Brett Adcock, welcome to the show.
I've been looking forward to this for a long time. The robotics guy.
Let me give you an intro here real quick before we get started. Brett Adcock, a serial entrepreneur and founder and CEO of Figure AI, building general-purpose humanoid robots for labor automation. Founded Vettery, an AI-driven talent marketplace which was acquired for approximately $110 million. Co-founder of Archer Aviation developing electric vertical takeoff and landing (eVTOL) aircraft. Founded Cover, an AI security company using NASA Jet Propulsion Laboratory technology to detect concealed weapons in K through 12 schools. That's amazing. In late 2025, you launched HARK, a new AI lab self-funded with $100 million to build what you call human-centric AI. You've raised billions in venture capital and Time named you one of the 100 most influential people in AI in 2024. Married and a father of three children. And before we get too far into it, we always start off with a gift.
He didn't give you any tips on that, did he?
All right, I got to hit one.
You can leave this guy here if you want to.
This guy's staying. That is the coolest thing I've ever seen as far as giving somebody a gift on the shelf. That was awesome. That was awesome.
Oh, something you gifts keep here, put on the shelves.
Very cool. Well, Brett, we got a lot to talk about here. So, man, how many companies are you running now?
Yeah, man. I'm not sleeping. I got too many.
Too many. Just like kids and work and just like that thing is amazing. Never sleeping anymore.
I'll bet. I'll bet. What do you think of the robot?
I think it's incredible. I can't wait to talk more about it.
Yeah. So, a couple things, just one more thing to knock out here before we get into it. I got a Patreon account. It's a subscription account and it's quite the community and they're honestly the reason that I get to sit here with you today. So, they get the opportunity to ask every single guest a question. This is from Steven Casey. In today's marketplace, we find that AI platforms can sometimes invent answers rather than admitting to a lack of information. Combining this in the physical realm of robotic action seems to multiply the downside effects exponentially. What safeguards are in place that we can put our trust in to prevent the potential for downstream harm to humans as a result of bad programming or computing errors?
Yeah. We don't want the Terminator popping out here when we do this work, right? I mean, I think like we were chatting about this outside. I think one thing to say, like four years ago when we started the company, there was no path for humanoid robots to make it into people's homes in the next 10 years. There was no good story. You had big hydraulic humanoids out there. They were all hand-coded to do certain tasks. What you really need is a cheaper electric humanoid that you can basically use neural nets, use an AI-first strategy with. None of that existed. I think we're thankful now looking back that we somehow pulled 10 years of the future forward. We have electric humanoids that are reasonably priced that can do useful human work with neural nets. And it's just an incredible place to be in, getting those questions which is how do we make this work now at scale in a safe way, because that's the spot we want to be in, not trying to make this work for 20 years.
So I think it's a very tough problem. We have to get the product cheap enough. We have to make enough of them. We have to make the performance work in very complicated things like walk around a house and do dishes, laundry, very complex things. Small kids can't do this, it takes adults to do this level of work. And we need that all done in a mechanical system that doesn't have any humans around for maybe most of this, that does it autonomously and doesn't make any mistakes. And then like your fan mentioned, we have to do it safely over time. It's just an incredibly complex problem. I think for us, we have a safety strategy both intrinsically, we want the robot hardware and the robots around humans to just be safe all times, and separately we have a bunch of semantic safety and other things that we need to put in place to make the robot work safe in the environment. Like you have a candle at home, you don't want the robot to accidentally knock it over. It's an intelligence thing in a lot of ways. Or there's a boiling pot of water, making sure we're very safe around it. And then there's the intrinsic safety of making sure this mechanical thing in your house is safe around everybody around it. I think the direct answer is there's still a lot of wood to chop of getting this thing to a point where we trust it to be autonomous next to my kids all day long in my house. That's kind of where we're at.
In your house. We've had many robots now throughout my house in testing for the last year or so. And I've had them kind of near my kids in some aspects, but we're always monitoring it.
Man, they like it's just kind of normal for them now.
Do they try to talk to it?
Yeah. They want to go like jump on it and touch it and do kids things, you know what I mean? Like they want to go touch it and talk to it and be around it. And we're still not at that stage yet where I feel comfortable enough to let loose and say here, you know, here's a robot or my kids are there and I feel okay. And we're not there yet. I think we will be in the next several years.
What's the longest they've been around any one particular robot?
We've had a robot in my house for maybe a couple months. Doing work kind of on and off, daily, sometimes every other day. And the kids were kind of at school or sometimes at home so they weren't always around whenever the robots were running but a lot of times. And that was just our home robot.
Do they get attached to them?
They name different names for the robot and yeah they love it. And it's actually a question where in the office of like you have a robot in the home and it's got some character to it and a little wear and tear. Do you want to keep that robot or do you want a new one?
That's what I'm wondering. What's the emotional attachment? I think kids are like the perfect... my kids wanted it. They wanted it there. We're not getting rid of this guy.
Yeah. It's got a little banged up a little bit here and there and has a tear here and they just loved it.
That is wild, man. Yeah. That is wild. Honestly, in our lifetime, we will be fortunate enough for every human to I think have a humanoid almost like a phone and car.
Wow. Yeah. We were talking, I mean just some of the stuff that you just mentioned, the complexity of the problem that you're solving here. I mean all these little problems that are even like knocking over a boiling pot of water. I never would have like just thinking about something that happens every day and then you think of all the things that happen every day in just a regular household and...
It's like problem city, man. It's like a fun house of problems. There's just problems everywhere. It's like hardware problems, AI problems, problems scaling and commercializing and getting the system reliable, manufacturing problems. It's like we have a problem fun house if you want to come by campus here.
I'll bet. I'll bet you do. I'll bet you do. Well, some people say AI is in an economic bubble. And as of this recording, Polymarket says there's going to be an 18% chance that the AI bubble will burst by December 31st, 2026. What do you think about that? Is AI in a bubble?
Absolutely not. I think you'll see some of the most transformative events in technology happen over the next 36 months we've ever seen in our lifetime. I don't feel like we're in a bubble here. I feel like we're scratching the surface. I'm watching AI in a human body do human work. It's early. We don't have, you know, at some point here this year, we'll have thousands of robots. We have hundreds now. But we need millions of robots to make an impact. That's just going to take some time and it's going to be crazy cool. So we're just at the start line of that happening, which is how do we get AI out into the physical world at scale? That'll for sure work and it'll go really far in our lifetimes. And then separately, we have AI now that can use computers like humans and can think. I was showing you a little bit of that here before the show. And that will manifest in a point where both in the physical and digital world you basically have these little mini humans that can do human-like work and they can think and use computers and use machines. And I mean that's going to lead to such a productivity, like we measure GDP per capita per human, but if you're able to make as many synthetic humans, millions, billions, tens of billions of synthetic humans in the case of the digital world maybe trillions. That'll lead to the greatest increase in productivity we've ever seen in our lifetime and ultimately reduce goods and service prices to unprecedented levels, like a true age of abundance.
Wow. Wow. What do you, I mean I'm just curious, what do you think what will humans be doing?
I mean I hope I don't have to like, I woke up today I was like unload the dishwasher, getting my kids breakfast, just like busy work that I, my kids are sitting there. I'm like doing work, you know what I mean? Like I wish I wasn't doing that stuff. And then all throughout my day I'm like trying to call the car service and then trying to get on my flight and coming here and it's like ordering lunch, like all the stuff I'm doing all day. And I don't want to do any of that. I want to be fully free. All that burden. And I just want to be clear-headed. And I want my AI to run a little Brett Adcock operating system and run my life and all these things I have in my head about what to order and pay his tax bill and do this meeting and I have to go back and do an engineering stand-up. I want all that stuff to be in my operating system and like a human in a box.
So you're basically saying the way this is going to turn out is your brain, I'm going to butcher this. You're basically exporting your brain and all the tasks that are going on in your brain. You're disseminating it to robots and...
Delegate all this out to robots.
That's amazing. That's...
Like we'll do that in like 24 months. Like we'll have all the stuff so good that you won't go order food anymore, book stuff, do a lot of work behind a computer, physical stuff in the world of doing laundry and dishes and just the leg work. Yeah. I don't know. Does anybody want to do that? Like I don't want to do it.
Yeah. Yeah. God. So, like you clear all that from my life. Like I got to spend time with my kids, enjoy life, be clear-headed. Do stuff I really love. Like I love working, but I don't like doing all this busy work. It's just manual labor I'm doing behind computers or in the physical world and I want to delegate that out to my AI to do and fully automate out.
That's, I never, I don't know why I've never thought about that. I've never thought about it like I've always looked at it as fear. I've always been like, 'Oh they're going to take everything over.'
It's a compression algorithm. Like we're basically running a large scale compression. So I think the way I look at it now is we basically have built synthetic human intelligence that can use computers and machines. So I'm going to delegate out all this busy work on both my digital life and physical life to robots and they'll just do all of it. But it's good. I mean, we have AI systems now in our lab at HARK that can use computers like a human can. It can talk to you. I just made a phone call to ours right before we started and talked about my schedule and how to ask for things and I ask it for things and have it do things.
Yeah. He had a chicken salad ordered to deliver to your office.
Yeah. Exactly. Exactly. With nothing besides a single 'hey make this order' and you can spin up computers to do that virtually and then physically I'll have all this work done by robotics both in the commercial workforce in the billions, like manufacturing and healthcare and construction, and every human at some point will have a humanoid just to do all that busy work for you and not only that but something to come home to that you can talk to that will know you.
Wild. Yeah, that's like the...
Yeah, it's going to happen now, which is really going to be fun.
Yeah. Yeah.
I'm excited to introduce to you the newest member of my family. We call him Stanley. We got Stanley this past Christmas, and pretty quickly my focus became making sure he was safe while still giving him the freedom to actually be a dog. I went looking for the top rated GPS fence, the number one, and that's how I found SpotOn. Stanley wears their Nova collar. With SpotOn, I set up a GPS fence for Stanley right on my phone. No physical fence, no leash. I just walk my property line or draw it on the map, and that's the boundary the collar recognizes. I can create multiple fences, save them, and adjust them whenever I need to. So whether we're at home or traveling, Stanley always knows where his boundaries are. Another thing that stood out to me is these collars are designed right here in the USA and assembled in New Hampshire by a team that's been working with high precision GPS technology for years. And you can tell a lot of attention went into making this thing reliable. The Nova Collar uses a dual-band GPS system connected to more than 150 satellites along with an antenna that's over five times larger than typical GPS fence collars that keeps the boundary accurate even around trees, terrain, and changing conditions. SpotOn's True Location technology has been independently tested and delivers 99.3% containment, which matters when you're trusting something with your dog's safety. It's incredibly durable, and on top of that, I can check his location in real time, send voice commands directly to the collar, and track his activity through the day. I use SpotOn so Stanley gets the freedom to run and explore and I get the peace of mind knowing he's safe. Let your dog roam with SpotOn. Go to spot-onfence.com/srs and use code SRS for $50 off the Nova Collar. That's spoton.com/srs and use code SRS for $50 off.
Well, I would like to do a little bit of a life story on you if that's, does that sound good to you?
Yeah. Like a small town, like 700 people.
Called Milmine, Illinois.
That's even smaller than where I grew up.
Yeah. Where'd you grow up?
I grew up in small town Chillicothe, Missouri.
About 8,000 people at the time.
Yeah. We didn't have anything. 700 people, man.
Yeah. So, like, you know, kids, sports, computers, got into computers really early. Did a bunch of sports. I grew up on a farm, so it was corn and soybeans. My family was third generation of this.
Yeah. So third generation of agriculture farming.
And then we switch over to AI.
Yeah. And we're doing humanoid robots now and AI systems. But yeah, I got really interested in computers really young. Started a bunch of startups in high school and college.
At first just mostly things on the web, like selling things. Did a bunch of different types of products I was selling on the internet throughout high school and college. Small drop shipping, retail, electronics, all kinds of things. Affiliate marketing and just fun stuff. It was nothing serious, just playing around the internet, trying to make some money. I didn't grow up with money. So the internet was a way to maybe make some money, it was really fun. I loved the ability to go out and create things and kind of control my destiny. So it was just something I attached to really early on.
Right on. Right on. Do you have brothers and sisters?
Yeah. Well, I mean what is he a farmer?
Colby. No, he actually runs an AI defense company called Scout. They're doing basically building autonomy and AI models for defense in the military.
So you guys both got into AI.
We both got into AI. We live like a block away from each other today.
Yeah. We grew up together really close. Went to the same college. We're different ages, a couple years apart. And then we were in New York for about 15 years together. And then he just moved out to California. We live literally a block away. I see him almost every weekend. And he has a startup basically 10 minutes away from mine, where I'm at now. And he's doing great, man.
What, I mean, what do your parents think when you're coming home with what you guys are involved in and what you're creating? I mean, it's just such a wild, you know what I mean, from farmer to this.
Honestly, I think one thing my parents both drilled into both of us really early was farming is very entrepreneurial. Like my dad ran his own business, you kind of have to go out there and put the work in or you're not going to get paid. So early on he's like, listen if you want to control your destiny and if you want to make money and be able to do what you really want in life you need to run your own business. And that was beat into our heads growing up, at some point you need to get out of here, get out of farming. It's not doing well and you need to start something on your own. And so just by default I was like okay this is what I'm going to go do since I was a kid.
But you got some proud parents, man.
Yeah, parents are great. They're like, 'What the hell's going on here? What are you doing?' But I've been doing pretty crazy stuff for a while now. So, I think it's gotten to a point where even at Archer, we're building 6,000lb electric aircraft and before that doing internet startup stuff, but it's kind of been working on crazier stuff now for a little over a decade.
Were you building stuff as a kid, too?
Yeah. Constantly building stuff.
Stuff on the farm. Building stuff in software and internet. Just like I just love building stuff all day. I'm very big into science and mathematics. I'm more of a visual learner, too. I like building stuff and seeing it and touching things. Even doing internet for like 10 years, I just always sat there every day wishing I was working on hardware stuff I could touch with my hands. Stuff like when growing up was rebuilding computers or just on the farm and building stuff. I always envied things that you can go touch and build.
Wow. You know, basically like Adams, man. So where do you go? Where'd you go to school?
I went to University of Florida.
Where do you go from there?
So after school I moved to New York and I started working on software startups. And during college I was working on a bunch of side small internet things. And then shortly after college, I started a company called Vettery. And the goal was to basically build, I got really going through college is like you got to look for a job. You had to go find something full-time and got caught up in the whole interviewing process of looking for a job. I just thought it was so broken, applying for jobs and never hearing back and you have to go through head hunters and then it basically became somewhat of this boys club of trying to figure out where you went to school and then certain people knew other folks of how to get in and it wasn't very much a meritocracy and I just thought the whole process was extremely broken. So I started Vettery, we were basically an AI recruiting marketplace. So the goal was if we can get all the world's talent and hiring on one platform, understand their needs really well. Can we make matches at scale without any humans involved? And the head hunting industry is hundreds of billions of dollars a year.
I won't even, I won't use it.
Because I just keep hearing everybody gets ripped off by that.
Ripped off. It's so expensive. Like pay like $50,000 a hire. It's insane.
And then they'll coax the guy out that they just brought to you and have him go to another job. I was like...
Yeah. They'll force you into this role so they get paid a commission.
So Vettery is a connector.
Yeah. Connector. Well, funny enough, we ended up selling to the world's largest recruiting company that does staffing, but let's leave that for a minute. But we basically started in 2012 and the goal was how do we put a lot of job seekers and a lot of employers on a platform, understand their preferences, and match them at scale. Like just how do we use algorithms? At the time we were like let's use AI but it was basically how do we use a lot of algorithms to figure out what people want and then make matches so you can just push a button connect the right folks and then make placements. And then we end up charging, most of our revenue came from subscriptions from big companies like big banks or startups or tech companies basically looking for talent. We started just in tech in the US. So at one point we had about a little under 20 or so cities globally that we were operating in.
Wow. But most of it was tech talent. The tech space is, you know, at that point. And so how long ago was this?
Started in 2012 and then ended up selling the business in 2017 or 2018.
So about five years, six years. Right on.
Okay. So Vettery was a really tough, I basically went like I didn't have much money. Went fully all-in with the business. Went into debt at one point in 2015. The business was having a tough time and then we end up selling, end up things end up going doing really well. The business completely hockey-sticked in growth when we got all the things figured out. Vettery was, and then ended up getting approached by the world's largest recruiting company. The same groups you and I were talking about, the same groups we were trying to take out of business. And they were like, 'Oh we want to acquire the company.' And at the time I was completely dead broke and put everything out of the business. This was at the point almost 7 years in and they, we were excited about an acquisition a year before that at $10 million from one of the big tech companies and they came in at $110 million. And it was a good time for me. I felt like the business was doing well. I learned a lot and I was kind of ready for my next chapter. So ended up selling that business to the Adecco Group. It's like the world's largest recruiting company.
But you didn't even have it for sale. They just approached you.
Yeah. We didn't hire a bank or anything. Listen, at the time we were doing like I don't know 20, 30,000 interview requests a week. Like so that was no humans involved. Like think about how many humans it would take to do like 20 or 30,000 interview requests. It's like, you know what I mean? And then manage all that process. So we were like, the growth was just unbelievable. And there's something better to a human jamming in roles, right? Like you need, and then to the extent you can get all the world's talent there and all the world's companies looking you can really create an amazing environment where you can get people to the right jobs. And right now it's not like that. It's like a really black box trying to both finding talent and looking for a job. It's just a terrible experience. So we kind of, that clicked. Yeah, the world's largest recruiting company came in and said, 'We got to buy this thing.'
Yeah. I sold the business and it was great. It was a good time for me. I really at the point was a point in my life where I really wanted to do something much bigger. So I took about a year, so it took about a year by the time I got the term sheet to sale to when we actually sold and closed. It's a long process you have to go through like tons of docs and then you announce the deal. Then you actually close the deal and then it went into escrow then it finally hit my account. It's kind of like one of those processes. And I want to go work on something really important, hard, and a couple industries that I've been interested in, robotics and aviation and some areas of security for basically since college. And I basically spent a lot of time trying to figure out if I was either going to work on, at the time school shootings basically 10x'd and I was like man there's got to be something to do here. And then secondly, I really wanted to work on flying cars. Having watched lots of sci-fi as a kid is like, man, I really want to go. There was a near-term problem of we got to go help with security in schools, K through 12, mostly in the US. And then how do we, I want to work on flying cars. And I ended up making a decision to work on flying cars at the time. So in 2018, shortly after the sale of Vettery, I started Archer Aviation. And basically the story here is you can build an electric aircraft that can take off like a helicopter. If you take off like a helicopter, you don't need to place the airports outside of cities. You can place them inside of cities. Like think about a normal helicopter can take off from a building or a helipad. So you take off, if you can take off vertically, you can basically nestle the aircraft inside of cities. Half the world lives in cities today. By middle of the century it'll be like 70% of the world. And you just can't get around, it's just gridlocked everywhere in major cities. It just sucks to go like 20, 30 miles. It takes like an hour in most cases. So basically you can build, design an aircraft that can take off vertically and then fly like an airplane. So you can get a lot of distance and you can basically then re-architecture the whole aircraft to be fully electric. The reason you want to do that is for cost and safety. You basically can make it a lot less expensive. You can put a lot less parts in the aircraft that are also good for safety. So you basically can build an electric flying car that you can move around. So instead of calling an Uber or driving that might take you an hour in LA or SF or New York, you basically can fly there in 10 minutes. And if we can pull everybody together in a kind of Uber pool style business model, you can do it for as cheap as an Uber. But the problem was I didn't know anything about how to build electric aircraft.
I mean, where do you, yeah, I'm just, I know you just sold your business for $110 million, but where do you get the confidence to, where do you get the confidence to go, I'm going to build vertical takeoff and landing flying cars. Now,
I mean, listen, I didn't wake up to this world like learning how to build software. So I learned how to do that and run engineering and run the company and there was a lot of through trial and error. And I just felt like one, I just felt I could learn it. I studied industrial and systems engineering at University of Florida. So and then you know ran engineering and ran the company at Vettery. So I basically just hit the books. I tried to learn as much as possible about three subject areas. First was electrification, which at the time electric vehicles were really doing well and even drones. Vertical takeoff and landing, like vertical lift which is traditional rotorcraft or helicopter. And the third is winged aircraft like airplanes. You really need wings. So you basically have to learn about those three subjects. So I started, I basically bought, my basement downstairs at home was like every possible book on these subjects you could match in and started reading as much as I possibly could. This was during the year transition as I was transitioning out of Vettery into Archer. I was reading every possible thing and then I found a small community of folks that were hosting on-site either half-week or week-long courses for this. And so I would go to these, sometimes they're sponsored by NASA or by colleges or whatever it would be on basically rotorcraft or electric propulsion or winged aircraft aerodynamics and I would basically try to learn as much as possible. It got to the point where I was completely obsessed with this algorithm I was building on electric aircraft sizing. Like how do you actually, how would you actually build an electric aircraft? So electric aircraft, what's interesting is in rotorcraft, you basically want to make, to create the most efficient lifting device possible you need as much of the rotor disc area, like in terms of surface area as you possibly can. It's why the helicopter rotors are so large. You really want that to be really large. That'll reduce power and get you up off the ground. In electric aircraft, the problem you're starting with is you have like 1/30th of the energy as you do in kerosene in a battery pack.
So you're off the bat 1/30th less range or 1/30th less energy. And so power becomes the dominating factor of how to basically build electric aircraft. Like how do you get power down as much as possible? You really want a lot of disc area. A lot of disc area is, well one it could be good for power but it's also bad because you have no redundancy in the system. You have one rotor blade if it doesn't go well you go down. With electrification you can basically build much smaller rotors and be fully electric. And the reason you can't do that with traditional turbo fans or engines is it gets too inefficient at these sizes. You can't build 12 propellers on a helicopter. The efficiency just drops to nothing. So with electrification you can size electric motors to small sizes and they're still 90% efficient. So like a small electric motor on the table or a big one the size of your chair, same efficiency. And when you do that you create a lot of redundancy across the system. So you can build an aircraft with 12 electric motors.
So this is, the rotors are underneath.
Basically the problem here is you can design it however you want. You could put a bunch of rotors along the wings. You can put them laterally across the fuselage. You can make one big one. You can make 30 small ones. So how do you design it? That's the problem I hit in 2018 was how do you actually do this? And so I basically was like a crazy man trying to design this algorithm to what is the ideal aircraft design and then how do I go build it? So, I was actually at a Hyatt Regency Hotel in Atlanta in 2018. It was an electric propulsion week-long design course and aerodynamics course for winged aircraft. And I met a guy there that was basically in the engineering department at University of Florida. He was doing his PhD in aerospace and asked him what he was doing there and he's like, 'I'm from University of Florida.' I was like, 'Oh, I went to school there as well.' And he's like, I'm like, 'What are you doing here?' He's like, 'Oh, I want to go do a career in eVTOL aircraft.' And it's called electric vertical takeoff and landing. So, you know, a helicopter is a VTOL and you put a little E in front of electric. And he asked me what I'm doing here. I was like, 'Oh, I'm starting a company to do this and I need to figure out how to go build these things.' And he's like, 'Well, listen, my professor runs a small drone lab. He's got a full building. He's got 12 PhDs. Why don't you come down and meet him and see if you can start building aircraft with him?' So, I flew down that weekend to go meet his professor that runs all of basically mechanical engineering, aerospace. And long story short is I ended up taking over his lab. And me and him and his team started building aircraft in 2018 and 2019 down at University of Florida. And I temporarily moved down there with my daughter at the time and my wife living in Gainesville, Florida. And it was great. We ended up building, I ended up funding a lab right off of Archer Road, a new lab because we needed more space and we ended up calling the business Archer Aviation. It was the main road down at University of Florida. And I spent the next year, year and a half basically modeling and building electric eVTOL aircraft.
Yeah. And it was a great time in my life. And the problem is there was no intersection of folks that knew electric, knew rotorcraft or knew airplanes. There was no Venn diagram of overlap.
So there was nobody in the world that...
I had to go from scratch and learn it from first principles. I ended up moving the company out to California a few years into the business. Things took off from there. We built bigger aircraft. I took the company public within three years of starting it. We're a six billion dollar publicly traded company today. I designed basically four or five generations of aircraft at Archer. It was hard. It really set me up well for doing Figure and the rest of the stuff we can talk about later. It was hard. Even going public was probably one of the hardest experiences of my life.
We went public through a SPAC process. SPACs at the time, four or five years ago, were all the rage. It was a special purpose acquisition company. It was basically companies going public through a reverse merger. It was hard because in 2018, 2019, coming off of software, I had never done hardware before. So, A, it was hard to raise capital, and B, there was nobody funding deep tech electric vertical takeoff and landing companies. The big venture capital groups were not funding SpaceX or Tesla or Rivian. None of these were getting funded by traditional investors. They weren't raising money from the named investors we all know about now.
Are they always behind like that? The mandate for most of these VCs in the Bay Area or Silicon Valley is not to do hardware.
Gotcha. They don't really, and if they do hardware, they don't do deep tech. They don't do rockets and autonomous vehicles. I don't think there's a single top VC in the US that's invested in a humanoid company. As of six months ago, they just don't do this stuff. I ended up going all in. I had just sold the company for $110 million. I made a lot of money personally and ended up going all in on Archer through the IPO. I put all the money I basically had, I bought a house and the rest of the money went all into it. It was a stressful period. We went public through a US SPAC and the reason it was tough is we ended up getting to a point where we just couldn't raise enough money privately. It was either raise $100 million privately at some valuation of three, four, $500 million, or go public and raise like a billion dollars.
And we ended up going public and raising a billion dollars.
Wow. You've got a huge appetite for risk, huh?
And we got sued during it.
Yeah. We got sued by basically Boeing and a big startup that was founded by Larry Page, Google founder.
That's got to be intimidating.
Yeah, it was. I woke up to a front page New York Times article about it. It was crazy. The backstory is Larry Page started a company in the Bay Area about 10 years ago called Kittyhawk. They did great work over 10 years in electric VTOL aircraft. I ended up taking basically the core 10 to 15 folks that were there, and they all came over to Archer within the first two years.
And they retaliated by trying to harass us while we were going public. So yeah, just a crazy story getting public. We ended up getting public, a billion dollars on the balance sheet, and we just started building aircraft and building the service. Think about the app and how you're going to check in and how you're going to build places, real estate to fly into. The engineering work we had to do around there of designing, it's basically a flying robot. You have battery systems, electric motors, sensors, embedded software, and control systems. Basically like the robot you saw this morning. It's a 6,000lb four-passenger piloted robot and it has 24 degrees of freedom on the system, wing flaps. We tilt the leading edge six motors 90 degrees for basically takeoff vertically and then going to forward flight. And then all the propellers, fan blades, have variable pitch propellers. So it's a highly overactuated system that needs really good software. No human can fly it basically without really good control software.
What altitude is it flying?
About a few thousand feet. So about 2 to 3,000 ft above ground level.
And that's what it would normally be at.
Yeah. Traditional helicopters fly at these levels.
I think a lot about Tesla and all the EV vehicles that are coming out. The government just seems so far behind on AI. You just brought up gridlock in all the cities. I've always wondered, when are we going to go full EV? I know there's a lot of pushback about that from an overreach standpoint, but if you just think about the traffic in the cities and if you have the AI processing all this, even without air vehicles, I feel like a lot of that would go away because the AI will route you the quickest.
Yeah. And take all the traffic patterns into account and it would just flow a lot easier.
Yeah. But there's all this government regulation.
I think it's also hard because if you look at the number of installed cars in the world, like a billion and a half or so installed cars, we make like 80 million or so cars a year in the world. It takes you on the order of 20 years to replace all the cars. So all the cars were electric and autonomous today. Autonomous cars have autonomous hardware in them. It's not like you can just go out and retrofit all the cars in the world right now. It's a hard problem.
Well, I mean, if you look at Tesla, for example, it can self-drive, right? It can come get you.
But when you're driving, if you take your eyes off the road, it wakes you up. You have to come back. It seems it's inviting more error into the road by doing that in my opinion.
Yeah. It could be more dangerous.
We're just in this transitory state right now where in like five years everything will be fully autonomous and trusted and fine and you won't have to do that. We're just in this chapter in the book for the technology roadmap here where we're living through it and it's a little messy and it's not quite straightforward and we don't quite know where it's headed next. But where it's headed is at some point in five years, when our kids grow up, they're never going to have to think about this. It's just going to be autonomous from the start, native. And it'll be trusted and easy and safe. I think we're just living through this period right now which is a weird thing, but if we close our eyes long enough, you'll have this autonomy and electrification everywhere.
How long do you think it'll be?
I mean, so I live in the Bay Area. You can take Waymos now. I can take them everywhere and it's unbelievable. They're all over over there.
They're everywhere. Yeah, they're in my... I'm in South Bay. But they were in the city for a while and now they're in Palo Alto, Menlo Park, San Jose, all over the place. They're really great. I take it. My wife and I, when we go to dinner and stuff on the weekends, we take Waymo. It's so fun.
It sounds so basic. Like, I take a Waymo, it's fine. It's just awesome, man. It's great. The car drives so humanlike and it's such a great experience not having a human there, to be frank. I love it. I order so many Ubers and stuff and the car smells or it's dirty or whatever else and it's just easy. It's really cool. So technology is in the early chapters but it's all here. We're going to have autonomy at scale everywhere. It's just going to take some time to roll that out. It's the time it takes to get the technology mature enough where they can run in enough cities and enough places, and then it's the time it takes to get the install base of autonomous hardware and software in all these places. That's going to take some time too. We just can't snap our fingers. We just don't have enough install base of autonomous vehicles in the world. I think Tesla's got like 10 million cars on the road and maybe there's thousands or so of Waymos, but you have over a billion cars on the planet. So you need to make a large fraction of that all autonomous. So you're looking at this isn't going to happen in a year or two. It's going to take some time.
When are we going to see your vehicles?
The aircraft? We have them now. We fly every week in California. The challenging part with Archer is that we are governed by the federal airspace. So to fly passengers and charge money, we have to have basically a type certification from the FAA. That process moves at the speed of the post office and the FAA is not incentivized to put anything in the air unless they know for sure it's going to be really safe. The safety standard for us that we want to certify to is 10 to the minus 9 in terms of hours of reliability before a catastrophic event. So that's one in a billion hours.
You can't... that's the standard. When we fly, it's the safest form of transportation we take and it's because of those standards governed by the FAA, which is great. That's the bar you need to be at and that's the bar you need to hit, especially taking passengers over cities with aircraft overhead. You need to be at those levels. So that's the long pole in the tent for us. And wherever you go, if you go to Europe, it's EASA, or in China, wherever you're going to go, there's federal mandates to get basically an aircraft to take passengers. So we're in the middle of FAA certification now. We hope to be certified as soon as possible basically. But it's not something you can... there's not a date on the calendar like you'll be certified here. You have to work through a very long and slow process with the FAA to get through this. And then we're also dual tracking that against a couple different entities globally now to make sure we can get certified and get in there. But it'll happen, man. The aircraft, again, we're in this chapter where flying cars, electric aircraft, it's early. It's earlier than AVs or EVs, autonomous vehicles, electric vehicles. But it'll happen. It'll happen in our lifetime. We'll be taking these things around.
Well, I mean, what do you envision when let's fast forward 20, 30 years. What do you envision? What does it look like? Do we have roads? Do those get ripped up? What does the sky look like? What does everyday life look like?
The really important thing here about the airspace is that it's three-dimensional and the roads are not. They're 2D. And we built cities now and houses and restaurants all around these places. You can't... there's nowhere to go. There's no more roads to build in these cities. So you have left with no choice. And then humanity are moving to cities. We have this secular trend where we all want to live in cities right now. It's like half the world lives in cities. It'll be like 70% by 2050. So we're all moving to cities. The roads can't grow anymore and we're constantly moving around going to work or going to restaurants and it's getting worse. The arteries are hardening here and it's getting worse and worse. It's some of the worst time to spend on a road in traffic. It's so soul sucking. It's just the worst time to lose. So the good news about the air is it's three-dimensional. You can stack basically an infinite amount of roads in the air at different altitudes and even laterally. So you can basically build little tunnels in the sky and you can basically stack them. You can put orders of magnitude more things in the air than you can on the road. It's the same for below ground with tunnels. So the future of travel in cities is below ground in tunnels and above ground in the sky.
Yeah, exactly. Just dig tunnels and it's great. The only problem with tunnels, with the node system on the ground, with say we call them vertiports, but basically real estate for flying cars, is you can, let's say you had 10 different or even five, say 10 different vertiports inside of a city, places to take off and land from, you can travel between any one of those routes. So it opens up basically exponentially more places to go to. I can go to any node on the system at any time.
So, hold on. So you're saying in order to take off and land, you'll have to go to specific locations. You won't be able to do it from your home.
Yeah, you're not going to take off and land from your home. Just because acoustics in the neighborhood, it's going to be too loud. You need a decent amount of infrastructure for that, for charging and for passengers and cleaning and checking in and stuff like that. They'll be in your neighborhood and you'll Waymo there or walk or take a bike and then you'll get on these and they will go to any node of the network. You can't do that with tunnels. Tunnels have to go from A to B. You can't jump to another tunnel downstream. You can do that with the sky. You can basically jump to any node on the network. It's exponentially more routes you can basically do with less real estate. And then you can basically stack orders of magnitude more traffic and humans in the sky. So my envision is that for most trips that you're traveling over 20 minutes, all that will move to the sky. And not only that, but you will have cities being transitioned to a point where you can live well outside of cities and get to cities really fast. The reason we live in cities is because we're working there and we have friends there and it's like I want to go to dinner with somebody. I want to see my buddies over here and we want to go to work over here or go to the mall over here. Everything is there and that's what we want to be. We're social creatures. We want to be there next to other humans. So now that you can fly this 150 mph in the air with no traffic, point to point, no stop signs, no construction, no things jumping out in front of you. You don't have to travel different distances. You're going straight from A to B in most cases. So you're removing 10 or 20% of the distance just by going point to point and then you have nothing stopping you going 150 miles an hour most of the way there. You can live far outside of cities and get down to city center in under 30 minutes.
So will these be personally owned or will these be... this will be like an Uber service?
It'll be like an Uber service. To get cost down, you'll basically just pay per trip. You'll pull up an app and you'll go like, I want to go downtown. It's whatever, it's 40 bucks and I'll be there in under 30 minutes. And you'll say, Great, I want to be there at that time. You'll hit a button. It'll be on demand. You'll ride your bike over or walk. You'll get in one. It'll leave in 7 minutes. And then you're basically flying right down to town.
Holy... And you're saying this will be in every neighborhood. This will be very accessible to everybody.
Yeah. That's... you're designing a whole... electrification allows you to reduce the cost and the safety burden of all this. We have like a normal helicopter could have like 100, 200 safety critical components that any component gives out the helicopter can go down. An electric aircraft has none. You can lose a motor, you can lose a battery pack on board and still fly safe without having this. And so from a safety, from a part count, from a cost, from an acoustic signature, it's not like helicopters are loud and very noisy. It's just a much better technology for this.
Have you been in one yet?
I haven't flown inside of ours yet. I've been inside of our aircraft. We basically have professional pilots at the company.
Yeah, test pilots. They're career test pilots and they're unbelievable.
I'll bet. A lot from the military, a lot from the big aerospace groups, and they're just professionals.
What do they think? They love it, man. This is the future of aviation, man. Everything's going electric, and it's so crazy.
Yeah, it's crazy. It works. It's crazy it works. And crazy we're in the right time period to make this happen.
Yeah. And the good thing about Archer now is we've demonstrated... the hard part is making sure you're in the right decade. You don't want to go do this and you find out it's a 2040 event. You can't get it done. It's just a waste of time. So the good news is we're in a sweet spot here where this is going to happen. Aircraft now work. We're certifying now with government bodies like the FAA to make it happen. And we have a good balance sheet with cash. The team's great and so it's just get certified and get this thing going.
Damn, that is some... you're really changing the world.
Well, we're the start of it but hopeful to make this thing work.
Let's... Yes, let's do it.
So, I spent like five or six years working on pretty crazy robotics work at Archer. The ultimate meta problem in the robotic space is can you build a general purpose machine to do everything in the world, much of what humans can in the world? And I have this big belief that we have this weird biological species, we have these weird hands and arms and legs and certain height and sensors, and then we ended up building this world around us so we can interact with it. If we get dropped into Mars today, we're going to build coffee cups that we can hold and stairs and doors and we're going to build this stuff again. It's like the human operating system. We're building things we can use and operate in that makes it look easy for our lives and we built it around the form factor that we are. Meaning if we looked differently the world would look different. Our espresso machine would look different. We might not even like espresso or caffeine in this case. So we built this whole world around us. The holy grail for robotics is can you basically build a general purpose machine that can do what humans can, which for me is a humanoid robot. And a humanoid robot is just a robot that has a human form. So has legs so it can walk upstairs and walk over uneven terrain or pick things up off the ground and bend down, which are important, legs are important for, or reach up. Has arms and hands so we can manipulate objects and do things like grab stuff, open these gummies, and fold laundry and do real work. And we have the right sensors so we can see the world and understand what to go do and use our biological neural net to kind of figure out how to reason from. Having worked on aircraft for five or six years, I thought it was pretty possible to go build an electric humanoid robot. And electric's important for cost and it's important for safety and it's important because the performance will be much greater. And at the time, maybe one of the best humanoid robots was probably the Boston Dynamics Atlas. It had a hydraulic system. It was really heavy and big and high torque and very leaky, like the oil's everywhere. And it also didn't run very long, maybe ran for 20 minutes on a single charge. So you needed to radically transform the hardware. And then you needed to figure out a way to build an AI brain. The humanoid is so complex. It has let's call it 40 degrees of freedom. A degree of freedom is like a joint. So an elbow is a degree of freedom and shoulders got three. A ball and socket has three, a pitch, a roll. And our robot has about 40 degrees of freedom in it. Each degree of freedom is a motor that can spin 360 degrees. So if you only look at how many positions the body could be in at any given time, this is a position, this is a position and keep moving. The amount of states, mathematically it's 360 degrees to the power of 40 actuators. So there are more states in the robot than atoms in the universe. There's more positions the body can be in.
By far. It's a much greater number. I've done the math a few times. Very confident in this even though it sounds ridiculous. So you just can't code your way out of this problem. How are you supposed to write code? How's a human supposed to write C++ or code to tell the robot at any given timestamp what to go do? If I want to grab this, I need to move my whole upper body, maybe lean over and I'm moving my fingertips and my wrist and hand to get in position to grab this. It's an intractable problem for code.
I mean you were saying earlier, I'm going to butcher this, but it's updating the foot 200 times a second.
Yeah, our controller is running... our whole controller, so we have a main computer that is processing what to tell all the joints to do 200, a little maybe more than 200 times a second to make sure we can just balance and then we can do the task. It could be reaching over and grabbing this or balancing. If we run that too slow, we just don't have enough feedback. Then we just fall over. We have to fully balance. It's dynamic so if you power it off mid-run, it's going to just fall down.
It's not like a four-legged dog or quadruped robot where at any given point it's usually statically stable.
So it makes it very difficult because you have to be able to... even to move my hand, I'm moving my pelvis and my whole body, my torso is moving, my head's moving. All of it becomes very complicated. Now it's not just move my hand, it's move my whole body to get my hand in the right spot. So every joint, all those 40 joints have basically position encoders so we know exactly what position the motor is at or even in the case of the knee. And we have force sensing, torque sensing on board. We have the ability to detect all the forces that that knee is seeing. It could be really high when it's walking or it could be powered off and have no forces on the leg. All of that feedback is being sent to the main computer and then we're telling all the joints what to do over 200 times a second. Some of the other feedback is happening at like five or six kilohertz. So the force feedback is happening five, six thousand times a second to the motor control on board. And we do the motor control, the brain for all the motors is done locally at the motor level because it needs to happen so fast. That's being fed back to a main computer that runs control software that tells the rest of the whole body what to go do at every timestamp to keep balance.
Starting something new isn't just hard, it's terrifying. Before I launched the Shawn Ryan Show, those what ifs were loud. What if nobody listens? What if this fails? Walking away from what's familiar to build your own thing takes real faith. But it ended up being one of the best decisions I've ever made. Whether you're starting a podcast or launching a store, it helps to have a partner like Shopify on your side to help ease those worries with their expertise and tools. Shopify is the commerce platform behind millions of businesses around the world and 10% of all ecommerce in the US. From household names like Figs to brands that are just getting started today, one of the biggest hurdles is efficiency, and Shopify helps you accelerate that. Whether you're uploading new products or trying to improve existing ones, the platform is packed with helpful AI tools that can write your product descriptions, page headlines, and even enhance your product photography for you. They also allow you to tackle all your important tasks in one place. From inventory to payments to analytics, there's no need to juggle multiple websites. Everything is in one spot to make your operations smoother. And you've definitely seen that iconic purple Shop Pay button. It's used by millions because Shopify has the best converting checkout on the planet. It helps boost your conversions, which means fewer abandoned carts and more sales for your business. It's time to turn those what ifs into... with Shopify today. Sign up for your $1 per month trial at shopify.com/srs. Go to shopify.com/srs. That's shopify.com/srs.
So, getting back to your original thing, my bet, we're like three and a half years old or something like that at Figure. I bet three and a half years ago in 2022 when I started the company was that this was possible now. And my view is like it was... I don't... over some 10 or 20 years this will work. And so I basically started on this endeavor to go basically rebuild from the ground up humanoid robots and AI software to try to see if you could make this work. At the time there was no good precedent. There was no precedent for showing that there was no AI that had ever worked on a humanoid robot in history. And there was no human electric humanoid hardware that was remotely okay to show it would work and there was no hands. There was none of this stuff.
So I actually had a lot of trouble early on even getting people excited about this because they were like, what the hell are you doing? And so I ended up having to basically self-fund a lot of it. The first year I self-funded all of it.
Yeah. And it was a lot. We got the business to a million a month of burn in month four. But it was like I knew what to do. We built a 40 person team as fast as possible. And I knew how to spin up hardware and software. The key characteristics of robotics like electric motors, battery systems, control software, embedded systems and sensors. And then within electric motors, we build actuators. They have a rotor and stator and a gearbox and sensors. Electronics and wiring and connectors and multiple sensors inside of there. And then firmware that lives on the microcontroller on the motor control side. And then we have thermal characteristics. It's hot. And then you got to all make that work at very high speeds and high torques. Meaning motors don't like working when they're not moving. Motors hate not moving. Motors want to run on highway speeds.
They love that. Whether it's a generator or an appliance in your home or an electric car, they want to run at highway speeds. They're designed to run at full RPMs. That's when they're the most efficient.
When motors are stuck and not moving but holding power and holding forces, they're at a really bad point of the torque speed curve. So they're not built well for this. Humanoids use all the time. When we're standing, we're not moving but holding forces. When we're holding something out, like giving me the gummies, it's not moving now but holding forces. It's just a really hard engineering problem on just that one little aspect which is hardware, just motors. So we have whole teams in those areas just in that one little area here doing rotor design, electromagnetics design, stator design, gearbox design, sensor design, motor control design, all of this inside the teams. It was an enormous lift to just get the team members there to do it. So we spun up a team to operate that really quick and then now looking back, I think we raised two billion or so. It's a much different story. We ended up building Figure One, which is our first generation robot, and had that walking in under 12 months. So from when we...
From inception to... Wow.
Yeah. From when we had basically incorporated the company in 2022, our goal was can we get a robot walking by itself at these dynamics in under 12 months and we did it with two days left in the year. At the time it was the fastest time in history for anybody to do this. And then from there continue to build the capabilities. We built generation two, Figure Two, which is that guy right there, is our second generation robot. And I think one thing we did before we kind of moved even to gen... while designing gen two is, I think it was probably 2023 at the time, we did a demonstration where we basically wanted to put this K-cup inside this Keurig and run it. It was just a pretty simple, nothing crazy, but we had a Keurig machine, coffee cup, and a K-cup and we had to go grab the K-cup, open the Keurig, put it in, close it, run it, and then make coffee. It sounds simple but for a humanoid robot to do that is extremely hard. And then we wanted to do all of that with just neural networks.
It sounds simple but I mean... simple task for kids. Give it to a four-year-old.
Yeah, exactly. The dexterity on the hands of that thing is just to hand you the bag of gummy bears.
Yeah. Then can you do it with neural nets on board? Can you not code your way out of it? Can you take in camera pixels and then output trajectories for the motors through a neural network? No code. And we did that in 2023 on Figure One. And it was probably the most significant demonstration we've done in four years now, almost four years, where we were like, internally we were like, how do we get neural nets to run on a humanoid? And I don't know. I think it was probably one of the first examples in the world to ever have shown that. And this was like game on. This is like, let's go build really good humanoid hardware. Let's make it cheap and really reliable. Let's make sure it can do what humans can from a hardware perspective. Meaning you want to look at like a phone where you can just add new apps to it, like the do laundry app, and the hardware doesn't need to change. The same exact hardware. Like humans don't need new hardware to be able to go off and learn how to do new skills now in the physical world. So you want to build the humanoid hardware so it can do everything basically a human can or as much as possible and then you want to go all in on neural networks because you just can't code your way out of this problem. And that was the first moment in 2023 we're like, God damn, this is going to really work. This is going to be humanoid robots, hardware gets good and then you're basically going to be... this is going to be a data play to train neural networks to run on humanoid hardware and do what humans do. And then we launched Figure Two. We did a lot more work. We started unveiling Helix, which is our neural network stack internally that we do here. And now we've designed Figure Three, which is our third generation robot you have here, which is the best humanoid hardware in the world by far. And we're now running robots that do... I watched it the other day, unload dishes and fold laundry. We had Figure Twos at BMW last year that worked six months every single day.
Six months every single day.
Every single day it worked a 10-hour shift every day for six months. And it was just the first time for us getting robots out to the real world doing real stuff. It's fun doing demos at the office and showing it can really work. But the real level boss is how do we get robots out and do clients fire us? Do they love it? Does it work? And the goals we have for clients is hard because we have to do human work. So we get human KPIs in terms of speed and performance. Humans, in the case of manufacturing, they don't... you might mess up every once in a while but you refix it. So you're not messing up every single time. You're pretty fast. In most cases the humans are there, not quitting or not showing up to work, but sometimes that does happen. So it's a hard bar to go hit and we have to wake up every day and be able to do that. And so we had robots on the manufacturing line. They have a basic body shop that basically builds X3s and X5s. In January of 2025, we started building our first BMW X3s on the line.
And I bought the first four that did that. They're at my campus now. One's at my house.
And they didn't build obviously the whole car. There's a ton of parts, but we did part of the whole process.
They're great. BMW is like, if you go into a car manufacturing company, they're like the best roboticists in the world. There's robots everywhere. There's these giant 12-foot KUKA manufacturing robot arms on the floor. They're bolted to the ground. They're massive. These things are carrying car chassis around like they're kids' toys. The cars are so big and so heavy a human can't hold it and pass it around. So you basically have machines that are building the car and then moving the car. So the whole body shop line is automated end to end... not end to end, but there's humans involved but the car is being built by machines. And then there's machines everywhere. There's special end effectors on every machine. And they're switching these things out basically in real time, grabbing a big end effector. The end effector is something that is grabbing a part. These end effectors are the size of my chair. They're switching them out in seconds.
One or two seconds. They're doing it really fast. And then they're basically building a car with this. There's robots everywhere. And so BMW, it's been a privilege to see how much automation has gone into automotive. It's unbelievable. These machines kind of make what we're doing sometimes look like little toys.
We're doing some very complicated stuff, but the car manufacturing is just...
So, what specifically were bigger figures doing?
Yeah, we had a body shop line that's basically building the rear header. It's like the back plate. So, the body shop, they basically build the car by putting basic sheet metal together, welding them onto the chassis, and then they basically build the car around that. Like ended up putting the seats in, bolting them down, putting the car doors in, wiring them up, the harnessing. And we were in the body shop line helping basically attach the rear header, like basically putting the rear header on this fixture. So today, or last year when we were there, we basically take a piece of sheet metal and we basically put it on this fixture and we do that over and over again. And they do that, you know, 10 hours full shift. And there's three different parts on those, three parts go on. This thing rotates and this big giant KUKA machine, like this robot arm, goes and spot welds it and switches it out to another and then grabs it and puts it down the line.
So like these facilities are being fed these parts into the machine and we were a piece of that. And the goal was just like, can we run robots every day? You know, are we gonna get our ass handed to us? You know, is it gonna be easy? Is it gonna be hard? And I think it was in the middle. I think we got the robot to a great spot where every day was great. I think the biggest learning lesson we took away is I really cared about if can we do that and can we clone it times a thousand, times 10,000? Would we have any issues scaling? That was the part for me. Like is it just like, you know, does it completely shit the bed and means like we need to rework our plan and go back to the office? Does it do it incredibly well and you can just copy paste this thing to everywhere in the world? Like how did it work? And the biggest learning lesson we got was that the robot that started the first day at the start of six months and the robot that ended the shift that day was the same. Like even though we had multiple robots in operation every day, we had like the same robot that did the start and the finish.
And it was cool, like it was the same robot ended six months later. And this was a thing where, you know, the worry was that humanoid robots couldn't last a month, couldn't last a week in these kind of environments. And wear and tear, just like it's running like 40 degrees of freedom motors every single day. Can they operate really well? I think from a hardware perspective, it did an A+ job. I think from a software perspective, we did like a B job, B+. And that's mostly from my perspective, like the architecture decisions I made to scale. About half the stack we had traditional code and heuristics in. So like the controller to walk was done by a C++ controller. It was done by code. The walking you saw today we had back then was done in code.
The rest of this we had a bunch of other stuff in there done by neural nets and like some of the perception stacks, some of how to move parts around, everything else. And you know, this was a year and a half or so ago when we were first launching. And I was like, man, the biggest problems we're having is the coding parts get stuck. The robot either doesn't see the right thing on the part and misses, the object detector doesn't really understand what's going on. The controller when it gets out of bounds of like what it's ever seen before, like you have carpet in here now and it's like really squishy.
The robot's doing fine, which is great, but I think our old controller would not do well. It's like you have really shaggy carpet here and it's harder for a robot to walk around. And so that was like we had a really difficult time seeing that even though it did well every day, seeing that scale to like lots of robots. So we went back to the office. This is about a year and a half ago and said we need to basically refactor everything into a neural network. And one of the big things, and I think we just announced Helix 2 two or three months ago now, I forget, like end of last year, and it's basically entirely down the stack including the controllers in a neural net now. There's like no code left really on the robot. Maybe some code in certain pieces but mostly just almost all of the thing is like a neural net at this point.
We removed the need for like almost over 100,000 lines of code when we launched Helix 2.
And so what you saw today was just a robot that we can put now back in say the factory in these places that will run all on a neural net. And I think we're running these robots right now. We're getting ready for deployment to customers and they're running incredibly well. We have robots running like basically now in 24/7 shifts without stopping, without any faults for days and days. We just went like over... Yeah. We just had like record time this past week on the robot running until we saw like a fault, like almost basically a whole week.
And they basically can run like four hours or so, five hours, and when you charge, another robot knows that, steps in, steps behind the robot say and gets ready for work. The robot then backs off, the robot swaps in, like in the next 10 seconds it's doing work again. So we can run that now in 24/7 shifts where they're talking to each other all autonomous. No humans are eating, go to bed, whatever. And they're running that shifts all day and all night. And we do it across multiple use cases now at the office in 24/7 shifts. And it's just like we're running them hard.
What kind of stuff are they doing at your office?
We do a few things. We have a logistics use case that we run in 24/7 shifts constantly. We really like it. It's done with a neural net. It's moving packages around and it's a really good use case. We like it and I want to run them for months and have failures. And we still see failures right now and most of it's in software. The robot gets to some spot where it feels unsafe, doesn't know what to do and it'll stop for a little bit. And if the robot's not on the line for a couple minutes, we call it a failure and we're not happy with it. We have robots that are greeters and visitor bots that walk around the office all day 24/7. So you go to the office, you're getting lunch or walking around, you're interviewing with us. You see robots everywhere. And those run in 24/7 shifts all day, all night, weekends, Christmas day, whatever. We run them.
They'll just come talk to you.
Yeah. They'll come talk to you and you can go talk to it and ask for things. We really wanted to go like, you know, the end state for us is like it's going to replace somebody like meeting the candidates that are interviewing there, taking the conference room, getting water or coffee, like all of that end to end and whole experience. Yeah, right now they're walking in the office, like at night time they're walking the office everywhere and it's a good stress test for us because these are neural nets that are running for navigation or planning or manipulation or whatever it would look like. And it's hard. This is a new thing. It's not like these things have been around for decades and we like understand that they're really mature. They're not. So, we really stress test them like crazy by running them all the time.
What's a conversation you've had with a robot?
We've been really working on like deep memory because I think one thing I really don't like is like these conversational AIs you talk to that don't know anything about you.
It's like not much to talk about. It's like what's the weather? Like you ask things about Wikipedia or something. It's like you know the way to work. It just, it's kind of nonsense. It kind of feels really stupid to me. So we've been working a lot on like deep memory.
So it will actually get to know you.
Oh yeah. Yeah. It needs to know who you are. Like who am I talking to? Is it Sean or Brett? And then based on Sean, like do you have the permissions to tell the robot to go do something or not? Like if you're visiting, no. You might be able to get coffee or water but like you want to have it go do something new. It won't do it.
Um, I've not even thought of that either.
Yeah. Yeah. Like we want...
You want to be able to command it.
Yeah. What are the permissioning systems and authentications of the robots?
I mean like, you know, like robots in my house. My kids are going to be like, hey, give me ice cream every 10 minutes and we can't have the robot doing that, right? I get home from work and the kids are just like, you know, through pints of ice cream and the robots are just getting whatever they need. Like it'd just be chaos.
Yeah. Um, so we're voice recognition.
Yeah. You have to do voice for something. Something's voice isn't enough where like if you think about it like an extreme example, you wanted to go like order food or spend money or send a wire, like voice recognition won't be enough. You'll have to do a higher level of authentication.
And then if you have perhaps even fingerprint scanning is possible too, but facial is what you really want to do. Gotcha.
Um, so those systems are not robust enough right now. We're working through them and the goal is to get it super robust. But like, you know, we want to have conversations with the robot. We want to ask it to go do things. Like you really want the main modality to be speech with robots. You want to just like, hey man, go make me food or like do, when I'm gone today, do the laundry after you unload the dishwasher, like do laundry in my kids' rooms today or something like that, or text it. So like language is like super important UI. So we're spending a lot of time on speech.
