As you think about competition, is it Tesla and China? Is that the way you see it in the world right now? >> I personally haven't seen it much out of Tesla in 2 or 3 years. I've seen China do a lot of dancing. We care about something very different than what we think China cares about. We just got to go beat them. It's a real race here. The work isn't really practical. The technology stack needed to get there isn't practical. And the AI stack to do it can be done by a a senior in college. My view [music] is there's like roughly four stages for this for this like humanoid race. My chapter one of this book is you have to build like really good functional hardware. Two is you have to be able to build an AI recipe. Chapter three and four for me are like scaling intelligence and scaling robots. >> What might we [music] see with uh with figure 4? What areas you focusing on making improvements? >> We're like in the flip phone era for humanoids. Figure 4 is probably like iPhone one moment we're going to have here soon. We need the robot to get cheaper. We're trying to make the more robot more [music] capable. A little sneak peek of that is we've been working on the hands. A figure 4 with a glove and a human with a glove is almost indistinguishable. >> That's like an iPhone moment. Figure 3 is probably the best humanoid on the market by far, but it'll make Figure 3
[00:01:01] look like a caveman. >> Now, that's a moonshot, ladies and gentlemen. >> This episode is brought to you by the Abundant Summit and Link Ventures. >> So, Brett, good to be here for our annual catchup. Uh it's been about a year since Dave and I were here to to see figure and uh quite the difference. So walk in, you just parked in the in the parking lot over there. I walk in and there's like a hundred robots outside to greet you. >> So my question is do you what do you call a group of robots? I'm calling them a flock of figures. Do you have a name for them? >> Yeah, those are our visitor bots. They um they walk around all all of campus. They can like open doors, walk on any stairs, use the elevators. Um, and they're they're there to greet greet people, talk, and then we're going to have like we're right now working on trying to get them to like, you know, do like get coffee for folks. Um, we have a kind of a stretch goal of like it'd be great if you can come to campus and never talk to a human as you get set up. I um, you know, get fully checked in, walk to the conference room, do a tour,
[00:02:00] like go to the bathroom, like all done through a humanoid. >> Let me show you where I was built. Let me show you what I'm doing today. >> Yeah. No, exactly. Like we do that all the time. So like why not? We should have a humanoid. >> And you said something a minute ago that's pretty profound. You have more figure robots than humans now. Yeah, we have more figure robots on campus. >> When did when did that happen? [sighs] >> Probably like four or five months ago. We uh crossed the number of I think we have like seven 800 people here. So we crossed the number of uh Yeah, we have more robots now. We have many more now than humans. >> And you're up to how many produced at this point? [sighs] >> I don't know like uh 1500 or so on campus, something like that. An order of like uh we have well over a thousand on campus and we have robots out in other places in the world as well. >> Oh, nice. So, we're seeing figure three around here. Uh, last time we were here, you were just transitioning to figure three. I think at that point, >> um, how long before voice? How long before I'm talking to the figure? >> You can talk to any robot here. >> Yeah. >> Yeah. You can stop it and just say, >> ask a questions. [laughter] >> You can talk to that. It's got real time
[00:03:01] speech to speech. Yeah. Yeah. So, every robot here is equipped with speaker mic. Uh, we um we actually use uh HARKC for our voice AI. Um and um and every robot here has got like full realtime speech to speech. >> Incredible. >> Yeah. >> Um so you know a lot a lot of people are asking me so when am I going to be able to get my let's let's jump into some timelines and applications kick things off. I think that's the most interesting. Uh when can I order a figure for the home? I think most people want to know that question. >> Yeah. I mean at a high level we have two tracks. We want to ship robots commercially. We want to ship robots in the home. >> Yeah. Um they're really different spaces. The commercial market's um in some way easier where it's like you know very well well um specified tasks. Um you can almost like write on a piece of paper exactly what the robot's doing every day. >> It doesn't really deviate from it but it's really hard in sense of like we have a lot of speed and performance goals to go hit and we mess up you get fired. Uh no great not great. So um
[00:04:00] really tough. Um on the home is a little bit of the opposite. we don't need to do at the speeds and uh speeds we would need on the commercial market, but then there's not really a specified task. Um we, you know, every home is different. I bet if we ask like what kind of appliances three of us have, they're all they're all different. All of our homes, the setups, the homes different homes. The things we'll want the robot to do will be like will change quite um dynamically. Uh so we're working on both in parallel. We think they're both an intelligence problem. Uh really like how do we get like the robot to I mean the robot figure 3 is really capable to do most things in a home. Um, so we're working really hard on trying to build like larger pre-trained models and doing a lot more work in post- training in RL uh to really really try to crack this problem. Um, when are we going to have in the homes like figure like so then like I think you'll get a little bit different of opinion if you talk to other maybe companies in the space but like our our view here is like figure will ship a robot when it's like when it ships like a really high quality product in the home. Like it needs to be able to walk into the home, talk to you, be able to do real work. like we don't want to
[00:05:01] be teleyoperating the home. We don't want the robot to be having any issues and failures or stopping because the robot's unsure what to do. The robot needs like be like a high quality product when it enters like an apple style quality product. >> The kids and the dog to worry about too. >> Yeah, exactly. But it needs to work like it needs to be able to go in there and be like this magical experience of having like another person in the home. >> Um so that that's our like launch criteria. Uh we're not going to launch before then. I think you'll see others do that and it will just like in my mind won't won't work. It'll be like what you've seen like AI devices. You've seen this now with Humane and Rabbit and these other groups in the in the AI device space. It just it's not going to work until you have like this really magical product. Um, so for us, we're we're trying to crack that. The the like key constraint here is like intelligence. Like the figure three robots like they have the ability to fold laundry, do dishes, like unload groceries, like open the, you know, open the refrigerator, open the oven, like do like do real work in the home. Um, like clean the house up. Uh, we've shown some of that. Uh but we need to be able to do that end to end for like for you know a
[00:06:00] dozen plus hours per day every single day seven days a week for years. Yeah. >> Um so we're pushing on that. I would say the biggest constraint is like an intelligence constraint. Like how do you drop a robot in any home? We would pick just talking to it have it do useful work. Uh there are other things. There's like an integration problem, there's a safety problem, there's other things that are like really hard but by and large it's it's an intelligence problem we're trying to crack. But again without a promise and I won't tell anybody [laughter] um [clears throat] 27 28 I mean we're doing testing now and we we'll at any given time now we'll be in like in in order of anywhere from 20 to 100 new net new homes we haven't been in >> u in California doing testing as of right now. >> Okay. >> Um and we've been doing this since the launch of like uh of Helix 2.5. Um, so we're doing that mostly for for like for evaluations and evals for our models. Uh, and like a ton of hardcore testing. [sighs] Uh, my my my forecast um, I
[00:07:03] think the first thing you really need we need to we need to prove to ourselves is we can do end toend work in a in a zero shot environment. Meaning a home we've never been in. We we haven't trained in it. We haven't seen any objects before just through speech. We can take commands like say a prompt and we can basically do end toend work. >> Um, we're doing that really well today on like the micro level. Like if you ask it like one thing like like make the bed or like uh say unload dishes like um from >> literally just did that actually. Totally open-ended prompt. Yeah. No like no preconceived what I'm going to ask you to do. >> Yeah. >> And it it takes a really good shot at any task right out of pure prompt. >> Yeah. So, um, so, okay, maybe like, uh, sorry, a little little verbose, but like the we we we need to be able to like, um, go off and do like do that like fully end to end now, uh, over long long horizons. Um, [sighs] we're training now larger and larger models. Uh, what what you saw there in Blackbox today was Helix 3. It's like the probably the only
[00:08:00] people in the world >> that have seen this outside this office. Um, and um, for me, I don't know what you guys were feeling, but like it it feels for me like a really like a baby agi moment. It's um, I would characterize it like a like it's like an adult body, like adult, it's a robot, but an adult human, like um, like a really dumb adult human, like a 2-year-old and like adult body. Um, and you can basically ask it whatever you want. And um, and it can do a lot of things and some things it just can't do well at, but it was um, it was significantly undertrained. It was our first model checkpoint coming out of Helix 3. We now have like much larger models that are being trained that we're evaluating even this week. Um so my bet is um you ask like what date. Um >> what year? >> My my bet is within the next I hope to see I obviously can't promise but I hope to see in the next sub six months perhaps even sooner we can put a robot into any home and do fairly large like long horizon tasks end to end, right? >> Um like over like hours and hours. And then the question of when you would actually sell them to consumers.
[00:09:01] >> Yeah. >> Yeah. Then we got to figure out how to, you know, get them to consumers, like make sure they're really safe. Uh there's like there's the reliability is really high. There's like no issues and faults. Um you know, >> do do you think you're going to uh keep robot maintenance and repair as a figure in-house capability or are you going to, you know, sort of approve outside vendors to do that all over the world? For >> for now, we're going to do ourselves. >> Yeah. Like we really want to make sure the robot is like is the highest quality experience like in the home and we know the robots better than anybody else. We built them. So we're going to do that ourselves. >> You can just crawl into a box and ship itself back. >> Yep. Totally. There's there's like >> the label on itself. >> There's like the hand comes out, slaps it on and closes the box. >> There's like many ways to solve the the the this problem. Like the the instate is for sure like us not doing everything. Um, but in the beginning as we like ramp the product up, like you really want this incredible experience like in the home is basically just like another human is in your house. And um, we're only going to get that
[00:10:00] with like the highest level of care and attention and we're putting everything our blood, sweat, and tears like all into this to make sure like it comes out really rocking. Uh, the flip side if it comes out and doesn't work really well, you're kind of dead. Yeah. >> So like uh, so we really want to make sure it's a really high >> Steve Jobs fan as a kid. >> Of course. Yeah. >> All right. It's like the parallels are unbelievable. You know, I was with uh Mike Row who at uh and and Joe Polish at an event recently and Mike Row was the guy who did the Dirty Jobs show. >> Yeah. >> Uh and he was I told him I was going to be coming here and he asked when are we going to see uh you know figure robots doing uh you know welding and uh and roof repairs and sort of all the the blue collar work. You know, there's this been interesting conversation that AI is coming for white collar jobs right now and it and it is from finance and legal and and diagnostics. But the question of when will these robots start coming into
[00:11:00] vendors doing you know traditional you know brick laying and all the manual labor out there. >> Yeah. I mean I like my my like my thesis here. Okay, let me lay out my thesis for robotics I feel like is maybe helpful to kind of give you guys my sense. Um my my view is there's like at a high level like roughly four stages for this for this like humanoid race. >> Um my chapter one of this book is you have to build like really good functional hardware. Like the hard the robot hardware is got to be incredible. Like you can't build like AGI on a robot without really good hardware. And uh we're doing that you know obviously now figure 1, two, and three. We have four coming out at some point. like um we think we're like you know but anyway the hardware needs to be able to do the kind of the speed torqus uh like uh necessary that say a human could with the same range of motions roughly at a high level be affordable and you got to make a lot of them. Yeah. >> Uh so that's that's one. Two is you have to be able to build an AI recipe um that can run end to end longer horizon work. Uh full loco manipulation mean like walking the body and moving the hands.
[00:12:01] Uh and you really haven't besides figure, we really haven't seen any of that in the world. Um you've seen like groups do like tabletop work with like offtheshelf robots and other stuff that are like bolted to a table that are overfitit to a certain environment. You haven't seen robots be able to like go off autonomously, be able to walk around and do stuff. Uh, and we started doing that with Helix 2 like full like full like full like roomwide. >> You know, I have I have the memory of being here when your very first robot the big deal was putting a Kurd cup into the coffee maker. >> That was like >> forever ago. >> Arguably like maybe the one one of the most important turning points we've ever had. Uh, >> getting coffee. Yeah, that's a turning point [laughter] for everybody. >> I mean, honestly, it sounds funny. It sounds ridiculous, but like back then I was like that was like two two and a half years ago. We we like everything was coded. Everything all robots was this is why the second chapter of like getting the AR recipe is so important. Everything was like C++ code. Yeah. >> Um and we were using like a kind of an autonomy like 1.0 stack for SLAM and object detection and everything else. >> Uh there was like really no path like making these really work at scale like it they robots just wouldn't wouldn't
[00:13:01] wouldn't um uh there there was really no like how do we build a billion units and make them all across the build the iroot movie. And that demo where we put we had figure one put a K cup into into a curig. Um I was sitting there with the team and I was like listen if we're able to pull off like uh you know full upper body um like uh dextrous manipulation with a neural network like this. It's a the first time in human like robotics history I believe that anybody's ever done this. But it it paves a path for like all this being done through a neural network like no code. >> Yeah. >> And this is what we really need. It's just like the the like entropy is too high in the space to be able to go off in the world and do everything that we want to talk about like go into every your guys' homes and all uh throughout the world and be able to do stuff like you think about the blue collar stuff you just said too like that stuff is really uh forceful interactions you're having with the environment like grabbing objects you're you're like making impact. It's like it's really quite uh difficult work. Anyways, um I I told the team I was like I I think if you do this it'll be like so historic that we'll probably raise like $500 million. And I put $500 million on a post-it note and threw it on a coffee
[00:14:00] cup. And I said, "If you're able to like actually do this demonstration, uh, you know, show us internally that this like neural networks really work." Uh, we'll probably raise half a billion dollars. And, um, we did it and we I think we raised like a little under 700 million like shortly thereafter. And it was a huge turning point. It it showed that like deep learning works for robotics. And that kind of entered for me like chapter two, which is like you got to have an AI recipe that really works well or you have like no hope to making the story work. People don't realize you jumped into robotics here before neural nets were really a thing. Um just self-driving cars too. They were 90% C++ 10% neural net. Yeah. >> And now it's completely inverted. Now the cars you got to keep the C++ it's the brake you know is connected fly by wire. So you got to keep some some deterministic code in there. But basically all the decision-m is neural net. So for the curig cup was it literally 100% conversion to neural net? not a single line of handwritten code. >> The we were sanding in a controller that was in C++ >> but the whole upper body um like all the
[00:15:01] hands and all the were all all a single neural net. >> Um so we were basically maintaining balance but everything above that was a neural network >> and then you know to be frank it took us about a year to go from a tabletop like that to all neural nets all the way down the stack moving around >> a year. Now, it was Helix. Helix one was like upper body neural nets and Helix 2 was like full body Yeah. >> neural nets. And that took us a year. I was in a stand up like right behind us in the lab here uh for for a year trying to figure out how to move the robot off a table to do work. And we always thought it was um really trivial like like I I should be able to like joystick like you know you can in your head like I can just joystick a robot somewhere and then do work again because you're like you're usually pretty static a lot of times when you're doing work. You're not like walking around, you know, making coffee. you're just like standing there. >> So, um but we realized uh the the hard truth was that this was a full body experience like where you put your feet >> like you're almost like your whole body is there just to get your hands and your head to a certain place in the world >> and stay balanced.
[00:16:00] >> Yeah. >> And it was uh what what really mattered was like the body actually mattered more than the hands and the head like like um and what we really learned is that all of those joint trajectories needed to be done uh as a single almost like unified experience. Meaning you need to understand what the knee and hip and feet are doing as it relates to fingers and hands. >> The same exact time stamp. >> Yeah. >> So it needs to be like all trained together. >> I think a lot of people don't appreciate like if you watch YouTube all day, you think that the Chinese have these insanely dextrous robots that are but under the covers >> there's like >> it's teleyoperated or you build these macros like 10 movements that are pre-programmed and then when you stitch together the video you think wow China's miles ahead of us in robotics. The tech stack underneath it is totally contrived. It's not a real neural net that's that's learning on the fly. >> The let me just be clear about what we see out of China. We see robots dancing and these are these are like deep learning RL controllers. Yeah. >> Like uh like one of the best forms of it is like training through like deep mimic
[00:17:01] >> and uh these are humans wearing a mocap suit like a really like a like a Hollywood film suit. They're running around doing ninja moves and then you're basically training an RL controller to do this and replay it back on the robot. They're all done open loop, meaning that you can just like push a button. They don't react to the environment at all. >> A lot of them have little lightars in their head, so they localize when they're in parades together, so they don't run into each other. >> Uh all of this work is completely orthogonal and different than what you need to solve general robotics. Um so um while it's impressive to see like um maybe for the hardware demonstrations that you're doing um none of it's like practical uh for like what we need robots to do at scale. So you're seeing like okay the work isn't really practical. The technology stack needed to get there isn't practical and the techn in the AI stack to do it can be done by a a senior in college with one robot. This is like very easy stuff. I mean we've we've done this work uh before in our past so we like we understand it. It's one of the coolest things about the investment community in Silicon Valley is that they get the
[00:18:01] difference because it's very subtle to the outside viewer the difference between the two things. But if you watched, you know, chat GPT become Astra while expert systems just kind of did nothing. >> Yeah. >> And you look at early on, you know, it's giving I can't spell strawberry. I can't add four and four. But you knew it was like on this trajectory where it was going to punch through. And the the way my friend John Le, our mutual friend John Lenny, describes, it's like building a submarine underwater and you don't see anything and then suddenly a periscope pops up. You're like, "Wait, that courage cup is the periscope. That means there's an entire submarine that's about to emerge." >> This this curcope is like we think like um starting chapter 2, which is like a recipe work. We ended chapter 2 with like doing kind of like this two I think 200 hour live stream with logistics work. you're seeing it right now like that that guy that those robots on that line haven't stopped since that since we did that four months ago. They run like uh practically every day. And then >> somewhere in the AI is they go themselves like could you please let me
[00:19:00] do something else. [laughter] >> But but you're hitting an important part like chapter three and four for me are like scaling intelligence and scaling robots. >> And what we really want to see is like what you're mentioning here with the submarine. We want to see as you're putting more input in you're getting better output. And for for us our input is data and compute and our output is getting robots to do more useful work. >> Well, you're blowing through. So stage three is data data data. >> It's data comput. It's data and comput. It's a it's scale. It's AI scaling laws. >> This episode is sponsored by Google for startups. [music] Think about this for a second. You now have access to the same generative AI models that cost hundreds of millions of dollars to train. Google's startup tactical guide for generative media gives you complete blueprint for deploying Google DeepMinds models in production. Images, video, audio, all of it. Real architecture, real results. Find the link in the show notes below. when you stood up Helix, you needed to the models that existed back then were not interfacing with physical systems.
[00:20:00] And then recently you saw, you know, GPT Astra uh actually be able to operate robots. So talk to us about about that. How you know uh is helic a general model that can operate a physical system? I would say like one thing I'll say like for these like like other like kind of type of language models and even multimodal models we're seeing command robots they're they're all being done for very basic tasks on a fixed camera frame uh very slowly. >> Uh if like if you slow robotics down enough none of it becomes a really difficult autonomy robotics problem [laughter] and like in the limit. So um these videos are shot with like 5 10x 20x speeds um on a very slow >> people realize they're sped up. They're sped up because like they slowed everything down because like uh latency and like these become like almost all like uh these like these become like you're basically building like closed loop control but doing it really slow. >> You need that feedback but the models
[00:21:00] are not good enough to output this really quickly and they're very like lowdimensional spaces on very basic tasks. >> We just have a different problem to solve. We need to move at very different dynamics and frequencies than uh what we're seeing out in the world in these videos. We need a robot to walk around. It needs to be able to react to the environment really quickly. It needs to be able to move a package lay over there at like sub 3 seconds a package. It can't really mess up very much. Only a few percentage points basically over like a whole shift. Uh we need robots to go into home, do a very delicate like fine grain manipulation, grabbing cups and dishwear and put them away and open doors and cabinets and navigating through tight spaces like a bedroom or like kitchen. Um and then we have to do that on like maybe one of the most complex mechanical machinery like ever built. uh and not mess up and you have to balance on like basically a quarter size point in your foot above the center of mass in your torso. It becomes like a very [snorts] different problem when you like attack it like this than what we'll see on like tabletop college setups. And you're seeing that now with like Astra and the rest of the groups. They're not real models for robotics. If they were, we
[00:22:01] would we could run them. >> Sure. Um, so I think um, >> well there's another subtlety in there which is your entire AI stack is totally owned by either Figure or HARKC which you know when you do a you know university setup tabletop whatever it's on top of at least one layer of third party neural net LLM. >> It's just not real. You can't put them into really high dynamic robots like this and make them work. >> Yeah. >> And you can't scale them. There's no path that where you can do like the cur K cup go from there to full locom manipulation to go to like their like to what we did with like Helix 2.5 recently. There is no like the architecture needs to dramatic dramatically change. These are setups that have like basically unlimited compute. They're plugged in with like really low latency coms through a network. Uh they have fixed cameras. They're overfitting the tables generally for data collection. Like they're not any scenarios that are realistic what we'll see. Um so we had to go build Helix from the ground up to build how do we build the the right brain for the robot. And um so right now I would I would say arguably we understand like how the brain works really well now. And
[00:23:00] what the brain is starved of today is data and compute. >> We're just like we don't have um we don't have enough like like world understanding of what forces feel like and joints and observations to be able to get it into like every home in the world. But we do see a path now. I mean, with Helix 2.5, we went into 30 uh rental homes that we rented like the day that day or day before and we were able to do work like like make the bed and like tidy the living room without saw those videos >> without ever seeing the house or objects before. It was like fully zero shot. They call it zero shot in like in like the in the community where like none of this was in training. It was the first time these robots seeing this >> and that that was just like um like think of this like a bunch of stair steps that we've been like taking now over the last three years that are all building towards how do we get that now to go from there to drop into any home for like for 10 hours and it can do work and that's the trajectory we're on now. So that really gets me the heart of your question is like when we're going to see that. There's um you know we've arguably
[00:24:03] seen some of that this year in 2026. I think you'll see more from us here even this year. And then it like I'd be very shocked if we exit 2027. We haven't seen this this whole picture. >> How much of the compute is running on the figure robot versus on on Wi-Fi? >> Also training versus inference. How much is training? How much is inference? >> Yeah, we um most of our compute today is in training. Um like uh like the dollars going towards compute. Um we spend a we just doubled our we're doubling our compute right now internally next week. And then um we are we did a big deal with nscale for 2027 that >> did that deal 100,000 GPUs three plus billion dollars. When's that come online? >> Um >> those are Ver Rubins. >> Yeah, those are like second half of 2027. Those are VR 2200s by the best computer in the world. >> Yeah. >> Um >> wait, what what's the ratio of training compute you'll have on that day versus say a month ago? >> Oh man. Um let me do the calcul. There are different type of GPUs chips than we
[00:25:00] have now. >> Figure the VRs are 2x like raw. He's using B300s today or something like that. >> Yes. So, we let's see. We have they're roughly on order of four >> 4x. >> Yeah. So, um man, uh yeah, like maybe like if like we took the whole uh deal with in scale and size it all the way up well over 100x what we have. >> Over 100x. >> Yeah. >> So, going going back to the question of >> Yeah. I mean we spend like we also spend like uh nine figures a year now on compute with before this like we're like you know spending like a large amount of dollars on compute here internally >> intimidating to everybody else. Yeah. >> Inference on on the robot on Wi-Fi. >> Um we do all inference and the robots are seeing it here uh on board today. >> So you've got what on board in terms of in terms of compute? >> Yeah we have um all of Helix is fully on board the robot. Uh so everything from uh like the planning and the reasoning to like even the low-level like sending uh trajectories out um and torques all
[00:26:01] that's done on the robot today. And what gets done uh on a main on the uh >> the only thing we do today is we we do like real-time speech to speech uh through servers >> and we use the servers to help um think about like we we think of the robots like agents like physical agents and like um like you know on the um you know if you think about like if I had a hundred robots building a home or building like like in construction like how do we get the robots to like how do we command that whole fleet like minute to minute hour to hour like cuz as the home's getting built like the job's changing the friend the friendlier Borg. >> Yeah. Yeah. Exactly. Um so uh so that's all done from our agent orchestration in the cloud uh today. Um which is the only way to do it. We have to like we have to like uh in this case is we have like multiple even on this conveyor here robots are subbing out they're recharging. Uh if a robot gets damaged or if doesn't feel healthy it walks to maintenance and we have another robot comes in it flags the fleet. So like all this is beingworked uh together. So um that has to be done in the cloud. Um so that will always be done in the cloud.
[00:27:00] uh things there. We're we're experimenting with like, you know, we've always have like with onboard and offboard comput. It's really hard. I think the struggle we have is like we can't get around the physics of having like very high dynamics on the robot. The robot needs to be able to you know um like we need to be able to talk uh to all the joints and sensors um and uh and you know and then communicate back what the new plans are in a really fast way. We don't have like seconds we can wait or even like half a second like we have like we need to get done in like like like um you know in the millisecond level. Uh so like there's no getting around that as we do like really fast dynamics of the world like that's um >> so the the you know in the best case scenario you have like a big brain living in cloud you have a small brain living on board in extreme example you have everything living on board which we're doing now. Uh, so, um, you know, jury's still out. We're still doing tons of experimentation, but all the robots right now are allorked on board. I think it's kind of nice because we can lose internet and the robots still do work. Um, which is nice. Like if you're even at home, you have like bad 5G service or
[00:28:00] something. You don't want your robot stopping mid laundry. Sure. >> You want it to like continue to do work. Uh, so there's a lot of different like like engineering trade-offs we have to go through to make totally understand this better. >> So milliseconds, is that even viable on the cloud? >> [laughter] >> is the the home Wi-Fi is a bottleneck, but then getting the milliseconds. >> I mean, the really the really the best 5G networks we can we see that you can do is like, you know, you can cutting it down to like maybe 20 to 30 milliseconds per leg. So, you're looking at like almost like 50 milliseconds just for like, you know, uh, a full round trip. Then you're, you know, then you're already at you're at 20 hertz just in the network itself. And then you have all the forward pass you need to do on inference. And there's a bunch of other stuff we need to do with uh as it relates to the images and torques coming back. So like uh it's not you know it's not just as easy as like pinging a cell tower and getting it back. We have a lot of other work to go do. >> Do you think then every home is going to have some kind of a brain box in it in the basement where the cable boxes? >> Um our hope is that everything's done on board and we have the right architecture now. >> But if every home has like 40 of them, you know, do you really want to [laughter] that's a good market projection. 40 per
[00:29:01] >> No, I mean everything's done on board like we do have on board right now. like the robots still need >> chips are scarce though if you can share it you know because you might have like a roof you [snorts] know gutter cleaning bot up there >> I don't think whoever like listen there'll be some neural nets that that no matter what will have to live on the robot like they're just running too fast >> sure it's all about training um world models versus physical versus you know what you've been doing in terms of uh task video recording what's your vision on on training to get to the level of fidelity you Yeah, we um we took a pretty big bat on Helix 3 uh here, which um we talked a little bit about on the all hands, but we haven't released anything publicly yet. Um our if you roll back before that, we released Helix 2.5 a few months ago where we basically showed like zero shot reasoning in in like 30 unseen homes before. Um we also showed the first like what we think is the first humanoid scaling laws for data. As we like as we increased data, we saw like
[00:30:01] uh very like log linear relationship meaning like um the models were like like model losses were coming down and the robots were getting better like actually executing better in these environments with more data than with less data. Um which is unbelievable to see. We've seen this in like LLMs. We haven't seen this in robotics before then. Um and if this data holds, if this if this trend holds, then basically as we're adding more compute and data to the problem, the robots are literally going to get smarter. They're going to get better at the task every single day. Um, we have a new, um, work we've been doing on Helix 3. Uh, I think we gave you guys a little preview today. Um, and, um, I would like largely characterize this as like little baby AGI. >> Like, um, we like have like a dude in a suit in the office. We have a secret lab called Blackbox you guys were in, which is like a little home. >> You said we're the only ones that have ever outsiders that have been in there. >> Yeah. >> Wow. That's great. It it's it's a mind-blowing experience because when when uh the LLMs were improving, they didn't really look anything like a child. Like they were they were doing
[00:31:01] complex math, but they were also hallucinating. It's nothing like a kid. But what we just saw in in the blackbox looks exactly like a kid learning how to do things. And it's it's completely, you know, nothing is is pre-programmed. It's all neural net. You ask it to do a task it's never heard before and it looks exactly like a kid who's learning how to do it and experimenting and trying and rotating and you'll pick up the chair and like it's not going to work and and you can just visualize how this is evolving just like a kid. It's it's so much cooler >> than the way the LLM's evolved. So your production rate right now, I mean the demand for these robots in in homes, you know, I'm asking people all the time if these things are costing you and are you still tracking towards a uh $30,000 price tag, 300 bucks a month type of lease rate? >> Yeah, the the in the home will have to make the robot really affordable. Um
[00:32:00] you'll need some sort of like $300 to $500 a month lease. Yeah, 10 bucks a day. I mean, it's like under a buck an hour. I mean, I ask people uh when I'm on stage, you know, how many of these would you own >> if it was that? And and you know, the average individual is like at least two. >> Yeah. >> Yeah. You got to either lease it or buy it. There's only two ways. We got to get the robot out. Um >> do you have a preference? >> Not really. I think through a lease, I think you can really bring the cost down quite a lot and help advertise that better for the us like for the customer. Um, I don't really know if in and I think the technology would be radically improving. Like um my my view is that we're like pre like we're like we're like in the we're like in the flip phone era for humanoids. >> Um I think um I would make the maybe call that >> figure 4 is probably like iPhone one moment we're going to have here soon. >> Uh it's just it's just such a significant um step up from where we're at with um you know figure 3, but we're like really early. So you know in the iPhone even from like gen one to gen like 8 10 it was like r every year was
[00:33:00] like man this this hardware is getting better cameras are getting like the >> slightly better camera >> now you can't really tell I mean it's like all it's like you know it's like asatomically approaching like some perfect like you know iPhone vision for them >> but like um you know for us we're really early so I think the robots will continue to get like much cheaper uh more dextrous like uh just like overall safer to be around. Um I think a lot of the um way you interact with the humanoid 2 will will really improve. Um so um so I suspect if you do a lease model you can really like get the newest version uh every every few years which will be good. >> It feels like you've been uh your nextg it's been like a 18month center on on next generations. >> Yeah we we like um we we really want to do our next gen at significantly high volumes. So we um we spend a lot of time on it and we'll spend a little we spend about a year between programs right now. We're spending we'll spend more than that on >> what might we see with uh with figure 4. I mean what what areas you focusing on making improvements?
[00:34:00] >> Oh man. Um that's a yeah a little I think the big drivers here at a high level like the at the requirement level are like we we need the robot to get cheaper. Like that's super important. We'll induce more demand. um you know like so like we're we're trying to like reduce cost as much as we possibly can. >> Um we're trying to make the more robot more capable like you want it to be able to do more things in the world than it has before. Uh you know like a little sneak peek at that is we've been working on the hands for like uh around almost like two years now and uh we're like probably the only group that's built like tendon driven hands and non-tender driven hands. We've done now seven iterations of our hand. Um, we I think we really understand hands really well. And um, one of the biggest step ups we've ever made in the hands is coming. And um, I would make the argument that like a figure 4 with a glove and a human with a glove is almost indistinguishable. >> Really? Are you serious? >> Yeah. >> That that's like an iPhone moment. >> Everything I've seen so far is like they're they're functional,
[00:35:00] >> but even the way they they grip things, you know, when it was picking up the cups, it's sort of reaching into the cups rather than grabbing them on the outside. I hate to say this, but like because like I think Figure 3 is probably the best humanoid on the market by far, but it'll make Figure 3 look like a caveman. >> Really? [laughter] >> Yeah. So, where are you in terms of [clears throat] >> It's funny because the the iPhone analogy there is really interesting because, you know, the iPhone 3 was the first really big hit, but it had no camera. The camera came later and that was a step function change in that trajectory. I'll bet the hand is the equivalent of the camera. You throw that hand on there and it has human level dexterity. I just think like you know we've been if you look at like all three robots they've been you know they look different for us. We've obviously changed a lot across the three. We changed like you know from one to two. We changed like the architecture of the structures from uh like um you know uh it went fully from endoskeleton to exoskeleton like all the loads on figure 2 like all the exterior take loads almost like your your aircraft does right >> and how we build aircraft at archer like the so it's a very different type of architecture we went uh as we like designed the uh robots out but you'll
[00:36:00] line all these up and figure 4 is it doesn't even look like the rest. >> Mhm. >> Like it's just like a what happened here? It was like a full deviation from our, you know, our lineup kind of look like it's maturing and there's like just a four fork >> and um that's what I mean by like it's so significant of a robot. We >> um had a couple hunches of like where things should be headed and those we ended up doing a bunch of work on those and those end up like ended up panning out really well. Uh, so we took a couple like pretty aggressive bets on complexity to get there and it it's it's Yeah. Anyway, fair to say like I think figure four will be like the largest step up we've ever made between generations. >> That's pretty extraordinary because two to three was a huge step up. >> You've never seen it like this before. It's like the largest step up is in figure 4. >> Maybe you'll maybe you'll sneak us a peak later after the after the pod. The on the hand touch is super critical, right? Being able to get that data. Um, you know, a friend of mine has a company called Touch Lab that is basically uh built out the most advanced touch systems for gathering data. Are you
[00:37:00] doing that all inhouse right now? >> We do all in house. Yeah. Um, we like >> the problem with like a lot of these exterior external groups is um we we have to be production ready. They have to work at really high reliability. It's got to work perfect for our system. It's going to be a low cost. Um, we need to like um rev it and fix it if we understand it. We need to like understand the firmware and embedded software around it. Like we need to understand thermals around it. Like so like it's not I you might might seem like we can just go off and like buy something and toss in the robot and makes this magical experience. It's never the case. It never >> same same thing for how Elon concluded. You need to vertically integrate every >> never fits like you can't even make it fit in the in the CAD. Let alone like the thermals don't work or it's not the right bus comes or like the um you know the power loads are too high or the cost is unfeasible or we can't it doesn't work past like five or six times and it breaks. like none of this stuff is like um you know it's like it's and then and then we have some of the best engineers in the world here that want to move really fast. >> So then we're working with another group that like isn't like beholden to us to like be able to do that work and it becomes really tough and that's what
[00:38:00] happened with figure one. We went out hoping to go get the stuff and pull it all in on the supply chain side and it just like all broke down really fast. Um we bought the first actuator. We brought it in. I think we still have it here. It's like okay we're going to use this for a elbow. This is an elbow actuator. It had all the right stuff in it. that, you know, had a torque cell and had the right encoders and the right gear boxes and the speed and torque profiles of the actuator were in the right place. And we got it and it was just like this big. >> It was like literally like this giant block and it's for your elbow in >> China probably. >> And it was like we were like where what do we do with this? >> Like it's just like the best one on the market. And we're like it's like literally three times too big and three times overweight. And um and it was really expensive. Um you know we need these like we can't have it hundreds of dollars. It needs to be like tens of dollars at the most. And anyway, um, so like, uh, we had to go build everything ourselves. So, we're >> now somebody in that room says, "Okay, Brett, we're gonna wind our own actuator." Like, that's that's not a small step. I mean, >> no, we like we sat there. We said, "This really sucks." Uh, we have to go do it
[00:39:00] ourselves. Is there any way we're wrong about this cuz we don't want to have to go do this. And we sat there for like a while trying to figure out every other possible thing to do besides like not build it. Like, if you build it yourself, it becomes really really sucks. You have to have like you have to have a firmware person. You have to have electrical person. You have to have a thermals person. Probably some work in embedded software. Uh you have to work in mechanical structures. You have to work in the motor which is like the rotor and stator and gearbox. And you probably need a gear designer at that point. >> And then you need somebody to then uh put this all together >> in a really small package because our design team wants it to look really good. And then it needs to be able to move around, which means like the wire and the cabling and all the harnessing and connectors all have to be really dynamic and not ever mess up. Um and then you got to figure out how to get the heat out uh of the system. Uh, and then you got to figure out how to where to put them and how to daisy chain them all now together to build like a real like a, you know, a functional like actuator inside a shoulder or leg or hip or whatever you're going to do. Um, and then it becomes like, man, this isn't really like a go get it and you put it together plan. This is like a it's got to be done from scratch. And every time you make something better on the
[00:40:01] engineering side, it gets like it could get bigger. And if it gets bigger, like uh you can't have your elbow like it's too big or your knee. Um, and sometimes it clashes with the rest rest of the robot. Um, and then it doesn't look great. you really want like a really high like high aesthetic product. So, um it becomes like a very difficult conversation to have across like 10 people. Do you think like a year from now if you think about the supply chain coming into this it's just raw molten metal and bare wire that you're starting from or like if you think about winding your own actuators like you're literally winding coils like okay but at the end of the day you're not going to make your own CNC milling machine and you're not going to make your own die casting or maybe you do your own die casting I don't even know but >> but what does it look like >> you're referring to the conversation we had at the gigafactory where raw aluminum was coming at one end and cyber cabs were coming out the other end >> buckets of molten metal just pouring and you're like, you're kidding me. Like it's right out of a sci-fi movie. >> I mean, that's it. To get cost down and um some of those times, yeah, some of
[00:41:00] that has to really happen. Like it's um Yeah, >> I can just imagine uh you're you're here to your team saying more robots. We want more robots walking around. Just so much fun. Let's talk about the robot. >> It's nice though, right? It's not That's not nice. It's amazing. You're living in the future. >> Even the weekends are all running like and like it's a bummer when they're not running. Like you just come in here like it's just like like >> how do you have how many do you have at home? >> And how are your kids uh dealing with it all? They enjoying you have how many you have what two or three young kids? >> Three kids. >> They're just like they're really young though and they just [clears throat] like want to run. They run over and they try to like jump on the robot. Yeah. >> I'm like you can't really jump on these robots. [laughter] You know what I mean? We're not there. We're not there yet. You can't jump on them. Um and they come to the office like every few weeks. They love being here. Uh all the robots and stuff. It's great. So, um, >> well also when you walk in now, there's a hundred of them outside, just outside the door, >> and you don't even feel vaguely like you can't walk through them. You just walk right up to them, and >> you know, they're they're moving smoothly enough that you don't feel like it's going to slap you in the face, >> which doesn't mean you can have some conversations with them on the way out.
[00:42:00] Let's talk about the the figure face cuz that's an interesting, you know, set of choices. You know, the unconny valley of making something look too humanoid. >> Right now, you've got a black screen. What's been the trade in the in your internal conversations on what should the face of figure look like? >> Yeah, >> I'm sure there've been a lot of conversations on that. >> Yeah, we still have conversations on this. Um >> the the face is really tough cuz like we're really missing a lot of the like facial movements of a human. >> Sure. >> So when you talk to the robot, it has like you know lights and screen and it has like obviously the gestures it can make with neck and the whole body. Um, but it's just there's just still something like missing to give it like a humanlike experience where we're like we're, you know, the face to something more humanlike cuz right now it's a it's a black face. >> Yeah. >> It's um, >> you know, it feel it feels friendly, but it's but it's a little bit ominous to look at a a you know, a shiny black. >> Yeah, we're working on we're working on
[00:43:00] the face. Uh we have some plan big plans for figure three uh on the head and face and um we're working through this. It's um it's just a hard problem. It's um I think >> it's a human psychology problem. >> Yeah. Like you really want the like I think like we're seeing more and more internally we're like talking to the robots more. We're spending time with them. Uh the robots like have like memory now. So like if you like if like um we've done work now where you can like like I can go up to robot it's like hey Brad good to see you again. And like remembers our last conversation. Mhm. >> So, um you're getting now to the point where like you might want to just like like almost like a companion. You want want to talk to it, especially in the home when it's around like family and you've had it for a long time. So, uh getting this we call HMI, but again it's human human to machine interaction uh really like well dialed is something where we have like a whole team on it. We have like an HMI lead. We have like a whole focus on it. Every week we meet and do planning on it and demonstrations. It's it's really hard. A lot of it comes down to like um not just like the face, but like how the body reacts too. Like we're just like pretty expressive with our like uh like obviously facial uh face like arms like
[00:44:01] all this as we're talking uh making small little gestures and nodding. Um all this really helps in my mind of building something that's more natural to interact with. Um so we have like um a lot of work we're doing there. Um and then yeah, the face is a tough one. I think the thing that we've like really stayed away from is like putting like googly eyes on the robot. Yeah, >> I think it's like a really um um how do I put this? Um I think like this is like one of the most uh like uh complex pieces of technology in the planet. And if we're there putting like googly eyes on a robot and saying oh it's like friendly and it's like oh it's like here it's like it's like uh safe is just like a huge disservice. Uh it's actually pretty it's it's it's like pretty it's like a it's like a con. It's like the robot isn't has googly eyes. it can like literally, you know, walk upstairs, open doors, like it's one of the most intelligent things we've ever built. Um, so I think we're we're trying to kind of um hold a certain standard for design that we show like some like show what the techn is really capable of. Um, so
[00:45:01] we're working through this problem now. I think it's a really hard one. It's one of those things where we have to get in a room with like a lot of people and we have to really kind of um like discuss like where the future's really headed and then we have to do a lot of like product iteration really. It's something you could test countless inter uh iterations of and how you know kids and adults interact with it. >> But I mean the end goal is to have at least the consumer side a home companion that you trust and is part of the family. Yes. >> 100%. >> Yeah. >> Yeah. You want to be able to like talk to like a human and you want to have like persistent memory. Um you might want to spend a more percentage of your time like like hanging out with the robot than you would like having to do work. Um, so my view is like in the home it splits to like you really want a companion there in the home that knows everyone can be really helpful. A kid comes home from school like sad like asking why has um and then you you want something to be able to do work like we all we all do. Like this morning I was like getting stuff out of the dishwasher like um can't believe I don't have my
[00:46:01] robot doing that right now. You know what I mean? Like um but hopefully you know hopefully by end of year like we're doing that pretty repeatedly. Um, so we're like working through, you know, all those things, but like, you know, we want to be able to connect with real humans every day. >> So on persistent memory, you know, if you have a pet, a dog or something like that passes away, you get a new dog, you've got a new, you've got a new dog. But if you have an iPhone, >> when you get a new iPhone, you just port everything over and the personality has now moved to the new device. So here comes your robot. You've fallen in love with it. It's got persistent memory. You've done a ton of tuning if you're like, you know, I don't like my laundry folded that way or you shouldn't be doing that. You should be doing this in the morning. All of that state is now in your bond with this device. So, does that live with the device >> or does it become like your iPhone where you port the personality to the and the other one just sort of, you know, >> it definitely it definitely does not live with the device? You'll be able to like have that pass through any device. >> And then what if you have two of them? Can you do you swap personalities midday? >> Yeah, we're doing some testing here where like what if you're like walking
[00:47:00] by like these hundreds of robots and they're all talking to you >> and all you're walking by. >> Yeah. >> All the way through. But you're walking by like all the way through campus and you're having like a singular conversation. >> Well, that's that's always mind-blowing actually because when you see them all lined up and they start doing things in in synchronous lock step, >> that is just a mindblow for anyone who sees it because you think of them as people. >> Yeah. >> And then you see them coordinated in a way that people could never be coordinated. >> So the equivalent when you're walking through is if you're having a conversation and the conversation is moving from robot to robot as you're walking down the line, >> you're like, "Wait a minute." And that could be good or bad, you know, because like that uncanny >> the shared mind. >> Yeah. >> Um when I think of these robots, I would want to personalize them. Um both cognitively, but also physically. You know, right now they're all wearing the same outfit and costume. I assume you'll have the ability you can design your your robots. >> It's like your iPhone case. >> Yeah. Outfit. >> Yeah. You're going to want to design your robot. >> Yeah. >> Yeah. For sure. Um Yeah. So, these
[00:48:00] robots can all be outfitted with different uh like different soft goods and apparel. Um, we have some here we can show you on the way out. Um, but yeah, you'll be able to like fully accessorize your robot. Give it whatever personality you want. Change its voice. Um, >> what about tools? Are there going to be figure branded tools that are >> What do you mean by tools? Like a hammer? >> Well, [laughter] like a hammer, but but you know, like a like a a mixing machine, you know, I'm making cake, you know, and it's like, okay, well, >> that' be pretty cool. Like appliances. Yeah, >> appliances. Like, you know, >> I spend half my life in the longevity business. The favorite example I use is if I've got a bunch of sensors in my body measuring my protein levels, hydration levels, peptide levels, and so forth. My robot in the kitchen should be cooking me a meal specifically for what my body needs at the moment. >> Yeah. >> Yeah. The thing about tools in general, though, is like right now we think, "Oh, I've got all this human-made stuff. My humanoid hand controls it." They're like, "Well, wait, now we have for the first time ever the ability to snap on an entire function. Just snap it in." Or
[00:49:00] I can I can mentally control it. Like I don't need to if the device has your IQ from Hark or from from Helix built into it. I can just mentally control the device. I don't actually have to walk across the room to mix the cake. I can just brain it over because cuz I can >> turn the oven on without even going to it. >> Yeah. Exactly. It becomes totally integrated in your whole life. >> Let's talk about safety. Um because I mean one of the one of your goals you said it in the beginning for getting this out is it's got to be safe. >> Yeah. >> Uh safety insurance. Um how do you think about building in the right safeguards uh for these robots because they are very capable. They are powerful to do what they need to do. Yeah. >> What's your philosophy there? What are you building into the system? >> Yeah. to make it really safe, you have to work from like basically like all the way down to like um like the substrate level like almost like in nonvolatile memory like really understanding what those rules look like and affordances for the robot all the way up to like the system like architecture design um like
[00:50:00] a system safety architecture. Um, it's really tough cuz you have like you have like both like physical safety and like I would call like more like semantic safety like you have um like the intelligence of it needs to also be pretty high like you can't um you know you don't want a robot walking around and knocking over a candle or like you know spilling the hot water. Uh these are things that like you need to understand there's like risks in those areas and how to deal with them. Uh I I think separately like you know we have like 40 plus actuators on board like uh you don't want any of those like failing. You don't want any of the robots falling or um you know uh especially with like you know say kids or pets and stuff in the house. Um so we working like some of like like we spent so much time last say two or three generations now thinking about like what is the right architecture to build like a real-time system that doesn't really have like that approach is like never having faults and issues
[00:51:01] and I mean you've seen this in the world like we have like turboan engines that are rated to you know 1* 10 - 9 hours before catastrophic event um and yeah so you have like you like systems very complex systems out in the world that are have um we've gotten to like really high safety standards um self-driving cars we need to approach this like things like this um so I think um we we see a clear path how do we like design a system that gets to a point where it's like it's safe enough to scale into like the millions and billions um but beyond just like the system architecture there needs to be a lot of like iteration design of like being able to build and then test uh learn and then rebuild and retest And um and then you basically need to build a track record behind that. >> Let's talk about like you know these are running software. Software can get hacked. Do you have a master off switch that can be you know physically actuated? Um you know we've seen agents escaping the lab. You know it's been it's been all the conversation on the podcast. >> I was with Danielle Aruse who runs Seale
[00:52:01] at MIT yesterday. They have a whole bunch of Chinese robots that they bought because they're really cool in the lobby and they've they're all transmitting data back to China. They're recording video and audio. >> So, I'm not worried about that for I'm not worried about that for Brett, but [laughter] >> it's actually an advantage in the sense >> I am thinking about >> we buy all of our competitors Chinese robots [laughter] >> having a robot hacked in the home uh by some outside person. So, how do you put the you know cyber security or you know physical off command uh if needed? Is that something you're thinking through? >> We have a whole cyber security team just all they do is deal with this. >> Yeah. >> Yeah. Both on the product and corporate security side. Um >> it's actually a pretty large team as it relates to headcount here internally. Um yeah. I mean listen we like self-driving cars have the same problem like you don't want all the cars on the street to be like hacked and driven into pedestrians. Um same same with us like so we have yes we have like safeguards in place for this work. Uh we're contin um it's an evolving thing. It's not like
[00:53:01] we have like a silver bullet where like we push like one line of code and we're like yep that will solve everything. Um because it won't. Uh so uh we have you know basically a lot like tremendous amount of work happening from cyber security system system safety engineering all the way down into like what are like the fundamental rules and in within nonvolatile memory that we need to put into the robot so it behaves like properly in the world. Actually, when we met nine months ago, you mentioned you were well down the road to having no Chinese or third party components from outside the country. Have you made a lot of progress on that since then or >> Yeah, we made a lot of progress last year or two on this point. Um I mean it happened because like at first like tariffs just got so expensive. >> Yeah. >> I think at one point like last year we were paying like one one week was like 270% tariffs. >> It was like you know during all that negotiation stuff and I was like we like it I think we were like this is going to end soon. >> That's an incentive. Um but it was like yeah this is not good and then obvious obviously like the geopolitical stuff like really flamed and um so so yeah we've been rotating uh like fully out of
[00:54:02] like just different piece parts and stuff we have out of China. Um so at this point it's like in a really good shape. >> Have you seen a uh a robot insurance industry pop up yet >> because I'm assuming there will need to be insurance models wrapped around robots. >> Yeah it's interesting. We haven't seen any of that like both commercially and and consumer will'll need this. >> And in terms of financing the robots, are you going to provide financing and have outside financers coming in? So instead of buying for 30,000 Yeah. I have a lease package. >> Yeah, we we currently have a lot of demand for folks that want to finance robots for us. Uh right [clears throat] now a lot of it's like focus on commercial contracts they want to finance. Uh there's no reason why this won't happen also in the home. Uh this is a multi-t trillion dollar space of doing like you know credit card receivables and lease financing for automotive like um and we're working with like the biggest companies in the world around this. Um but yeah, we'll we'll definitely um you know I don't know if we'll personally finance them, but we'll work with different like banks and funds and stuff to do this work.
[00:55:00] >> Yeah. >> Yeah. Do you think about space at all? Because I I know this seems like a stretch because we're in the kitchen right now, but >> but no, there's there's only like three maybe four people on the planet that have the wherewithal to train >> a $3 billion foundation model for physical world. >> Yeah. >> And and meanwhile things are starting to go into space, you know, not not just on Elon's rockets, on other rockets, too. If you think about Eric Schmidt launching stuff into space that needs to be maintained or changed or whatever, he doesn't have that foundation model, that physical world foundation model. >> So there there's probably, you know, these guys in space, but there's also like no, my model controlling something else. Yeah. >> But that physical knowledge model is just like this this insanely valuable centerpiece. >> Yeah. It's funny you asked like you also mentioned like you know like this AI stuff like really happened really f when I started figure we called it figure AI our mission was to advance AI for robotics and we wrote and my whole
[00:56:00] master plan was like you know this is there was going to be a convergence of AI and robotics and it's going to be really powerful and it just you know I thought that maybe would happen 10 20 years later it happened like in next year >> and uh deep learning started working and in that we also talk about like a key focus of trying to get the space >> um So, it's an area that we're like really interested in internally. We actually had a couple we've had a couple conversations on this last six months >> and uh one group told us it costs like tens of like low tens of millions of dollars per hour to put a human in space to do work. >> Yeah. >> And uh we're like, "Wow, that's crazy." >> Well, it's really bad for PR when things go wrong, too. Totally. You know, it's >> I think a humanoid should be great at this. >> Exactly. And that's the whole objection to data centers in space and other things in space is maintenance. >> Throw a humanoid out there to do it. >> Yeah. Exactly. And how cool is it like floating around? [laughter] >> We we we had a team here, a space team here, and they're like, you know, how is like, you know, are you IP rated to this and this and this for space? And I'm like, oh my god, no. Like [laughter] like you know what I mean? Like uh >> vacuum rated joints.
[00:57:00] >> Yeah, exactly. So like they're like I've got a lot of work just I'll be I'll be candid. We are not ready. [laughter] >> You you very famously said you're not going to do uh defense work uh with Figure. Is that still the case? >> Yeah, it is. >> I I respect that. I think it's kind of hard. I think I think it would be difficult to do defense work by also trying to sell robots into your home, doing work in like everyday commercial, you know, places that we all know. I think it just becomes really tough for humanoids. >> I think we need to like you need to pick a route. Um, you know, we're picking like, you know, the whole civilian side, which we think is thousand hund hundreds of times thousands of times bigger than military applications. And it's also like an area where all of us here at the team at the company really want to solve. We want to solve a robot in the home. >> This episode is brought to you by Blitzy, autonomous software development with infinite code context. Blitzy uses [music] thousands of specialized AI agents that think for hours to understand enterprise scale code bases
[00:58:00] with millions of lines of code. Engineers start every development sprint with the Blitzy platform, bringing in their development requirements. The Blitzy platform provides a plan, then generates and pre-ompiles code for each task. [music] Blitzy delivers 80% or more of the development work autonomously while providing a guide for the final 20% of human development work required to complete the sprint. Enterprises are achieving a 5x engineering velocity increase when [music] incorporating Blitzy as their preIDE development tool, pairing it with their coding co-pilot of choice to bring an AI native SDLC into their org. Ready to 5x your engineering velocity? Visit blitzy.com to schedule a demo and start building with Blitzy today. I am curious. Uh, you know, when the bird scooters started showing up in my hometown of Santa Monica, um, people started sort of like, uh, trashing them,
[00:59:01] hanging them from trees >> and so forth. So, I I'm curious when we start to see robots walking around the streets, >> uh, how do you keep the robots safe from the humans? >> Yeah. Yeah. Yeah. I guess we don't want to be like slapping humans, right? >> Well, it's it's more it's more, you know, >> the question. >> No, no, it's it's you know, there's there's a a significant amount of >> we can train them >> of civil um fear, distrust, anger towards technology. And you know, you saw the Whimos burning in LA whenever that was, you know, 18 months ago. Uh I can imagine the robots themselves need to be kept safe from humans. >> Yeah. I mean, we're certainly not going to train any robots to like defend themselves against humans. So, um, to the extent that a human wants to trash our robot, I guess there's really not much we can really do. >> A lot of cameras on that robot. >> I know, but like, yeah, we wouldn't want to like, you know, we wouldn't want to
[01:00:01] >> try to defend ourselves or try to >> um, stop us in any way. I think um, >> you know, I think it just is what it is. So, I think uh I think you're going to have some of that, but I think you know net net is probably like a like a very small percentage of what really happens. You see this way, too, but it's you know what I mean? It's not like um I I I think it's just something we're going to have to like like basically like build through. >> Yeah. >> Yeah. We haven't had it happen since. >> We're a little worried like is there going to be a blue van drive up and like snag one of our guys outside? [laughter] >> Yeah. But again, take it apart. >> Yeah. Yeah. It's funny. >> It's really hard to imagine anyone getting away with that more than once. You know, there's so many cameras. It's all connected. >> Really funny. The robot's like in the back. Listen, man. You don't want to do this. You don't want to do this on timing. [laughter] >> I have like I have I have a 2 kilowatt hour battery pack. I can light on fire and take this whole thing [laughter] down. Don't Don't test me back here. >> Brett is not going to Brett's not going to, you know, pay a ransom fee. >> Yeah, [laughter] he won't pay. I've already asked him before. Don't do this.
[01:01:00] >> Policy. >> Be really funny. Just like trying to talk him out of it. [laughter] Can >> we talk about exponentials for a minute here? Peter's favorite topic actually. >> Go for it. >> Like, so botq, are we allowed to talk about botq? Is that >> Well, what do you want to talk about? >> I want to know the exponential self-improvement equivalent going on right now. Yeah. RSI in the physical world is the coolest thing ever. >> Yeah. >> You know, Vonoyman probes and >> but it feels like they're very close to the point where not not where they build themselves like winding the the actuator, but they're controlling the machine that's winding the actuator and then it's a closed loop. Yeah. >> So, I've always described it as like the way this is this is my kind of takeaway from the Gigafactory is that it's not robots building robots. It's robots controlling machines that build robots that control machines that but it's the same effect. >> It's a closed loop just exponential takeoff. Yeah. >> Of just the volume. And you can see it, you know, when we were here before there's like a couple dozen. >> Yeah. >> And now it's like I don't know if the cameras can catch it all, but there's just massive scale going on. There's >> Robots everywhere. >> And you said you slowed it down, too.
[01:02:01] >> Yeah. So like when when does this hit like >> we're still making a lot. It's just like we have a lot of robots coming off. So it's um yeah we're like we're in the process right now of scaling back up as we go into like December December January. So um >> So when do we hit RSI and robotics? When do we hit robots building robots? What's your what's the time frame for that? >> Yeah. What's the very first robot where no human labor went into it at all? It was purely >> Yeah. Like full lights out manufacturing man. It just feels like we're man it feels like we're so the the pro the thing is like every quarter that goes by the progress that we're seeing here is just like it's just astronomical >> and I have to be real with you guys like when I started figure it was like I was so excited about being able to work on the humanoid pro like you know problem. I think it was probably one of the most important businesses I I could be building. And I was just like, I'm going to just go do this for 30, 40, 50 years, like just the rest of my life. Like there wasn't like, okay, in 2020, you know, 2024, 2025, we'd see all these
[01:03:01] breakthroughs and it would just be like on a full exponential. And in that way, we got really lucky. Um, you know, like basically deep learning showed up a year later, uh, in full force. Uh, the robot hardware just got really capable. It's now doing work with neural nets like fully end to end like we're seeing out there. Like um it's it's nuts and it's just getting better and better. We're now, you know, we just took a robot uh uh down to a client site uh that it's never been in before. It's never been to this the warehouse or anything else. This is like three weeks ago. And it like fully zero shot the whole thing. >> Yeah. And >> like before like you know lighting's light would change or like the conveyor would be different, package mix is different, the robot would just get confused. It was just hard and it made sense like right never seen any of this stuff before. Uh we did larger scale pre-training with Helix 2.5 and that we took that that model down to this client site and it just zero shot the whole
[01:04:00] thing. And if you asked me a year ago if that was possible a year from then I would have said like you know that would be incredible if we did it but like you know certainly the next 5 years. So the same argument here it's like everything's just improving so fast >> and we're seeing like things like the hand for figure 4 and even figure 4 itself. We're seeing the work we're doing on Helix 3. We're seeing the amount of data coming online. We're seeing amount of compute coming online. Like we have like several billion dollars of cash right now like on the balance sheet that we're working to deploy. Like >> congratulations for that. >> Yeah. Thanks. It's just so like it's like you know does it seem certainly within five years for sure >> for still feels for sure within five years we have like a full lights out manufacturing um but it's like how do I like you know but it's like foggy if it's going to happen in like 6 months 1 year 2 years 3 years but like it's right there like >> lights out manufacturing lights there's no humans there no power no lights needed >> like full lights out manufacturing we call it so like um and we designed figure 4 to help with this problem uh like we um there was a bunch of stuff with figure 3 design that we felt were really tough to do like full closed loop
[01:05:01] lights out manufacturing uh that we designed into figure 4. That's why we're saying like iPhone one moment. It's like part of this is like we think this is the robot that will like build itself and uh >> let's flip it around and ask the other way when the new hand comes along like which is imminent. Is there anything you can think of in the current manufacturing process where you know here's a person controlling milling machine or whatever whatever the whatever component [snorts] anything you can think of that won't be able to be done by that new hand >> there's like blue color job opportunity like what's that last thing that somebody needs >> not really man I mean I would have said like maybe ladders and this like you know some of this forceful interaction stuff would be hard but like the robots like climbing a ladder next or like it it'll just climb over and over every day like RH just going up and down ladders and uh we're like doing like very forceful interaction work now and carrying big boxes around our 50 lbs with the neural network. We're we're unloading things in a truck. We're doing things in a home like dishes and folding
[01:06:01] laundry. Like it's just like all the puzzle pieces are right there. >> Yeah. I mean the ultimate functionality for me is when is a robot doing surgery, right? When is when is a humanoid robot the best surgeon on the planet? Well, the the timeline probably lights out improvement to that moment. >> I mean, it you know, so we've just seen, you know, AI become the be best diagnostician. I mean, the numbers are blindingly obvious that, you know, if >> AI versus human >> AI wins every time right now and then you know the next question is when we have a robotic surgeon. So that you know it's federated learning. So every robot knows every surgical case ever done. It can it doesn't have any jitters. It is didn't have a fight with its boyfriend or girlfriend the night before. >> It's there and it's able to do surgery. And what's fascinating is it's you know a dollar an hour for the surgery. Yeah. >> Right. Any any place on the planet. >> I mean I think that's one of the >> I also feel like I've been like half the
[01:07:00] time I ever visited a doctor in my life I've been misdiagnosed. Yeah, >> I feel like just having a network of like the smartest thing in the world with you that can be there with you in a room checking you out would be like an incredible be like the world's best doctor for every every country in the world. >> Yeah, we're we're there. We're there now. Um and then the question is you know when are these robots the best surgeons in the world where you you know you go and you need something you say I do not want that doctor touching me. I want the robot does it perfectly every time. >> Yeah. >> I think uh I want to capture this though. that hand comes out imminently and we can demonstrate every single step of the manufacturing of the figure 4 being done with that hand. That's a big moment in human history. And I I think we might want to just document that, you know, just like literally every step of this entire pro. Here's a something that arrives in a box. Open, assemble, build. >> It's like Yeah. I mean, guys, here's what I see. We see basically like a task we'll pick out like we need to be able
[01:08:00] to do this or that. We go get data for it. We train Helix. The robot then goes goes and does it. >> Yeah. >> And we've done that now over and over now for the last like you know couple years >> and we're like man this robots are really capable. How do we get it to generalize is the real thing. Like how do we we wanted to be able to walk into Baku >> with a minimal set of instructions like whatever training a human would get probably some slides and some documents and maybe like like shadowing somebody for a few days. Yeah, >> we want the humanoids to be able to go that same process. And like that's really the chapter that we're in right now, which is like how do we scale intelligence? How do we get this thing to do it from a hardware perspective? The hardware is going to get as dextrous as a h you know hands. This is going to it's going to think about it as a keep a steady rate of improvement. >> So if it has a hard time handling a needle one year, the next year it'll just get better and better at this. Like this this is all now happening really fast. This is like singledigit years. I think that's actually a much higher bar though than the the moment in history that's true lights out manufacturing needs to demonstrate like I here are all
[01:09:00] the steps of creating a figure 4. >> Yeah. >> Pick anyone at random. This robot with this hand can do that step. That's a moment in history that is actually sooner and easier than the I gave you an open-ended task that you've never even thought of before and you have to go and figure out how to do it. That's actually a much higher bar, but it's probably only, you know, six months later. The one thing I've learned is like these neural networks like love um >> love data. >> They love to learn. >> They they love to learn. They love the amount of variety of data that's coming in really well. Like I would said like um even on this logistics line that we're seeing behind us now it's running. The packages are all like you know like moving. They're like they're squishy. They're moving on the conveyor. Like the robots in a lot of cases hasn't seen these type of packages before. And um for whatever reason it's just it's just good at learning these representations in training. Um and uh I would have you know would have thought it never been the case. So we're seeing um you know we're going out now and doing large scale data collection as we're bringing like more diverse data. The robots getting smarter at the things that it's not even getting we're going out and
[01:10:01] like getting data like some guys like I saw like you know has a hose and like washing down his dog. Another one is like somebody's like you know painting a house and then we're like they were taking data like this and it's getting better at like um folding laundry. >> Yeah. >> And um it's crazy. It's like this data diversity. We've seen this with LLMs and pre-training. This data diversity um is just like so healthy for this like for this for the neural net to learn these representations across our world. So >> do you think I don't know if you can quantify this but how much do you think is general purpose physics you know physical world knowledge and how much is like okay if I want to train you to do this particular type of surgery then I need exact data from just that surgery. So right now when people automate these things they think entirely down these narrow swim lanes but we've never had the general purpose neural net before. Well, we've been on the narrow swim lane for two years. >> Like we that's what we do, right? Like we've been doing that for all we're like narrowly like I would say arguably probably the best in the world at post training these narrow use cases. >> Um and what you really want from a humanoid is like it can like adapt to things really quickly, things it's never
[01:11:00] seen before. And in the case of like say Bakyu or even surgery, you want like very minimal demonstrations the robot uh to be able to do these work. That means like like the prior here like the the robot coming into it with like say without any like say surgery training should be like should have like a decent understanding of like what it what like just with a prompt like what it should be doing and how to do it and then with some more specifics like I need you to do this task then go here and then do this and watch me like here this needs this this like needs to happen this way without that specifics um and context for a robot should then be able to do the work. So, it needs to have a good base understanding of how to do this stuff. Like a human was like a human walks into like uh getting prepared on how to what their first day of baky looks like. They're a really capable human. They know a little bit of like um like verbal prompting or like phys like show them how to do things to really understand um and to you know latch on to the concept. >> Uh so for us it's like it's really not about these narrow peaks anymore. It's really about like this broad level intelligence that we have like in LLM >> and then from there the robots can be fine-tuned to a spot where it's like the
[01:12:01] best in surgery or the best in you know like logistics or the best in you know some specialized task but from a from a it should be generalizable enough though to do most things >> like the surgery bot coming in should be able to go grab you a coffee really quick without any training you know what I mean but might need some specific training on that surgery >> coffee bot my favorite when we heard last uh you were just gearing up to building your manu facturing facility. Uh what's the status on that? >> Yeah, working on that now. We haven't announced anything publicly. Um but like you know the current Baku capacity um in my estimates like won't won't really uh hold here for the for the for for the future. So we have like you know we can do you know like maybe over 40,000 or so units now at full rate at a Baku next door a year across like four lines. Um and uh you know if we if we can crack this next few steps here I think we just see uh it feels like almost like a potential potentially unlimited demand. >> I I can imagine that. So do you think do
[01:13:01] you think figure four is going to be the first you know high mass production robot? >> Yeah. >> And then uh you know any projections for 2030 2035 20 240 what do you m I mean the demand is near infinite. I I buy that. Yeah. >> Uh so what what are the numbers that you dream about? [snorts] >> We we we see like we see I I arguably see a path where if we can um it's a really hard path, but if we can solve some of these like steps that we're on now that we could scale a million units like uh a million units a year out of a facility fairly quickly. But that's dependent on us being able to crack some of these chapter 3 problems we're doing now. And I remember when you and I had this conversation and I said, you know, do you do you agree with, you know, Elon's 2040 prediction of, you know, more robots on Earth than humans? >> Um, >> 2040 10 billion robots. Uh, I I think I
[01:14:02] think you were you agreed with that potential size of the marketplace. >> I'm fairly confident we're going to have a lot more robots than humans. >> Yeah. >> Yeah. You're like even if you size it like there's three billion 3.2 two billion humans in the commercial workforce. That'll probably scale to five to eight, maybe 10. Uh, and then you'll every human will want at least one. So, you're at like roughly eight there. So, you're roughly like order of like 20 billion humanoids. Um, but you're going to have them everywhere. They're going to be doing all kinds of things. Like they're going to be a dollar an hour as you said. Like you're like you're just going to have them uh waiting if there's any more work to do. Like one thing we're seeing now which is like going to help with this exponential is like the companies are all set up based on the human workforce. So even if we can go in and automate like say like a certain package line or certain part of manufacturing, we're still bottlenecked by the other human processes that run on certain shifts. So like the truck doesn't come in certain certain times. So it's not like we can keep working. >> There's like literally no packages to move anymore. So if you're working 24/7, so you're like left with like there will be this period where you're like you're you're coming in and doing real useful
[01:15:01] work and there will be another period where you come in and just fully automate end to end an entire company >> and work 24/7. Yeah. >> And that's just another step forward. And like we haven't even seen that in the world just getting I think you could ship I think the world could hold like 20 billion humanoids just in the first phase of that. But when you get to the point where like entire like logistics and supply chain companies um healthcare like data centers can be run fully 24/7 with lights out manufacturing you're like in a different you're like you're built you're basically building like entire economies that run 24/7. So I think it's a very different future we'll live in there. And then this is this goes to the point of you know the conversation you and I had with Elon about UHI when the robots are actually able to massively demonetize everything like building your home doing whatever you need at that dollar an hour price tag. It >> it creates abundance for everybody on the planet which is a beautiful thing. >> Yeah. Pretty deflationary. It's going to be great and it's going to raise wealth
[01:16:00] everywhere. We had a late night conversation and you said something that would really uh was a powerful idea and I want to come back to this. You know you said there's sort of an ethical moral issue where when AI is becoming or when we reach ASI um and ASI needs things to be done. You don't want ASI telling the humans to do these things. You want ASI telling the robots to do these things. Do you remember that conversation we had? Yeah. >> Do you want to expand on that? Because I thought that was a really important point. >> Yeah. I think like um I have like a like a view that robotics is super necessary to kind of like avoid like a dystopian future. I feel if we like there there's a question right now of you know I I feel like we have this question like are we at you know AGI ASI moment here it like the the digital AI is like so damn
[01:17:00] smart but it's like it's it's also like um not very magical. It feels like just like a like almost like a search engine replacement and it doesn't feel like hi or ASI in my mind. It doesn't feel like the Jetson right now. It just feels like it's fine. It can replace my Google searches. Um but like as this gets more and more intelligent like the digital systems that are living in a server in order to like like move things around uh to make money and help more in the economy these digital systems there's a future where they're like bossing around humans to do this work paying them money to go do the work for them. >> Yeah. Almost like enslaving humanity. It's um >> well, we have a long history of enslaving each other >> and it's like kind of an outcome >> like there there certainly like there's already like some companies popping up that are like like let like is like uh asking humans to go do work and paying them per hour to go like um like let agents go out and like get get work done. >> Um in my mind like if there is no like
[01:18:01] robotic solution here that it could lead them to a really um bad place. >> Yeah. So my view is that the as we build into AGI and ASI like the the humanoid should be doing this work and not humans and um I think it's a really important part here of the story that all needs to come together and it's coming together really fast. >> Yeah, I haven't heard anybody speak about that and I I think it's a really critical point you made that you know to avoid that dystopian future. Um I mean no human wants to be told by the machine what to do. Um, >> yeah, >> I think Peter is happening. There's like websites out you can go to now. >> Of course. Meet me puppets. >> Yeah. Just like it's just like the agent needs to hire somebody to go open a bank account or go do something for them. >> 100%. Or go do a task. Like there's like a like many like task rabbit type services are up now that are doing this. They're getting real traction. >> Yeah. There's another type of dystopia that I think we're on a path to avoid because of figure, which is I don't know if you ever saw any of those Russian cars back in the Soviet era, but
[01:19:01] everybody in the country has to drive the exact same color piece of Soviet car. [laughter] It's like one form fits everybody. And when you look at robotics in manufacturing today and you look at like a Honda Accord assembly line, it's like this robot is welding, it's welding, it's welding, it's welding. And what comes out the end of the factory is an identical Honda Accord for everybody. Different color. >> Yeah. >> Maybe a [clears throat] couple configuration differences. But now the LLMs are capable of mechanical design for the first time, which means somebody with an idea for a new device that never a physical device, you know, a gutter cleaner, a pool cleaner, like a a companion, whatever, you can vibe design it, but you can't make it. >> So unless [clears throat] all the parts are 3D printed or something like that, it'll never come into the world. Yeah. All of a sudden, Figure Robotics comes along capable of making itself, which means it's this far away from being able to make virtually anything. So then you're in a world with massively bigger variety of physical things
[01:20:01] >> because it's it's just a completely different supply chain. >> It's all, you know, anything that can be done can be done very cheaply, fully automated and >> 100%. >> Yeah. >> They're just going to work their butts off. All they're going to do is work. >> All they're going to do is work. We literally have had that robot since the like the logistics live stream running like just every day. It's just the right it's running right now. It just hasn't stopped. >> In instead of a c instead of a webcam at a in a fish fish tank, you've got webcams on robots doing packages there. >> Hey, can I ask you about chips? >> Yeah, of course. >> Chips. So when you look at you know the uh suppose there are 10 billion robots uh or more and suppose figure is a huge fraction of them maybe a dominant fraction the compute behind them wants to get smarter and smarter and smarter uh so Elon is looking at all the supply chain components he's like terraab this this is the one dependency I can eliminate all the other dependencies they've got a road map oh wait the chips so if we were talking nine months ago
[01:21:00] like we where the idea that you would have the capital to, you know, chips is about a $16 billion problem, but now you've got billions of dollars of cash. Then suddenly it's it's viable. Are you going to go down that road? Is it something you're talking about? >> Yeah. How far down the rabbit hole are you going to uh vertically integrate? Yeah, I suspect at some point we'll >> Yeah, I mean it's really a question is like can we control our destiny well enough with the current supply that will exist over the next five or 10 years. >> Um memory silicon like um and then you know even comes down to like the the cost per chip and power we're getting. Even the chips we have now run on robots, they're just still too costly and too high power. >> Mhm. Um, if you really want like a really affordable robot, you're going to have to really spend a lot of time in there. >> Yeah. >> Um, >> well, also it's, you know, you remember you were talking about the uh the actuator comes out of the box that's
[01:22:01] this big. >> Yeah. >> The people designing chips right now, >> they are sold out for decades. >> And so they're not going to start thinking about innovating in the direction that you need to go for robotics. They're just going to try and the cheapest, fastest fab they can put up and spew out what and you just have to live with We actually met with the like sea suite of the like one of the like largest memory OEMs in the world and um it's like a few months back and they're like we have we have we like and I was trying to talk them through like you know our future and roadmap and stuff and they're like we have no capacity for this. >> Yeah. >> All of our capacity is going to like phones and servers and everything else and like the phone like even like if a phone like doubles in memory takes the whole capacity of the whole market >> and the phone and >> they are phones are getting choked off. You think about like the amount of margin in a phone is insane and it can't even get the supply. >> Turn was complaining about that. His earnings call said memory is destroying us. >> Memor is bad. Yeah. We actually um we've done a few pretty cool like memory partnerships here uh to short memory over the next few years uh internally.
[01:23:00] So we have like um uh in the near term we feel very confident in our ability to scale up robots and then you know you're asking a lot of good questions about kind of the long term. So um yeah we'll [clears throat] see. I think like uh as you mentioned it's a huge capital investment the the technologies um the models itself and the way we're building the models and running them is really evolving fast. >> Yeah. >> So I'd be you know one concern I have is like building you know these these these chip this chip work takes like two three plus years. >> So you have to you have to look out three plus years minimum and say like okay what is the design going to look like? And um if you told me three years ago was the design of my inference compute here, it would just been completely wrong. >> Yeah. >> Um and then we have like some you know work we're doing now with Helix 3 which you guys just saw that I would say is like radically different type of way we're running now than we are used to. So I think um yeah it's like it's a little bit hard to like lock us into a
[01:24:00] certain amount of constrained power and cost and you know um chip architecture >> given how things are accelerating so quickly. Um, so, um, it's been nice to be able to like move to the right chip for us now, but like, um, but you know, if we look out over like a long enough period of time, we're going to need access to like unprecedented amount of silken. >> Well, there's a pro tip. You know, there's a there's a gap to if you if you think three to four years out, that's the earliest you could get your own chips. >> Yeah. >> But between here and there, uh, positron, etched, Grock are getting incredible headway on six nanometer or thereabout process nodes that are still available. But you have to race right now to book that capacity. >> But a 6 nmter node that's directed toward your use case is dramatically better than the two and three nanometer nodes that are, you know, that's >> all the iPhones now. Yeah. >> And all the Nvidia GPUs are built around the two and three node. They can't get any headway out of the sixes. >> Yeah. >> But that's the stepping stone that fills
[01:25:00] the gap. >> But it's, you know, those conversations are moving really, really fast. And those companies are, you know, it's funny in the in the physical world, um, entrepreneurship world. >> Yeah. >> There's there's only like two or maybe three of you. [laughter] >> So, I mean, it's a huge bar, but someone has to fill that void, otherwise we're all going to suffer for lack of robots for for some period. >> How old are you now, Brett? >> 40. >> 40. So, congrats. I I think >> to 40 made it. [laughter] >> Yeah. No, no, I think the congrats. >> I think I think No, the congrats is um you know, you've built serially increasing more valuable companies, which is which is extraordinary. Um I I think you were making the quote earlier, the the stat earlier that uh what was it? >> It's staggering. It's staggering. So all entrepreneurs all time only Zuck is ahead of you at age 40. But serial entrepreneurs >> Zuck god damn it. Well, so so [laughter] it gets better.
[01:26:00] >> But you're you're a nicer guy. >> Serial entrepreneurs, which is basically Jobs, Elon, and a handful of others, are far more likely to change the world in the future, which totally makes sense to me, having invested in hundreds of entrepreneurs over the years. If you can do it once or twice, >> you're going to do it three, four, five times. If you do it once, it might have been a lottery to it might have been just, you know, an outlier. Doesn't mean you can do it again. But you've already proven to do it like four times now, I guess. Um, so the forward trajectory of that profile is so much better statistically, not to put too much pressure on you, >> but then among serial entrepreneurs at age 40, >> you're 20x ahead of where Elon was or Jobs was in terms of resources that you have available. Let me just phrase it that way. And part of that is just sign of the times. There's just a lot more capital available today. Also, we were talking at breakfast this morning about if you I think you were destined to be a physical world guy, >> you know, it's just like it feels like it's in your kind of Illinois farm DNA, you know, machinery.
[01:27:00] >> I like touching hardware. >> Yeah, I do. And but if you if you were born 20 years earlier, >> you would have run headlong into the hardware is hard. Silicon Valley is never going to invest in it. You know, >> the moment is so perfect right now. So, you're just hitting like it and >> yeah, >> it's just going to be so fun to watch where this goes. Let's talk regulation. Um, so I mean so much conversation around regulation of AI, uh, but I haven't seen that in the robotic space. I've seen sort of the US government, uh, putting a regulatory block on importing Chinese robots. >> Yeah, the FCC. >> Um, which, um, I I get, but are you starting to have conversations or hearing from regulators about we got to, you know, regulate robots. They shouldn't be on the street. we need to be careful about in the homes. Where is that space right now and what are you expecting? What are you what are you sort of like gearing up for? >> I think the FCC ban on Chinese robots was like a little bit of a it was a shock. Um yeah, I think from my perspective on
[01:28:01] that, like I'd love to see like a really level playing ground to compete. >> Yeah. I mean, as you think about I'm going to come back to regulation in a second, but as you think about competition, is it is it Tesla and China? Is that the way you see it in the world right now? >> I personally haven't seen it much out of Tesla in two or three years. >> Um, I've seen some teleyoperated robots, but I haven't seen any. >> You haven't seen this, >> right? >> Have you? >> No, I haven't. >> Okay. Um, I've seen China do a lot of dancing and um, but we've talked about that before that the the neural network architecture is completely it's literally 180°ree different work. If you if you said Brett, we need you to out dance China. Um, [laughter] they do they do a damn good job. We'd have to pivot everybody's strategy. We don't have a like uh we we are dancing road map and go dance. Like it is um it is literally 180°ree opposite work. We need to get a bunch of humans wearing Hollywood suits dancing around. We need then work on replaying that on robot. Um
[01:29:02] and um you know it's like and we need to get out in the middle of the woods and go to around and leap over rivers and >> yeah Boston Dynamics was making the best uh you know robot. >> Uh >> I mean listen I think I think like >> I think I think if I answer your question more more directly. I think I think if I had to pick the like the the most um like the closest like uh competitor to figure I would say it's certainly China today. Um like the collective China there's just a lot of work happening there. They're moving quite fast. I happen to view they're moving a little bit different direction than we are. >> Um doesn't mean they >> and there are a lot of them. >> There's a lot of them. Yeah, there's a lot. >> There's like 150 humanoid robot companies. >> Yeah. just it's kind of all crap right now >> and um the hardware is not great. The policies they're running and models don't look that great, but the the amount of progress they're making year to year is really great. So I think um so I I would expect China to continue to do much better and it's a it's a group
[01:30:01] we intend to beat. >> Sure. Uh I mean China needs robots desperately. >> Yeah. Really bad. >> For for manufacturing and for elder care and all those. Yeah, of course. But are there any of the particular Chinese robot companies that you think are doing a better job than others? >> Not really. I think like we care about like we we care about something very different than we think China cares about. Um I China's almost in this race to like show like uh how capable their robots are at gymnastics. We care about like how do we ship a robot in a home? >> Yeah. >> To do like zero shot long horizon reasoning. >> And you're not seeing that out China. Like you can't even like like these groups like even like Unry, they don't they don't build good hardware when you when you like buy it. They're like super tiny. They come out of the box, they wave with a controller, you're like what else do I do with it? It's unclear. They have no hands and they have compute the size of like a USB stick. So you basically can't run any real like models on them. >> Uh seriously, like all the groups that are doing research put like like compute backpacks on them. >> Um and they're not the sensors aren't great. The hardware was like basically
[01:31:00] duct taped together. Uh it wasn't manufactured well. And um but like you know will they build better robots over time? Certainly. They're going to build great like I think over time probably great robots. Um we just got to go beat them. >> Mhm. >> At this. It's a real race here. So we're spending all of our time and energy thinking about where do we want to head long term and we're trying to move that left on the time curve as far hard as we possibly can. >> And every morning I come in thinking about how to move faster >> because there's always a way to move faster. And um you know sometimes >> for entrepreneurs out there hear this there is always a way to move faster and speed and agility is your >> you can literally always move faster which is like >> quite an unbelievable thing to have access to. >> Um you can even just you know working harder like work double the amount of hours you will probably make an order of like one and a half maybe even two times more speed more speed per year. So like um all of that really matters as we're compete against like the you know the global collective of China.
[01:32:00] >> Welcome to the health section of moonshots brought to you by Fountain Life. The discussion in this section is forformational purposes only and is not medical advice. Please consult a qualified health care professional about your individual needs. I'm a co-founder of Inclient at Fountain Life. You know AI is having an outsized impact on every aspect of our lives. How we teach our kids, how we run our companies. It also is having a huge impact on health, helping you prevent heart disease. One of the key things I'm here with Dr. Don Musalem, our chief medical officer at Fountain. Heart disease has been personal for you as well, hasn't it? >> It really has, Peter. My daughter was five. My husband died of sudden cardiac death. And so, this is a topic that is one that I am missiondriven to try to eradicate. Prevention first and early detection is absolutely critical. 50% of people die of heart attacks with no warning signs. >> No shortness of breath, no pain, no nothing. >> No silent killer. >> They just don't wake up in the morning. >> They don't wake up. And so, you know, AI, this is our mission to advance
[01:33:00] science to try to help to one day democratize wellness. We know at Fountain Life when we do this CT and geography with AI analytics, we are actually finding that 88% of people coming in have detectable coronary art disease. But Peter, what's more alarming to me is 23% of those individuals had soft plaque. This is the plaque that would not traditionally be seen on CT looking at calcium scores alone. And this is the plaque that we must intervene with with the multimodal testing we're doing, including diagnostic laboratory studies partnered with healthy lifestyle recommendations. >> So listen, make sure you understand what's going on inside your body genetically, metabolically, and cardiovascularly. you can know and it's your obligation to know. So check it out at fountainlife.com/peter to find out more and really make sure that you're the CEO of your own health. All right, back to the episode. You have the advantage of being a founder company versus a large industrial
[01:34:01] company trying to get into the space. You're willing to take bigger risks than almost anybody else. you're looking at, you know, you disruptive strategies and approaches and iterating rapidly. Um, >> well, and also the talent when you walk around the building here, the talent level, you'll never find it at a big industrial company. >> I mean, yeah, there's like a bunch of like small little things here. Like I think the the only way to really build like really capable general purpose robots is in the Bay Area. Like we invented almost like robotics, >> the Shenzhen of of Yeah, >> that's better than Shenzhen. We built we built robotics here. It was AI was literally invented here. uh you have silicon was built here like every like you know have batteries and motors and embedded software and thermals and uh AI uh deep learning uh all of infra all of it was built here like you're talking about the best place in the world to go do this >> um self-driving cars were like here before they were here in China AI was here before it was in China uh you know like the internet like these are all things that were like birthed here that
[01:35:00] um and the people like growing up here have worked in these areas and really understand the technologies well. And we've assembled like what I think is a top tier of them and they're like, you know, the whole team is a bunch of maniacs. They're just like they're really great at what they do. They love being here and missiondriven and we're all here to kind of beat everybody in the space. >> What was your journey to Silicon Valley from because you grew up in Illinois in farm country, right? And then you went to college in Florida. >> Yeah, I went to um went to college in Florida. I moved to New York City after that and then I started a software company called Vetery and um ended up selling that in 2017 >> and still in New York. >> Yeah, it was in New York. We it was basically an online like recruiting marketplace >> and um so we had like uh basically a two-sided marketplace to bring on employers and job seekers to help with recruiting and ended up selling to the world's largest recruiting company in 20 2017 for a little over 100 million. And um and then um as soon as I got the term sheet for this, I was like, I need to I had a few things I really wanted to do in the world and a few of them were in the hardware space. Um and then I ended
[01:36:00] up starting uh Archer Aviation uh like that next year. Um and then >> started it here or in New York? >> Started I started in New York with the intention to move >> and it was first I was going to move to LA because like all the like you know some of the best aerospace folks in the world. >> Yeah. And then and then I found out it wasn't really what I was working on at with Archer like building like electric like um you know like basically like electric like electric it's basically like a robots. It's like electric motors and battery systems um and control software and stuff. None of that was like stuff that aerospace was building. Um and it was all happening in the Bay Area. Uh you had some groups out here building like like EV tall aircraft like Kittyhawk and Airbus and a few others and you had like all this electrification community. >> Yeah. >> And robotics community. And so I moved here uh moved here then and I've been out here now ever since. Um and it's been it's been great. I mean this like the you know Archer is basically like a flying robot. >> Yeah. >> And um it's got all the same you know it's got electric motors, batteries, control systems, embedded software, sensors, like it's a flying robot and um
[01:37:01] so you know doing robotics out here is arguably the best place in the world. May maybe second to like >> I think number one has got to be the Bay Area. Maybe number two is maybe Shenzen. Um, those are the places that you'll see if if general bonus will be solved, you'll see it in one of these two places. >> I totally agree. It's amazing actually if you'd stayed in New York or moved to LA today, you know, with $9 trillion of new liquidity from Anthropics IPO and Nvidia's runup and everything. If you woke up tomorrow morning and said, you know, we were on this podcast yesterday, I think we need to start building our own ship fab. >> I need $20 billion. >> Yeah. >> You can do that here. >> Yeah. cannot even come close to doing it in New York or LA or >> this is the new capital >> in all those ways. >> You know, I was in New York for a long time. It's like I love I love New. I New York is one of the best places I ever lived. And um you know, and I was like, man, what's this deal with SF? Like I'm in New York City. They you know, probably like in software wise, probably the number two market in the US besides SF for like doing software engineering. Yeah. >> I was like, I don't really get this like SF thing. And you know, end up moving
[01:38:00] out here for Archer and mane do I get it now. It's like uh it's like this like it's [clears throat] it's like this most condensed place in the world of talent. >> Yeah. >> That um are all kind of in this right cultural phase of life to come over here and really really really punch it. >> It's actually the easiest place to import like if you go back to Chicago or or Boston huge talent >> but you can migrate anybody >> to here. if you're in like Austin or Chicago or whatever trying to do robotics or Boston, you you'll get some folks, you know, like one out of 50 or so, two out of 50 to move to move there from SF. But like in you get 50 out of 50. >> Yeah. >> And um it's and like in robotics like >> there's like an order of 12 different divisions of folks you need to build one robot. So it's like, okay, good luck getting those 12 people to come and those 12 groups to build out >> uh by hiring them all from SF. It's just not going to happen. >> Um it's just too difficult. So when you moved out, all everything you said was
[01:39:00] true when you moved out, except it wasn't the financial capital of the world yet. >> Now with all the new liquidity, it's all that plus it's the financial like Sand Hill Road and >> Speaking of that, you took Archer public. What 2.7 billion thereabouts? Uh Peggy Johnson just took Agility public. When's the right time for a company like Figure to go public? >> What do you What do you need in your mind having gone through a public offering? >> Well, I think Peggy went public cuz she >> Yeah, she needed capital. Yeah, of course. [laughter] No, I I understand that. I know I'll take you for years and I'm not I'm not going to speak about but >> like what are the criteria having learned what you did with with Archer? >> Yeah. >> I mean listen a high level Archer went public because we needed the money. >> Yeah. like we we were raising our like series B and um you know we just couldn't we needed hundreds of mill we needed like a few billion dollars to get through I needed a few billion dollar >> to get through like FA certification which is you know is very difficult to go do and um we were I was just not seeing it in the private markets and
[01:40:01] ended up going public and raising a little under a billion dollars. It was the right move at the time but we were really early. It was a really we just was just a really young company. So we were too young at that stage to be public. But we needed the cap I needed the capital. So I took the company public and I was a choice I would have done over again just given the constraints I had at the time but if if we had access to the private markets Archer would still probably to this day it would still be private. >> Um been healthiest for the business if we if we were even today. >> Um so uh on the figure side you know I think it's somewhat inevitable that we'll be public at some point like you know access to the right public capital. public trust as well for disclosures and such. >> Totally. Uh I think it's great for employees and liquidity and everything else for for folks and investors. Um but you know I think >> and full disclosure for the uh uh for the for the audience here I am a investor in figure. Yeah. >> Yeah. Um early investor too. So >> early investor very proud of it. >> Before we were cool [laughter] >> you know you know who I invested it was you it was specifically you and the team
[01:41:01] you're able to assemble. >> Yeah. And you know again just a note for you know for the entrepreneurs listening um the signal was so strong based upon what you had done as a serial entrepreneur but even stronger you were putting your own capital into it >> and the team you had assembled it was like that's a rockstar team >> team was team was lit >> still are uh still lit it's um >> I couldn't have recreated it again if I tried it was just uh right place right time and team just unbelievable and yeah it's been great Um, you know, listen, I think, uh, I I think one of the big focuses for us right now is we want to solve what we call general robotics. This is like a general purpose robot. >> And he saw a little glimpse of what that looks like today, a little future, and I think it'll get much much better. >> Um, that will really kick off like significant scale manufacturing, uh, even maybe a larger significance in inference and compute that we might need as we scale. Um, and just overall working capital needed to in capex needed to to to grow the business. Um so
[01:42:00] I think we really want to hit that milestone privately and um we feel like we're approaching this. Um it feels like >> man every new model we train we're like is it is it the one and um and then there's like you know bugs we want to work on and all this stuff. I think we have you know sitting in our servers today 10 times more data that we can train a new model on than what you guys saw today. >> Wow. I think it's literally >> we have 100x the compute coming on. >> Sorry I wish I make it bigger. I think it's maybe way more than like much more than 10. maybe a factor of like 15 or so more data sitting there. I mean we we bring data in. >> Is this is this data captured from humans wearing cameras? >> Yeah, it's it's it's from a variety of sources. Um but basically we need we need really rich information about like um physical world information that we can get from online and scrape online. >> You just can't get it from like the internet. And uh we need to know like like how like how would a human grab this cup and how are they orientated? What are the forces they feel? What are the hands doing? >> The feedback like the video is all over
[01:43:00] the internet but the feeling in that feedback. >> Yeah. Totally important. Can't get it. >> Yeah. I can't get it. So like so we we basically have like a an effort we call index here which is like large scale data collection effort. Um and we like but we get data in. What I'm trying to say is like we have to go like audit that data. We have to make sure it's the right distribution. We have to make sure we clean it uh annotate it get it ready for training. That takes time. Um and then when it's ready for training, we got to we got to be at the right place where we're ready for a new checkpoint for pre-training and go through. So not like we can just instantly train on it. We have like you know most of our compute um over the last like month uh more than half of our compute has been taken from like one giant pre-training run uh a large percentage of it. So like um >> but what's the price tag on a big big run for you? >> I mean we pay like like and we pay nine figures a year for compute right now. >> Yeah. >> Like low nine figures. So like um so like you know that was like maybe 50 to 60 70% of our comput pre-training. >> Okay. >> So like uh like you know a lot nine figures a year of like of is going into
[01:44:00] like just pre-training for Helix Helix 3. >> Um so quite a lot. And then if we had 10 times more compute we would train on 10 times more compute. >> Well next summer you have 100 times more. Are you going to >> next summer >> you going to do a billion dollar training run? Yeah, I think like Yeah, I think hopefully I I I expect like next summer we'll be doing billion dollar training runs on Helix. Um if all goes well, I mean like it's like you know everything's got to go well, but it looks like it looks like >> you'll have the data enough data. >> We'll have the data, we have the compute, we have the capital. Like getting these like larger like clusters set up is is hard. So we went out like six, seven months ago for compute. Um it's the current cluster we have now. I think it was like seven months ago. We had like four vendors that came back that had like compute like pretty decent like chips uh like uh new Blackwell series chips and we like we we beat the piss out of them, you know, in pricing just like got it down like you know the GPU per hour. It was like a great deal. We signed it like great. It's going out. We went out like I think it was like three months ago, four months ago, a couple months after that. We're like we we need like double our compute capacity. We were just like we just need more for Helix 3. >> We went out and we got like we went out
[01:45:01] and got like no bids back and then I was like we got to push harder. Went out again. We got like one bid back from like a big cluster. >> Yeah. >> And and then they beat the piss out of us [laughter] >> on price. Like it was just like buyers and sellers. >> I was like, "What just happened?" And we just couldn't get any more compute. >> And um you know, we met the Nvidia team and I was like, "Listen, it's just we're at like everything's at capacity. The whole world is just like there." And um >> we I I made this prediction earlier in the year. I was like, you know, in 2027, you're not going to be able to get comput anymore. >> And that happened in like September of 2026 for us. Yeah. September 2026, the world ran out of comput. You just can't get it. >> I'm so glad you're saying that. So many people are misinformed right now about, oh, you know, this AI thing is overblown. The bubble bubble going to be chips. >> We can't We can't get H100s. >> Yeah, >> we can't get the old chips. They're trading above par. >> We can't get those. We can't We can get I like somebody wrote me yesterday, I got computer. I'm like, great. I'm out. She's like, eight nodes. >> Yeah. >> I'm like, dude, eight nodes. Like I can't even stop and talk to you. Eight nodes. Like I need like [laughter] more than eight nodes, man. Uh it's not going to work at all. Um so yeah the whole
[01:46:02] world's out of compute and then compute is constrained by memory. The whole world's out of memory. It takes three or four years to build those fabs from scratch. Um so yeah there's like a really and Nvidia has been like one of the best place you can't you can't train on anything else besides this. So um it's really just interesting environment. >> You know what else is u you mentioned staying in the private markets longer but a lot of the investment is coming from Nvidia and other public companies. It's actually coming from the public markets through large public companies, but they never used to invest in startups or early stage companies or even like growing fast growing companies like bigger. Now they're just wide open for business. >> That's also a Silicon Valley kind of invention which allows much bigger things to be built highly efficiently in the private market. Yeah. >> Maybe a fab, you know, that would take it to the next level >> because you were saying a second ago like the new fab the new HBM memory fabs won't be online for like three years. Yeah. But if you look at how many are being built, >> it'll be over overbought before they even come online. >> Yeah. >> Like it's not even even if they come out
[01:47:01] on time, which they won't. >> Like none of these are forecasting humanoids. >> Yeah. >> They're not even forecasting. >> Not even like we literally met the CTO of one of the big groups and they're like, we don't even have any of these in our projections. >> Well, you know, everywhere you go too, whether it's software for mechanical design or whether it's space, whatever, everyone's like, "Yeah, they haven't forecasted my demand yet." >> Yeah. >> Like, okay, we have enough of these conversations. you stitch it all together, you're like, "Oh my god, we're like >> I mean, if you look at like the scaling loss for compute too, like um literally just the same amount of data and same parameter models being trained on more compute >> has like log linear relationship. You're seeing like a real ROI there." >> Yeah. >> So, um I I I would say like, man, it doesn't seem like this trend is anywhere close to stopping. >> Yeah. Um, so it just I think we're really early in the what's coming out of this like both digital systems like you know say we're doing a heart or even like figure stuff like it's just so magical. It's like it's like going to be way it's going to be 100 times bigger than the internet. It's like this stuff is going to be so important >> thousand times bigger. >> So I think we're just like I don't know
[01:48:00] my views were little we're still really early >> and uh it's not time to panic yet. So I keep calm and keep building. >> So I'm a Hark Pro user. Um, >> nice. >> And I would love to ask you, what's the relationship between Harkc and Figure in terms of where is is Hark going to be powering part of the Figure ecosystem? >> Yeah. Um, well, maybe I'll give you a quick overview of both just real quick and then what we're doing. I mean, Hark's on campus here um 100 feet away. >> It's nice. Um, you have you have your children next to each other. >> I just like it's nice. I don't have to travel and can always be around and help, you know, help wherever I can. >> Uh, yeah. like you know HAR's like basically like building trying to build personalized intelligence so almost like the digital version of what we're trying to do at Figure >> right [snorts] >> uh we do a lot with the computer use and speech and memory and vision a few other things and we're we're building our own like AI devices internally uh which I'm really excited about um we launched HARK Pro which is our kind of like our digital version of Harkc like our non-h hardware version um about a month we're
[01:49:00] about a month in now >> um and it's been honestly a crazy launch we we we were like you know we're kind we're really interested in hardware we think hardware is going to 10x this whole thing. And um we were we were I was even debating whether we want to put out anything before hardware. They like, "Okay, let's just put some stuff out. We think the models are pretty good that we're developing now." And I I actually use them all the time. It's all I've been using for a while now. And I was like, "Let's put them out and see how it goes." And it was just crazy. We um we we've now had over like 1.2 million signups. Uh we've done you know um we do you know probably like I think almost like five million browser sessions like since launch like um the usage [clears throat] is just like going through the roof. Um so much so that we had to like curb some countries like last week just because we're >> Yeah. Anyway, so um so that's good. We um so we're developing models and hardware. Uh right now we're very focused on on one of the areas is in real time speech to speech. So when you talk to robots on the way out they're using the HARKC voice realtime speech to speech model. And we're doing a lot of improvements there. I think speech I think real-time speech to speech is
[01:50:00] going to be really important part of agreed >> the UI layer for robotics and >> there's no third party components just totally internally developed real time speech to speech. >> Our speech pace model is our own model here. Um yeah >> there's a lot of people demoing speech to speech on top of something else >> 11. It's very similar to what we're talking about with the Chinese robots like the difference is night and day. You have a vertical integrated. >> Can I tell you something about speech to speech? Speech to speak today kind of sucks in the world. It's the the the hardware is not the right hardware to be using it on. We don't use it like as a primary thing. It's like a thing we'll put on every once in a while in the car, talk to. It's not really that humanlike. It doesn't do well with tool calling like going and get information. It's like it's not very good like low latency. You interrupt it sometimes doesn't hear you. >> It's just like not really quite great. >> Um so we're, you know, we at both Figure and Hark strongly believe like speech is like the right UI layer, like the the default layer into robotics and hard AI hardware. >> So we we're like all in on speech. Like we really want to make that work. So we're working on that at Harkc. Uh we have a collaboration with Figure on this. Uh Figure was like the first beta user of Hark. So we've been using Hark
[01:51:00] now for five months. Uh the so we use like uh we don't like force people to use it, but if they want to use this uh they can use Hark internally. The numbers are just like like basically through the roof internally here which has been which has been great. Um we can get like help with productivity around the around the company. Um and then uh and then we're also working on um we're really big at HARK on like larger scale multimodal pre-training. We're like really we think it's really important for the space. >> Uh obviously figure has been working on like you know how do we generate like action predictions out of the robot. Um anyway we're doing a lot of there are there are some areas of class collaboration we're doing. >> I you know I have a we have a running uh bet inside the moonshot made team here especially with Alex. So uh Elon very famously you know started XAI then SpaceX bought it and then you know X and so forth consolidating the companies. Uh so the bet is that Figure will buy Hark eventually.
[01:52:00] >> I mean I would probably make a different like yeah I would say like this is just my guess so don't hold me to it but I bet Hark is bigger than Figure in sub six months >> in sub six months. >> Maybe sub 90 days. >> Are you serious? from a revenue standpoint, from a >> I mean it's already bigger than figure from revenue standpoint. We've been growing like crazy. Uh and um >> like the hardc numbers are just through the roof. We have some like very big stuff coming on hardware um and some very big commercial agreements. So um yeah, I would say like uh like I you know given that it's digital, we can scale like very different like we're in different slope than scaling robotics. Um but I think um listen, this is my guess. uh we don't have anything going on right now. Um but uh I think you know I I do see a path where yeah certainly I think I think I do see a path where next month there could be a chance that uh that is bigger than figure and then at some point longer term maybe figure gets bigger again. I don't know. Um
[01:53:02] >> a tr a trillion here a trillion there you know it's >> I do I mean the reason I wanted to work on hark is like I was sitting here watching like doing this human like doing the human robotics thing. We're like, we're we're an AI business. We want to get like get build the AI brain that fits in a hard the hardware is just the distribution for the AI. It's like the it's like the the vector it's going to head out into the world in. And I was like playing around with like chatbt and the rest of the AI stuff. And they were just so stupid. And I was like I was like I'm just like talking to this like chatbot that was like a a search engine replacement from Google. >> And I was like this isn't what I really want. I want something that like can be like with me everywhere I go that like knows me really well. has access to my accounts and can like do things for me like pay the property tax bill, remind me of something before I forget it, help book my travel for me, like uh do something more complicated that might be long horizon like for me with like uh like with my permission and like almost like Jarvis would have and um nobody was doing it and um I you know open led my series B at figure and I you know got the vibe of being close to those folks that like they had no real interest in making this a a huge priority internally
[01:54:01] >> and I also feel that on the consumer humor side like interacting with you know say computers and phones with AI is this like really broken experience today. So I sat here and said okay um I'm kind of in my prime as a builder. Uh I I'm trying to understand AI more than more. I'm like fascinated by this digital version of AI. It's like you know even going to a computer and using it. It's like it's a little mini humanoid like moving a mouse and you know keyboard and trying to get everything to work. I'm just fascinated by it like personally. So, I spent a lot of time in just like researching it, understanding how this deep how basically deep learning works in this space. And um so I got to a point where I'm like I'm going to be like I'm just going to sit here and watch all this stuff or I can go build and come back from come off the sidelines. And I decided like it'd make me a better founder and CEO if I came in and and helped build something I think was really missing for the world. And I think it's going be really special. There there's going to be a point in time, you know, like all those little charts you watch uh charts you see over
[01:55:00] time like over a span of like your whole lifetime like how often you see your kids and your parents and everything like kind of goes up then it goes down like it goes down down down until you're like you know the rest of your life. I think there'll be a chart here of like both in robotics but certainly for like even for stuff like HARK where this line will be going up all the way to the end >> and um it'll be such an important you have the smartest person in the world with you everywhere you go that's that's that's that has the same interest as you do. Yeah. >> And and helping you and >> trustworthy though. That's the key is that that's actually common between heart and figure is that you have to trust this entity with your most personal information for it to be effective. >> We tell the team every day. It's like it's like you have the person's overtime diary and we need a new level of security, a new level of privacy that's never been done before. Like a thing that we do, you know, at heart is we try to protect that at like every angle and it makes it really hard on us like it's like we can't we can't do what other companies do and as as it relates to, you know, advertising other stuff for their products because because we're trying to like protect the interest of the consumer as much as possible. So um
[01:56:01] anyway, I I think um >> well congratulations on a meteoric growth. I mean it's been when did you announce Hark? It's not what >> we're like over a year old now and we launched Pro a a month ago. So, we're like a month, you know, we're kind of like a month out with pro. Um, it it's been great, man. I like I'm I'm learning every day. There's like we're like in the early early chapters of this AI like um almost like when like 1999 look back at internet, it just wasn't the right right. It wasn't quite the right internet yet, but it was like it was all getting seated. It's all getting seated now. It's I I thought for a minute uh last like six months, we were real late at Hark and then um I've now realized we're pretty early. Um, and so we have a lot of, you know, we're launching, I think, four or five pretty big product announcements this week. We launched one this morning. So, there'll be a lot, you know, and then we think about as we like think about getting into hardware. This thing is we have hardware now at HARK in the lab and dude is it crazy? >> So, peripherals, >> when you say hardware, uh, is there
[01:57:01] going to be, you know, sort of HR glasses or >> can I give you my like thesis here for devices? Um so my my view is that um there's like there's like two different like areas for AI devices. One is like there's this like centerpiece for AI devices which I I think you know for today for consumer electronics that sits the phone and computer like a call like a billion units a year devices. So it's right there in the middle and then those are the main main show main platforms and then around there you have like accessory devices. This could be like AirPods or like you know watches or glasses or whatever else. I think AirPods is like 3% of Apple revenues. >> So like Apple's not sitting there getting super excited about the AirPods, but it's a important platform play. They're super interested in like the computer and phone. That's the main thing. >> What what what will happen is there will be AI devices that will kill those in the middle. >> They will do a billion units a year devices that will be way better than your phone and computer today. >> And then there'll be just as similar as it is now like accessory devices on the
[01:58:00] outside. um those devices will be singledigit revenue percentage and they will be they will only work with the main show >> for AI devices we've only seen the accessories today like the meta glasses which are horrible by the way um and other type of like things on the periphery that are not the main show that don't work well today because they interface with with like things that are in your main platform like your phone your iPhone fundamentally will never interface well with AI device like if you try the meta glasses it has to have your app open at meta on your phone or it can't get a network. >> Even with that, it's a horrible experience. It just like this the products maybe one of the worst products I' ever tried. >> And so, >> how do you really feel about this? >> I mean, like, let's just get real. Like, you're like, I don't I can't I feel like and I watch the spectacles like it's just I feel like I'm taking crazy pills. [laughter] >> Like, nobody's going to wear these. >> They keep getting bigger and uglier somehow in space. Um, but like, you know, so for us, you have to go build it well. >> So, what's a form what's a form factor look like for the ideal AI peripheral device? >> We're not going to tell you. Like we
[01:59:00] gota we gota working on that. >> We wouldn't tell anybody. >> Like we're gonna tell you like I I think my analogy my best analogy >> but you have a vision for what it becomes. >> Oh we have the hardware in the lab right now. >> Nice. >> Oh wow. Really? >> Yeah we have it right now. It's in the lab >> right next door. >> Right next door. I'm using it every day. >> Huh. >> Uh it is. >> Does it have windows? Can I look? >> Unbelievable. It's unbel it's unbelievable guys. Like I uh it it is just um this stuff is going to be really incredible. Um it >> are you coordinating with Johnny IV and OpenAI? You guys like talk to each other? >> No, we're not. We're not coordinating like not [laughter] at Not at all. We're not talking um at all. I mean, we like we fired them like two years ago. Okay. >> Um after the series B work um and it just wasn't a good fit for us. And >> how so? Is there a back story there? >> Um I mean they they so they came in I met Sam, you know, and they they came in and led our series B at Figure >> um and they co-led it with they co-led it with Microsoft and got they ended up joining the board and we did a lot of work together internally um where we we worked with them through for about a year or so. Um, and it was like, you know, the goal was like to build like
[02:00:00] journal intelligence in a in a in a robot. And um, you know, I was in every weekly meeting. We had a whole team dedicated to it. They had a team dedicated to it. And we're like, great, let's get the figure every day and let's rock. They're like, "Yep." And then they never showed up at the office. They were like, "It's too far from SF." And, you know, it's it's just like it's just loud over there and it's like, you know, we're like, you know, we're like we got grease on us and it's like loud and we're working on robots and they just would never show up. We couldn't get them there. I was like it's like we couldn't I was like okay let's do a big event and get you guys. We just couldn't get into the office and then and then like every week I was just watching the progress that we were making and then we were sharing with them and then they were making and sharing with us and it was just completely lopsided. We were making 10 to one 100 to one progress for us they were doing and we were like sharing all the stuff we were doing >> uh like >> meanwhile they're sharing with Microsoft. the chain of >> and then you know like and then it became pretty obvious they they really wanted kind of after I think after seeing the results from us really wanted to get into robotics again and then it it was just like in you know it was just something I just >> when did Johnny IV come on board at open
[02:01:01] AI during this tonight was that after >> I don't think Johnny's working on the humanoid at open AI nobody knows >> I think he's working I think he's my guess is he's working on the like AI devices work >> um I don't know if he's working >> the iPhone killer but now they got afraid of saying that but it's it's some device that has AI embedded in it that's so compelling but your your point about there being a centerpiece your personal information isn't going to be everywhere you're only going to trust a limited number of center centerpieces and so I think they want that device to be the centerpiece not the iPhone you obviously have the same vision with Hark >> so I can see how those are >> listen it's certainly not going to be the iPhone like it's uh the iPhone's going to go away like >> all right we're calling it here the decline of Apple >> I mean like I like let's just call and you like listen if you take the extreme case in 10 years we're not using a traditional phone. >> Yeah. >> You know what I mean? Like I just like uh like so we know that like >> it's like it's like Turnis really has a challenge right now because they've been they've been gliding >> forever on the iPhone and it's going to
[02:02:02] get challenged from a dozen different places. Yeah, >> man. I've like I've like had this thing where I've like w I've like thought I'd wake up at some point every day and see you know Tim Cook or whatever John get on some like emergency like you know like PR event and basically say like hey we like you know we bought a hundred billion dollars of GPUs we're bringing the top team on the world we're going to go at pre-training hard in the meantime we're going to use other models and this is going to be so core to the experience and interface with consumers long term that we really need to own it understand it and embed it and Um it's it's been crazy to see um nothing. Yeah. >> And it's now crossed the point probably about a year ago to a point where it's like really too hard to turn around. >> You think >> it's for sure it's like they've lost so much momentum. >> Yeah. >> You want to co-design the hardware and the models together. They're really um you really want to think about um these AI devices as just a pure interface. It could be BCI in the brain. It could be language devices that we're working through now. like this is just the
[02:03:00] interface to this raw intelligence in their side that's going to be humanlike just for you >> and um they need to be co-designed together >> and it's not something you just like flop in >> totally interested in BCI. >> Yeah, I think like I'm very interested in BCI. Uh I think it's one of the most I think it's a very important vector to be on for kind of like AI interfacing long term. Uh I just >> I'm invested in four of the companies right now and I've just seen some I just had some meetings yesterday. Uh it's blowing my mind. We're on the precipice of being able to actually uh get to a mouse conneto. >> Yeah. >> And then a human conneto. >> And and then the world changes. >> That's sick. We should debrief at some point. >> I would love to make some introductions. There's an amazing company. >> Yeah. I think we're going through this phase of like kind of beyond where we're at now. It's like like language device phase. I think there's a there's a there's a phase kind of after that of like of BCI and alignment um into the brain. I think it's going to be really
[02:04:00] important. Um I expect HARK to be working through that whole process all the way through. Um so um I think we're like uh but certainly right now there's like a there's like an upgrade cycle coming >> for traditional computers and phones and what I can tell you by just using some of the products here internally is it's going to be phenomenal. It's going to be a much better experience. >> Yeah. >> Um >> nice. Yeah. We haven't changed in 20 years. >> It's literally 20 years and we now are just plopping AI into it and it's like it's kind of like the interfaces are just really reactionary like um you have to kind of control the phone yourself. But we've >> man we've built this thing that can think now and it can use computers and it can speak to you and it can see and it can build persistent almost perfect memory. Um and you know you you want to be able to interface with that real time and um even you know and and have it with you everywhere you go. And uh we don't have anything near like that in the world today, but we will shortly. Like it's awesome. >> Yeah. Well, after AI is the interface. You basically use the phone as a camera, but the cameras are like three bucks. >> It's like like, okay, why did I pay 1,500 bucks? I got a $3 camera.
[02:05:01] >> I mean, phones are incredible. It's just like there is there is there is just like a next leg up coming. It's going to be great. >> Yeah, it's about time. Let's wrap with the conversation on on regulation because I am curious what you're thinking here. We've got Bernie Sanders and at all saying, you know, we got to stop, slow down, prevent RSI. And I I have to imagine, I don't know if you're in those conversations at all, but I have to imagine there's going to be a similar kind of like uh slow down. We don't need robots on our streets. It's not safe for the citizenry. How do you think about that? Are you having those conversations? And um is there any you know is there is there real concern that the public should have in this regard? >> I can speak for our plans. Our plans are to launch a really high quality product that works really well into the market >> and is safe >> and is safe. Like we don't want to be out there like you know teley hopping
[02:06:01] the robot around and doing unsafe work um shipping too early. So, we're, you know, we're making incremental steps now. We we we shipped first commercially last year. We're shipping even more this year. >> Have you seen any inkling from regulators um to start speaking about humanoid robotics the same way they're speaking about super intelligence? >> We we we really haven't seen that like public facing. We've um even rumblings of this, we've seen like there's been like, you know, the only thing we've we've been aware of last like 90 days is the FCC ban on humano on Chinese humanoids. Um, beyond that, we haven't seen anything, you know, come across our desk. I think, um, we're, you know, to also a little bit personal to me, like I don't fly to DC every week and try to get my nose in this. Like, I really want to be here. I'm here. Good for you. >> I'm here. I'm here trying to build and I want to build awesome robots regardless. Like, there's going to be robots coming. They're going to be great for the world. U, I think the same with AI. It's going to be great. If like humanity woke up one day and just didn't want AI, we can just shut down the internet. It's, you know, we can just close it. And, uh, we but we don't want
[02:07:00] that. it's going to be like really helpful for the whole entire world. Um, and you know, there will be like whatever Bernie Sanders the world or whatever uh here, but like for us, we're just like really focused on the mission of trying to build and ship your robots >> and we're just like we're here every day trying to go do that. >> I got a question just to inspire all the entrepreneurs out there. So, you know, I grew up a big Steve Jobs fan, too, and I think a lot about how much more he could have done in his life. But in that era, you could only do one thing at a time, and nobody had super voting control and all that. And then >> then when Google came along and then Zuckerberg came along, it switched to this era where the founders control the companies. Yeah. >> And they could be much more aggressive with the business plan. And that works really well here in the valley. Yeah. But then Elon came along and reinvented it again where you can do multiple things at the same time. So this is working out incredibly well for figure and HARK. If you just said Hark nine months from today could be bigger than figure it which is mind-blowing. >> I think you said you said three months from today. >> Three [laughter] months from today.
[02:08:00] >> I mean like Hark's just going like this guys. Like I just like it's it is going really well. Like so um we we'll see. >> You've also got this other thing about school safety you know that that is Yeah. which is, you know, not going to be bigger than the other two companies, but it's important for the world. Yeah. >> Happening concurrent. Like the power of an entrepreneur today is so much bigger than it was in the Steve Jobs era, let alone the Zuckerberg era. So, I think it's really inspiring. Do Do you think a 20-year-old entrepreneur today can walk in your footsteps, or is AI going to make it all too late and it's it's, you know, the ship has sailed? >> Yeah, I mean, I think one thing I'll say is I'm like completely dying. like three kids, like three companies. Like I'm just like happy. You look >> I'm just I'm just like I've been here every night for whole year. I'm just like um man has it been hard. Um but it's like you know it's the best time of my life I never want to do again. You know it's like uh [laughter] it's been it's been going great. Uh yeah the listen I like the future generation is going to be great. Everything's going to be AI native. Uh you'll be able to
[02:09:00] build any type of hardware just through um >> prompt product. >> Exactly. just through talking and robots like um it's going to be so cool like it it really where the AI is really missing is like it has like no imagination today >> and we're going to be bound by this imagination and it's like this is what entrepreneurs do best right like they always have like a little notepad I still say have a notepad of like all the business ideas I I can think of and just I still keep them to to this day and it's just like it's it's a you know they're great at idea generation great about thinking about the future like this is going to be an an unbounded opportunity uh for every like new entrepreneur like behind me um growing up and it just maybe one of the it'll be the it'll mark the best time in the world for entrepreneurs in history and now you have an ecosystem here like today that like for entrepreneurship like >> and when I was graduating school like it wasn't like it wasn't conceivable you'd start as an entrepreneur like it was just like you don't do nobody does that like >> uh so I had a job for a few years before I started it I wish I started even sooner uh starting companies so um but
[02:10:00] like you know now you can go do that you can get capital for it there's incubators there's people want to work at startups. It's fun. Like you have this whole like community built around it too, which is just like going to propel uh these future entrepreneurs further. >> Yeah. >> I like to joke uh we've all been promised a three and four day work week. We've invented a 10day work week. >> I mean, I like working, man. Let's just call it what it is. >> This is the most exciting time ever [clears throat] live. >> I like grinding out and working this. Like I just we we I think everybody here with me likes it, too. We just like we have a bunch of groups that like like to work hard and we've like we've made it this like dirty word out here in the valley and so you know like a lot of folks come in here and be like you guys are like the most unsilicon valley company I've ever seen like we work hard. >> We come in on the weekends like we work nights like we just we like it and um there's not enough of that anymore you know just like almost like we like interview people like oh what's your like day off strategy and like work from home like we don't have any of that you come in here every day and you like work with us and we're we're gonna work with you. Like you know my Baku guys can't work from home. >> They gota be on the line every day. >> Yeah. >> And uh same with us. Like we got to get
[02:11:01] online too. So like um >> anyway it's just um >> what you're building is so so much cooler than virtually any day job, you know? I mean it's got to be even if you're working twice as hard, it's got to be so much rewarding at the more rewarding at the end of the day to look at what you physically built. That's just also I didn't anticipate how much the office environment is elevated by the walking around >> so much better. >> It's it's bizarre. It's like it just adds energy to the building somehow. >> So, a year from now, we'd come back and do our uh our >> figure four recap. >> You come back, figure four. Figure 4 gets you all set up, checks you in, gets you water, coffee, walks you around, gives you a tour. Uh and then I'll see you for the interview. >> Yeah, we'll include figure four in the interview, too. We just like honestly figure four can speak on my behalf. [laughter] >> That's so fun. >> It's like, you know what I mean? >> Buddy, thank you so much. >> Thanks for having me. >> Yeah. >> And and now when we cut cameras, we'll go see your secret device in the figure 4. >> Perfect. >> Good deal.