 Peter Diamandis? Give, give the jobs that are dull, dangerous and dirty to the robots.
1:34:34 Speaker 02 And then create a lot more of the things that make people happy. The parks and the, and the homes and, you know, all of that, bigger homes and better things to play with. It's all constrained by that inability to manufacture through the lack of the humanoid, you know?
1:34:51 Peter Diamandis? Let me, let me ask you a, a numbers question. So I interviewed Elon at FII Summit. You're gonna be there in October as well. And, uh, and also Brett Adcock and they both gave a number around 10 billion humanoid robots by 24. Do you believe that number?
1:35:08 Speaker 02 Ten billion by twenty twenty?
1:35:09 Peter Diamandis? Yeah.
1:35:10 Speaker 01 I think it's probably roughly correct. I think it might happen before. I think it really comes down to
1:35:19 Speaker 01 what kind of like artificial constraints we put on how we scale.
1:35:22 Peter Diamandis? Yeah.
1:35:23 Speaker 01 Um, at that point, you have to actually really think about like, how are you refining rare earths? How are you mining more aluminium? How are you ensuring that you like, get your labour bootstrap really well with with robots into labour? How do you build out a power infrastructure? We need more chip fabs, by the way.
1:35:39 Peter Diamandis? Yeah, I mean, I mean.
1:35:40 Speaker 01 Like, we're, we're, we're not going to be able to build ten billion humanoids without way more chip fabs.
1:35:46 Peter Diamandis? Yeah.
1:35:47 Speaker 01 We can, we, we can help with having robots build this out. But I do think that timeline depends a lot on how permitting processes go and like how much
1:36:01 Speaker 01 we kind of allow ourselves to scale and how fast. But I do hope we get there.
1:36:07 Peter Diamandis? Yeah, I mean, for reference, there's like a, a billion automobiles on the planet. You know, you would think there's more, but I don't know how many, sure how many iPhones or whether it's on the order of eight billion smartphones on the planet.
1:36:19 Speaker 02 I'm really glad you said what you just said, though, because the numbers are so wildly out of balance. You know, each, each one of these robots uses a full GPU, could probably use two. And if you're talking about a billion of them by twenty forty, we're only making twenty million GPUs a year. And then TSMC has sixty-six percent market share now in the fab. So they have literally one point of failure for the entire economy that we're trying to build. And so we're desperately short on the fabs.
1:36:47 Speaker 01 And, and that's if you just go one layer deep.
1:36:50 Speaker 02 Yeah.
1:36:50 Speaker 01 Like, look at ASML.
1:36:51 Peter Diamandis? Yeah, oh my god, sure.
1:36:53 Speaker 01 Right? So like, the supply chain for chip fabs, that's even more brittle.
1:36:58 Speaker 02 Yep. I'm really surprised that we're not moving much faster, given that Elon is right in the middle of it. Elon is, or was in Washington. That we're just letting this bottleneck.
1:37:09 Speaker 01 Well, how, how, how, how long have we been talking about magnets?
1:37:13 Speaker 02 How long have we been talking about magnets?
1:37:14 Speaker 01 We've been talking about magnets for a long time, right? That's a problem that like only China can really make high-grade magnets.
1:37:21 Speaker 02 Yeah, yeah, yeah.
1:37:21 Speaker 01 Uh, and it's not just the rare earth. It's the process to produce.
1:37:25 Speaker 02 Yeah, yeah.
1:37:26 Speaker 01 And I think now finally like people are opening their eyes and like, wait a minute, this is actually a real problem.
1:37:32 Speaker 02 We meet, we meet with a lot of government officials and they're completely unaware of these bottlenecks and, and it's funny that if you point them out, there's still no reaction. Yeah, it's like, but it's, it's so acute and so urgent. Um, it's, I mean, you're, you're in a perfect position to actually identify those bottlenecks. So it's really great that you said it on this podcast because then we can take that material and say, look, look, he would know, you know, like this, this is what we need. This is going to be a crisis very quickly. So, yep.
1:38:02 Peter Diamandis? Thank you for the tour today. Thank you for the work that you're doing. Uh, super grateful, uh, excited to have you at the Abundance Summit with your, with your team of robots. Um, if you're interested, it's abundant360.com, check it out. Uh, again, it's onext technologies, uh, dot com to come and learn about the positions here and, uh, following you on X.
1:38:26 Speaker 01 We're even easier, onextech.
1:38:28 Peter Diamandis? Onextech. And by the way, the reason you named the company one X, I think that's worth closing on is the story here.
1:38:35 Speaker 01 Well, you know, there's all of these videos on YouTube with robots.
1:38:38 Peter Diamandis? Yes.
1:38:39 Speaker 01 And there's always like this eight X or four X in the corner. Uh, and all we do is real time because we build proper robots.
1:38:45 Peter Diamandis? There you got it. What you're seeing is real one X speed. Um, and, uh, we had fun today with your gamma.
1:38:52 Speaker 01 And it's also amazing because you, if you apply for our careers and you come here, you get to be a one X engineer.
1:38:58 Peter Diamandis? Okay. Well, a real pleasure, my friend.
1:39:02 Speaker 01 Awesome.
1:39:02 Peter Diamandis? Thank you for the day.
1:39:03 Speaker 02 Thanks, Jeffrey. The things I get to do because of this podcast.
1:39:06 Peter Diamandis? Yeah, we're having fun.
1:39:08 Speaker 02 Oh my God, yeah.
1:39:09 Peter Diamandis? Yeah, awesome.
1:39:10 Speaker 02 They're awesome.
1:39:10 Peter Diamandis? Every week, my team and I study the top 10 technology meta trends that will transform industries over the decade ahead. I cover trends ranging from humanoid robotics, AGI and quantum computing to transport, energy, longevity. There's no fluff, only the most important stuff that matters that impacts our lives, our companies and our careers. If you want me to share these meta trends with you, I write a newsletter twice a week, sending it out as a short two-minute read via email. And if you want to discover the most important meta trends 10 years before anyone else, this report's for you. Readers include founders and CEOs from the world's most disruptive companies and entrepreneurs building the world's most disruptive tech. It's not for you if you don't want to be informed about what's coming, why it matters, and how you can benefit from it.
1:39:55 Peter Diamandis? To subscribe for free, go to dmandus.com/metatrends to gain access to the trends 10 years before anyone else. All right, now back to this episode.
1:40:18 Speaker 03 College football is back. So Hilton called and me, the superstition concierge, to make your fan rituals a reality. Need a room to match your lucky number? We got you. Want to make sure our team doesn't wash your lucky jersey? Oh, that smells lucky. Hilton's unmatched hospitality can keep up with any superstition. Even a marching band wake up call at 5:55 in 55 seconds. Hit it! When you need a team that will do whatever it takes on game day, it matters where you stay. Hilton, for the stay.
Previous The AI-Crypto Collision That Will Redefine Global Power w/ Eric Pulier, Dave Blundin & Salim Ismail | EP #187 Next Open AI's Head of Product on the AI Race, Google & the Reality of AGI w/ Kevin Weil & David Blundin | EP #189 Page updated October 26, 2026 · transcript and summary are machine-generated
