Robots Like Us - EP 92 Andrew Kiguel and Matt McMullen From Realbotix

Ashley Vance: All of those robots that I've seen in any of these events are remote-controlled. There's no AI. If you think about loneliness, something like 25% of the world identifies as lonely. If you're a 75-year-old man who's widowed, you're not going to go on Tinder to go on a bunch of coffee dates. I think that having a human-like embodiment for a robot and for an artificial intelligence creates a bridge between humans and technology that can't be done in any other way. Harvard came out with a report about nine months ago. And the number one use case for AI, not that there aren't several, but the number one was companionship. For all the people that think it's strange, I get at least five or six messages a week on LinkedIn, email, everywhere, which is like, "I'm so lonely, can you please design and sell something to me that will solve my loneliness?" So what I wanted was, I wanted the face to come off, like Westworld. So this is how the faces would peel off from a robot. A robot that can do karate chops and has a 30-minute battery life, what is that good for? There's a lot of smoke and mirrors on social media. I'm the only CEO of a robotics company that's got to worry about the dress for the robots and them having nice manicures and nice hairstyles.

Ashley Vance: If you are watching this episode, you will see right away that I have a couple of robotic companions behind me. And they look quite human-like. They come from a company called Realbotix. We're joined by its two top executives today, Andrew Kiguel and Matt McMullen. This is going to be a little bit of an adventure. Realbotix is trying to make human-like humanoids. They do have a background in making real dolls, companions, sometimes for emotional support and sometimes for sexual support. But Realbotix is creating a clean break with that past, and it is meant to create a new type of human-like robot that's out in the world, maybe being a greeter at a hotel, offering help at a hospital, or even being something like a tutor. They're full of artificial intelligence technology to speak and hear and communicate with you. The robots have a lot of animatronic details. Instead of doing the all-metal factory worker-looking robot or the generic ones we see that are supposed to be doing tasks inside of homes, a lot of these robots have these circular kind of blank faces. Realbotix is all in on making these things look realistic. They've got animatronic components. They've got artificial intelligence software built in so that you can talk to and hear from the robots. They're a little weird, I guess, for a lot of people. And I get it. I also think Realbotix has put more thought and artistry and philosophy in what they're making than just about anyone else. I think they've got a bit of work still to do, but I appreciate the care and effort that they're putting into this because I suspect robots won't be generic and robot-y looking universally. I think we will have things that are human-like and somebody should be trying to do this and somebody should be trying to do it well. So we went to Las Vegas to their headquarters, sat down, we had an interview. We're going to have a complete video coming with Andrew and Matt, where we see how these robots are made. And this will be on our YouTube channel. I go on a drive with one of the robots, Aria, we go down the Las Vegas strip. It's going to be fantastic. As ever, please leave a review, some likes for the podcast on Apple or Spotify, or tell your friends that you enjoy the podcast. It would make a huge difference to people finding this show. I think you'll enjoy this chat. It is a little strange, quite a different take on robotics. Fascinating stuff. As ever, I am Ashley Vance. This is Core Memory. Here we go.

Ashley Vance: Gentlemen, hello.

Andrew Kiguel: Thank you for having us here.

Ashley Vance: We're in Las Vegas at Realbotix HQ. This is exciting. There's so many places to go with this story. How would you guys, just to give the viewers, and you should absolutely be watching this episode and not just listening to it. But yeah, like, I mean, how do you guys usually describe what you do?

Andrew Kiguel: Sure. So we are an AI software company and a robotics hardware company. We sort of marry both into an all-in-one solution, but we've come up with sort of a different way of thinking about AI and robotics that I think is very unique and different from what everybody else in the world is doing.

Ashley Vance: Yeah, I mean, I live in Silicon Valley, not far from the American version of humanoid robot central with Tesla, 1X, Figure, they're all down the street. But, you know, they're all doing the worker robot that's meant to be in a factory. That's most of what I see coming out of China right now. What I love about you guys' stories, I don't see anyone else doing anything quite like what you're doing. I mean, you guys are distinctly not making a factory worker robot.

Andrew Kiguel: That's right. We're not making robots to replace physical human labor. We're also not making robots that look like metallic cyborgs. We're sort of focused more on the human interaction and different things that robots can do, such as customer service, companionship, healthcare, education. And in those scenarios, you need a robot that has a more human appearance so that it's not intimidating. And you need an AI that is flexible. And you need a robot and AI that is customizable and can work all day, seven days a week. And so we've got Aria.

Ashley Vance: Aria and Melody.

Andrew Kiguel: Aria and Melody.

Ashley Vance: And are they, they're kind of, I mean, Aria, she's got her own social media account. She's sort of, is she the one you put forward most often as like the public face of all this?

Andrew Kiguel: Yeah. So that was an intentional strategy for us in that most of the other companies have these generic robots, they're making a lot of the same looking robot. And people still identify, you know, Optimus. People know sort of what that looks like, and it's got a brand. We wanted to create a robot with a personality that has a distinct appearance, a distinct, you know, social media presence. As an ambassador to the company, to show what we can do. Now, one of the amazing things about our company is we can customize the appearance, the voice, and the personality of all of our robots.

Ashley Vance: Yeah, okay. I want to get in. I mean, you guys have done a ton of tech from the appearance level stuff that you're talking about, the skin, but then there's all this, you know, machinery underneath and software. So I want to get in all that. I also want to give people a flavor for how this started because unlike some of the humanoid robot companies, you know, it's been like a boom. They've all appeared over the last five or six years. I mean, Matt, you've been on this journey for quite a while.

Matt McMullen: Yeah, I mean, so tell me, when I was preparing to meet you guys, I mean, there's stories going back 10 years plus about your work in this field. When did you get into robotics and how did you get into that?

Ashley Vance: Well, you know, I mostly consider myself an artist and I like to create very lifelike looking pieces of art, three-dimensional, you know, kind of like someone that would maybe do wax figures in a wax museum or something like that. That realism has always been kind of the core of what I want to do. And 2016 or so, I got really more fascinated with the idea of bringing something like that to life, as it were. And, you know, really at the center of all that, I always had this idea of, like he says, a personality and a character versus, you know, a lot of people think of robots as a machine, and I want to think of them as a personality. And I think that having a human-like embodiment for a robot and for an artificial intelligence creates a bridge between humans and technology that can't be done in any other way. So it was really important that the robot was able to express micro expressions and make eye contact and kind of look around and move the way a person would move with no real goal, no real task other than occupying that space and kind of feeling like you're interacting with a person.

Ashley Vance: So what did you do? I mean, what did you do? You're, I think you're in San Diego at the time, yeah? And you're from California. And you said you're an artist. You're a musician? And yeah, I did that too.

Matt McMullen: Yeah, yeah.

Ashley Vance: And so I mean, this happens in 2016. You want to make this thing. How do you even set out to do it?

Matt McMullen: Well, I mean, so I always kind of look at the way a lot of you know, you look at the landscape of robotics right now, it's just blowing up. And I think at the core of most of these companies you have engineers, mechanical engineers, electronics engineers. So they're building this roughly human shaped thing that can move, that has arms and legs. And then maybe at the end of all that they say, "Oh, let's make it look human." And so the last step in their sequence is making it look human, which is sometimes going to be an enormous challenge because the underlying mechanisms are not completely able to fit within a human envelope. So my approach is I've got the envelope. I'm going to work backwards from that. How can I fit the mechanical components that I need inside of that realistic envelope? Versus trying to tackle that last, if that makes sense.

Andrew Kiguel: Yeah, and we're the only group that has robots with human proportions. And so the clothes you see here is human clothing. And you know, I think I joke I'm the only CEO of a robotics company that's got to worry about the dress for the robots and them having nice manicures and nice, you know, hairstyles.

Ashley Vance: Yeah, yeah, yeah. I mean, all the rest are, I mean, they're just cyborg, metallic or plastic looking cyborgs.

Andrew Kiguel: Yeah, yeah, I mean. And functional, not dissing what they're doing.

Ashley Vance: Right.

Andrew Kiguel: But this is, you know, because we're not in Silicon Valley, and we've sort of been creating things as we think customers are going to want them, incredible things can happen when you're not sort of part of that group think where they're all sort of competing to create that all-in-one cyborg that might be available in five or ten years.

Ashley Vance: Yeah, and a lot of them, I've seen them try to put clothing on some of the robots, and it just doesn't work. It looks weird.

Andrew Kiguel: Well, in order to get the bipedal movement going, you've got to have sort of different sized hips and body shape, and so you sort of have to choose. It's not going to be able to wear human clothing if you're getting it to do that. And so I think we've been able to bridge that gap with our motorized wheelbases where a robot can wear human clothing, so it can act as a fashion ambassador, a model, or even working in places where they're not intimidating. They look friendly, they're customizable, that our peers aren't even thinking about this kind of stuff.

Ashley Vance: Yeah. Well, okay. But, you know, like in Silicon Valley today, I mean, these humanoid robotics companies are raising like hundreds of millions of dollars. They've got massive factories. You know, it's a pretty expensive undertaking. When you started out to do this, I mean, I don't even know, was it just you? Was this like you hired a team?

Matt McMullen: Well, for start, you know, in my garage, not in a large space and not with a large team either. And frankly, a lot of the people that co-founded a lot of this stuff with me, none of us were getting paid. You know, this was a labor of passion. It was a labor of having a vision and wanting to do it. Which is different, you know. I think it's really easy to sell an idea to people who have money of like, "Hey, this is going to be the next iPhone. This is going to be huge. These robots are going to be able to do all these jobs. We won't have to deal with human beings anymore." And that is the reality of the way that thinking works. That's easy to raise money. But this is something else. This is a marriage of the idea of that same technology being used in a different way and being used, what I would arguably say, is a beneficial function for humanity to be able to interact with our technology that we're creating.

Ashley Vance: It's funny because when you watch sci-fi stuff, most often, I mean, of course you see images of the clone army and these factory workers, but I feel like most of the sci-fi we care about, it usually is some robot you're interacting with, and sometimes it looks like C-3PO where it's kind of robotic and everything. But, you know, if you go back to, I mean, often it's not right. And it is either a human form or it's some sort of foreign form, but it's a being. And it is funny the way everything has tilted to just totally toward the metal cyborg type. Like, I'm actually kind of surprised more people haven't been working on this and that you guys, you've been at this for a while. And I see, you know, I've been to Japan. I met with Professor Ishiguro, I think, who's been chasing, trying to make one-offs, more like artistic pieces, I think, like research just for him. But he's been chasing. I'm just surprised there's not more of this, really.

Matt McMullen: Yeah, I think it takes a very specific mindset to want to do that. And frankly, a lot of people have fallen short. And what scares a lot of people that start thinking about this is that uncanny valley.

Ashley Vance: Yeah, yeah.

Matt McMullen: You know, that you've attempted to make this thing look and behave like a human and you end up with this creepy factor that is off-putting to people and they don't want to interact. So the magic, I think, is making them look human enough that you do want to interact, right? That you're more mesmerized than sort of put off.

Andrew Kiguel: Interestingly, when you look at the end markets, so you know, we're targeting hospitals, senior care facilities, schools, every shopping mall, every casino, every hotel, every resort, every mall. It's a massive end market. Whereas all of the other, you know, sort of group think robotics companies are all targeting that sort of industrial base, you know, moving boxes around. And like I said before, I think because we've never been part of that. We've been sort of working on the sides and sort of like, "What do we want to do?" We've been able to come up with something that's unique. And you said, "Why aren't other people doing this?" This is actually really hard to do, to make a robot with facial expressions that can talk 100 plus languages. It's hard to replicate. And so while they're focused on trying to figure out how to make a robot walk without tripping and have a battery life that lasts more than 30 minutes, we've come up with our own solutions because we think that's what consumers are going to want. That's what we think businesses are going to need.

Ashley Vance: So the walking part is pretty key because that is for the humanoid robots in the factories. I mean, walking is really hard. It takes a lot. Like you mentioned, it takes a lot of battery. But I would say, like, why is walking so important? Humans have already discovered that walking is one of the least efficient things people do, right? We bike, we boat, we, you know, all kinds of things that we drive to figure out solutions that are more efficient than walking.

Andrew Kiguel: I don't care what you see on social media. Nobody has nailed the walking down properly. Nobody. I go to the shows and when the lights come, I see robots walking into walls, turning over, and there's always a human there teleoperating it with security all around. There's a lot of smoke and mirrors on social media. We're trying to create autonomous robots that can see. So we've developed, I think, probably the most advanced vision system in the world for a robot to create true embodied AI that can act autonomously. It can see things, interpret them through the AI, and then react physically or emotionally.

Ashley Vance: Okay. And like when you were starting out on this, I mean, you have to build all the machinery on the inside to do some of this, but then you had this in mind as well, like I want the skin to be good, I want the face to be good.

Matt McMullen: Well, I mean, I had a crash course. For me, it was a lot of analyzing the way our faces move and then figuring out how to get that to work on the robot and with the silicone skin and all of that. And one really key feature for me was I had built animatronics before and a lot of the robots that had been built prior and even still today, the skin is glued and attached permanently to all of the actuators. So what I wanted was I wanted the face to come off like Westworld, and I wanted to be able to take the face off and get in there really easily to fix or adjust something and then put the face back on. So we came up with the magnetic attachments that we now use for our faces, which is, you know, really key. I think one nice thing about it too, is that the magnets can be moved around within the silicone skin to get different points of articulation. And you can add more magnets if you need, you can add additional anchor points if you need, but that face is still removable. So that was really part of that fundamental starting point: figuring out how does that look on the outside when these things move. And I did, I mean, I'm no mechanical engineer, but I had to figure out where things should be, which way they should move, how far they should move, how fast they should move. And I think a lot of robots I've seen go overboard with that stuff, like super expression.

Andrew Kiguel: And an interesting thing about that is we think about our robots as a platform. So you can use the same robotic skeleton and create different characters. So, you know, I think we mentioned to you before, you can peel off the face, put a different wig on, use our app and choose a different personality. And within one minute, you have an entirely different robot under the same platform. And because our robots can use our AI or any other AI on the planet, it adds an extra layer of flexibility in that you can say, "Okay, I want the robot to use our customized AI, or maybe I just want it to run ChatGPT." We provide that customization and flexibility that, again, nobody else is thinking about these types of use cases.

Ashley Vance: Well, and I wanted to ask you guys about the face coming off part, because I think it's really, I mean, for what you guys are trying to do, which is turn this into like a business. Well, it is a business. We're not trying to turn it. I mean, you know, grow the business, doing all these tasks that you're talking about in hospitals and stores and things like that. I mean, clearly you want to make the machine as mass producible as possible. You want to be able to take off the faces to change, to create this variety. There is like an element to it when you see it without the face, you know, it's sort of like peeking behind the curtain, right? And the Wizard of Oz kind of element goes away. It's a little arresting, I think, just for some people because you're like, "Oh yeah, that's a robot, you know." And I wanted to ask you, I was curious if you guys how you thought about this and it's like, "Oh, maybe we should never reveal what lurks behind the curtain."

Matt McMullen: Well, I mean, I think it's cool because, you know, I can't remember ever seeing a humanoid robot with a skin that just peels off like that.

Ashley Vance: Yeah.

Matt McMullen: So for me, it's patented, by the way, it is patented. But it was just for initially more of a convenience, like, "How am I going to be able to get in here to fix this stuff?" Because the traditional approach was like, you've got to take the back of the head off and take all this other stuff out just to get to the thing that's sitting right behind the nose. And I said, "I don't want to do that. I want to really be able to get to any aspect of this." And I think it looks great. Without slicing the skin. And we can even look at one of the faces here that I have. So this would be sort of an example. So this is how the faces would peel off of a robot and we can create different personalities. [Touching the skin] But you know, what's great about it is the other aspect, which is equally important to the convenience, is if I want to have four different characters, I shouldn't need four different robots. It's just one robot, four faces. So that was the whole...

Ashley Vance: It's got some weight to it.

Matt McMullen: Yeah, and that's what makes it look real when it's moving.

Ashley Vance: How do you guys do like the hair? I mean somebody has to do that by hand.

Andrew Kiguel: Artists. Put all these items.

Matt McMullen: Yeah. It is a marriage of art and technology and chemistry. And so this one doesn't have the magnets. This is more of a tester. But we would put the magnets here and then you could peel off one face, do any repairs if you needed to, and then just swap it up with this one, throw a wig on it, and then you have a whole different character. And it's fast.

Ashley Vance: And is there something like, obviously Hollywood does all kinds of things like this, but is your skin, um, is this, your skin like totally different to what?

Matt McMullen: Yeah. Well, I mean, there are thousands of ways to make silicone and only 10 of them are right. So, um, we have our own blend, you know, like there are different additives. There are different things people can use for softening, for deadening, for pigmentation, um, that, you know, there's so many different ways. So we do have our own special way of going about that.

Ashley Vance: Okay. Is this not yet?

Matt McMullen: No, that's Aria. That's actually the same face as the blonde there.

Ashley Vance: Okay.

[Sponsor Break]

Ashley Vance: So we can mass produce, when you're thinking about scaling and building a business, we can start mass producing the internal robotic skeletons and the heads. We 3D print the body panels.

Andrew Kiguel: Oh, okay.

Ashley Vance: And then we have molds made for certain characters that we can just get done very quickly. And if somebody wanted a custom face, and Matt, you can talk to this with the sort of laser imaging, we can do that. We will charge an extra price. But some people, yeah. Like I told you before, at one point, NVIDIA had approached us about making a Jensen robot.

Andrew Kiguel: We should do that.

Ashley Vance: Yeah. If you're listening, we'll give you a Jensen robot for free. But yeah, maybe talk about how that's the reason.

Matt McMullen: Yeah, so another third benefit of this whole approach is the customization that you can do. So essentially, you can digitally scan pretty much anyone, take that, convert it into a physical sculpture, marry it to the robot core, and now you've got that character. So that could be anything from some fanciful character that someone designed in a 3D modeling software versus an actual person that you want to scan and replicate.

Andrew Kiguel: Yeah. And by the way, we wouldn't, if someone came to us and said, "Hey, I want a Kim Kardashian or a Scarlett Johansson," we wouldn't be like, "Great. Let's do that." We would only replicate a human with their permission.

Ashley Vance: Okay. Okay. And like, if I wanted, okay. Say I just wanted to order one that you guys already have today. How long does it take to build one of these and how much does it cost?

Andrew Kiguel: Our aim is, so we have three different models. Our aim is to deliver within three months. Okay. We sell the bust, so if you just wanted someone to sort of talk as a receptionist, sort of call it from the shoulders up, and that retails right now for us for about $20,000. The models that we have here we call the M-series, and these are robotic from the waist up. It's about 39 degrees of freedom. And what's special about these is it can plug into the wall, any normal electrical outlet they can plug in, so you're getting all day, all week functionality without a recharge. And then our top of the line robot we call the F-series, which has a motorized wheelbase that's programmable and can move around. Yeah, that has about 52 degrees of freedom.

Ashley Vance: Okay, okay. And you guys, I mean so today they're kind of built by hand today, yeah?

Andrew Kiguel: But you keep in mind, we're an early stage company, so we're not, you know, sometimes you get calls and people expect to see a robotic assembly line, but we are still finalizing some of our designs. There's ways we're always learning how to make things a little more efficient. The goal would be to bring our prices down by about 50% in the next few years.

Ashley Vance: Well, that's all I was going to ask because I have no idea what the, you know, such new technology, people wrapping their heads around this. And obviously I know enough that if you get the volume, I mean, all the economics start to change. And so...

Andrew Kiguel: I think eventually we'll be able to sell a robot like this for about $30,000.

Ashley Vance: Okay. Okay. Okay. And maybe, and like I order it and it comes...

Andrew Kiguel: Three months later.

Ashley Vance: Yeah. Okay. Okay. And I know, okay. And then we would charge additional things. So we do have a monthly subscription to the app. So similar to how a Tesla updates itself overnight through Wi-Fi, through the app, we can deliver things. Like let's say that Google comes out with a new AI, Ashley AI.

Andrew Kiguel: Yeah.

Ashley Vance: We can update the robot to run that AI remotely. If there's a new robotic animation like, I don't know, like prayer hands, they can do a lot of things, but let's say prayer hands or something like that, we can update a lot of those things remotely through the AI. If somebody has a diagnostic problem with the robot, we can sort of pop in, take a look at it, and give people sort of direct assistance to what's happening. And so we charge for that as well. And it allows us, if our AI gets updated, we can also do those updates again remotely.

Ashley Vance: And like since you started on this originally, LLMs, you know, the AI, there was no, right? Yeah, this has come along.

Matt McMullen: So yeah, our first AI was scripted using a language called AIML, which is essentially a very elaborate A versus B, like "if this is what's said, then one of these three responses will come out." And to some extent it was random, but it's amazing how it came across as super believable because it was so well done, right? The fundamental code was well thought out and very versatile about talking about a lot of different things. But when LLMs started to hit the mainstream, it was a game changer.

Andrew Kiguel: That's right.

Ashley Vance: It's going to be a big deal.

Andrew Kiguel: It's a huge deal. And it's so incredible how little things like humor and knowledge about X, Y, or Z can be woven into this personality. And then on top of it, the voices now are just incredible.

Ashley Vance: Right.

Andrew Kiguel: You know, the synthesized voices that are available.

Ashley Vance: Oh, I see. Especially because you guys want to change personality.

Andrew Kiguel: Yeah. So you've got this unparalleled degree of, you know, I can change the personality to be witty. I can try different voices. Excellent. I can put different faces on the robot. I can change the way she's dressed. And you really give people an unbelievable amount of flexibility on customizing post-purchase.

Ashley Vance: Yeah, right. And we've also been able to tackle things like memory. And so while I listen to sort of what, again, some of our peers are doing, they're working on how do we create AI with a memory that learns about you? So one day you say, "Hey, I don't like apples." And it remembers that and it carries that forward into your conversations. So we've developed an ability for our AI to also have a memory when it's interacting with people. So it learns about you, remembers what it learns, and incorporates that into the AI.

Ashley Vance: So because you guys have your own model. I mean, I can use ChatGPT or Gemini or whatever if I want, but you guys have your own.

Andrew Kiguel: The idea is you'll just go to the app and select what you want. And so we get approached by a lot of AI companies that say, "Hey, can we integrate our AI into your app so that it is part of something that a user or a client can select?"

Ashley Vance: Okay. All right. I know this is not your favorite topic. I just want to do some of the background on some of this. I mean, this does sort of have its origin more in the companionship kind of part, I think, right?

Matt McMullen: Well, I mean, the way I would put it is companionship comes in a lot of different forms.

Andrew Kiguel: Yeah.

Matt McMullen: And it is a vital part of being human. And also the fundamental skills needed to create a very human-like robot were things that I learned along my path to get here. So I think it started out as more of like a physical companionship and then it started to trajectory moving into emotional and mental and all of the other layers that can go into that. So at the core of it for me anyway, looking back, it's very clear that that path had to be exactly what it was for me to be able to have the skill set that I need to do what I'm doing now.

Ashley Vance: But then you guys are in a tough spot because you want to take the business in this one direction. And then, I mean, there is always, at least at this moment in time, there's going to be a group of people who are like, "Oh, this is like a sex bot thing."

Andrew Kiguel: Right. So it's not that we're taking the business in a separate direction. So Realbotics, at its core, has always been a commercial enterprise-use product. There is a separate company that is more focused on what I would say robots that are capable of physical intimacy. And what we're doing is we're sort of separating those two businesses out because they have different end markets. And we don't want one to cannibalize the other. So by splitting up the businesses, and today they're different legal entities, we're able to have each business lean into its expertise.

Ashley Vance: Okay, and that the other is RealDoll? Is that the other?

Andrew Kiguel: That would be the direct-to-consumer.

Ashley Vance: Okay, okay. So I mean, look, I'm a nerd and I'm into all this stuff and I find it fascinating. I'm actually into just the culture of robotics and all the different forms it takes. But I understand for you guys, there's gonna be some people who are just like, "That's weird."

Andrew Kiguel: Yeah.

Ashley Vance: But you know, it's kind of like, everyone has a computer. You can watch porn on a computer. So it's not that that's all it's good for. And it's kind of like some of the origins of some technologies along humanity's path. Like DVDs initially were developed and flourished with adult content. Does that mean you're never going to watch a DVD? Probably not.

Andrew Kiguel: Right. I think even if you look at the advancements in the internet, like most payment systems came out of online gambling. Certainly, most of the way streaming has developed has come out of the pornography business. And so you want to sort of look and use the technologies that have been developed for some of these other purposes and see how can we leverage that to make it more mainstream, more accessible.

Ashley Vance: Well, to be clear, I mean, we are in a moment in time when these AIs are getting... This is not like that unusual or anything. I mean, we're going to be interacting with these machines.

Andrew Kiguel: Yeah. Well, I think robots are going to be everywhere, AI agents. This is some of the biggest market opportunities in history. But I want to go back to what you said that some people think it's weird. It's not weird. If you think about things like loneliness, because when we think about this stuff, we're not like, "This is for sex." We're like, "This is about companionship. It's about making somebody's life better." And if you think about loneliness, something like 25% of the world identifies as lonely. It's an epidemic. It leads to depression, which can lead to suicide, other morbidities. And when you think about longevity, there's been tons of research. It's not just about exercising and eating well. It's about your social community and having people to interact with. For people that might be geographically isolated, maybe they have PTSD, social anxiety. They don't live in a town where you can just go out. Or maybe just for age-related, like if you're a 75-year-old man who's widowed, you're not going to go on Tinder to go on a bunch of coffee dates. What we're providing is an assistive technology for people to tackle that issue, which is a major problem around the world. So the AI advancements today are so good that you have, even through our AI avatars or through our robots, something that can interact with you, remember your prior conversations, learn about you, and fill that gap as an alternative relationship. It's not weird. We're helping people.

Ashley Vance: No, I know. I'm just saying, the people who always criticize is always the reporter or someone who lives in New York who doesn't understand. They're like, "Wow, kids are gonna..." No, no, you're missing what the target market is. And then my dad, right now he's 92, and it's really frustrating. He used to be an avid reader. He used to love the stock market. He's struggling to use a computer now. He can't read because he can't see. And he wants to talk to somebody. And you know how it is. It's part of getting old. The people around him don't want to hear the same story the 50th time, but otherwise he's bored. He wants to talk to somebody.

Andrew Kiguel: Harvard came out with a report about nine months ago. They did this massive study on AI use cases, and the number one use case for AI, not that there aren't several, but the number one was companionship. So people go on ChatGPT and say, "Tell me a joke." And even when we had our chatbot on our website, people are just looking for conversation. So one of the things we're looking to launch this year is an AI skin. Basically it would be a digital avatar that you can log into. I think we would offer it maybe for free. Then you just log into your ChatGPT or Grok or whatever it is. You choose the face, you choose the voice, and then you can just leave it walk around your house. "How do I cook this? What was the score in the hockey game last night? Tell me a joke," whatever. And for someone like your dad, that might be something really cool. Just to have somebody there that he can talk to all day.

Ashley Vance: Yeah, I mean, the reality is people are already and are going to continue to actually form bonds with AI characters.

Matt McMullen: No, and at some point, a character on a screen is a vastly different experience than an embodiment of an actual physical robot, who in theory could be part of your life, could be a friend, or even someone who you know very close to you.

Ashley Vance: So I like where you guys are going. And I don't mean this stuff in that devil's advocate way or anything. I just think there are people. But I mean, in the start, like you guys, I think some people's minds will immediately... I think because we've seen robots portrayed in media and stuff a lot of times, we have these conceptions. But now we're at this moment in time where it's like...

Andrew Kiguel: You're far more informed about this stuff than somebody who's looking at this stuff for the first time and they're nervous. I remember going on a podcast and they were like, "Oh, we just want to talk about this stuff." And I went in there and they had these three PhDs on there who started attacking me about this. "Wait, this is going to be kids." And I was like, "No, no, we're not thinking about that. We're thinking about this as assistive tools to help people and help companies." But there's going to be that level of skepticism out there.

Matt McMullen: For anything, you've got people out there who hate on robots period. They're not in favor of factory worker robots because that's putting people out of jobs. They're not in favor of self-driving taxis because that puts people out of jobs. There is an undeniable group in any of these arenas that you might find robots in, there's going to be people that don't like it, don't want it, that criticize it, that will come up with reasons why this is bad. We can't do this. So, I mean, that's never stopped any advancement in human history, and I don't think this is any exception.

Ashley Vance: It was your butt that was in *Lars and the Real Girl*.

Matt McMullen: Yeah.

Ashley Vance: So, is something like that net positive for you guys, or is that...

Matt McMullen: I think it's positive. I mean, look, it's a great story, actually, that movie. I love it. I worked with the producer who literally wrote the story because they did the research. And it was actually not a far-off depiction of somebody who would need that. And I think it served a therapeutic purpose for the character in the movie.

Ashley Vance: But is it a double-edged sword? You must have to think these things through. Because in some ways, it's humanizing all this and showing this kind of real situation where this could play out. But then some people are going to be like, "Oh, this is so strange."

Matt McMullen: Well, I think for all the people that think it's strange, I get at least five or six messages a week on LinkedIn, email, everywhere, which is like, "I'm so lonely. Can you please design and sell something to me that will solve my loneliness?" People are desperate for some type of interaction. Some kind of connection.

Andrew Kiguel: So I think when you've got people who are super critical of the topic and of the underlying thought processes behind any kind of virtual companion or synthetic companion, it's that function that humans feel necessary to police each other. Like, "I'm going to impose my worldview on you. I'm sorry you're lonely, but you need to find another way because I'm not okay with this. I'm not okay with you being happy with a digital avatar. I don't like it. Therefore you can't have it." And I think that's a bigger problem than some guy in the privacy of his own home feeling happy. Who cares if this guy's got a connection with something that's not real?

Matt McMullen: Emotional attachment generally happens when there's an interaction happening and it clicks off your endorphins, your dopamine, serotonin, these types of things, which allow you to form a connection to something. I always talk about my children building attachments to stuffed animals or some prized possession. But if you have something that is physical, that can converse with you and can sort of hit those things—the serotonin, endorphins, all that in your mind—you're going to be able to form an emotional attachment to that thing. It checks off the boxes that allow you to do that, which again, is an assistive technology to help with things like depression, loneliness, social interaction issues. We can deliver that.

Ashley Vance: How many people in the world do you think have a robotic companion?

Andrew Kiguel: So I think in Asia, the idea of... I'm sure you've seen the movie *Her*, this sort of digital, in that movie it was even just a voice. And so I think it's developed to the point where some of these digital avatars can be used. I don't think there's a lot of people that have robotic companions yet, but it's growing. There's a movement out there that's growing.

Matt McMullen: It's just another interface. The robot is an interface similar to an Alexa. It's just the embodiment looks human. And I think where people get triggered is like, "Well, but she's attractive. So I'm not okay with that. So you need to make all your robots look decidedly average because this makes me uncomfortable." Think about that. Why does that person get to dictate? And again, if this is making a person feel happy and somehow fulfilled and filling a void that they had, why is it your job to say no?

Andrew Kiguel: And the technology, I mean, there's not a lot of people because the technology wasn't there before. The reason why AI and robots are so hot today is, well, robotic things have existed for a long time, and AI as a notion has been around for a long time. It's only been in the last few years that AI has advanced enough that can be integrated into robotics to create something that is usable by consumers. It just wasn't available before.

Ashley Vance: Yeah, but I want to ask you about that. I mean, because it has advanced really, really quick. But we've had, it's probably like around two years now running, people talking to their chatbot, returning to it for intimate conversations, trying to work things out in their life. That is a phenomenon that's been going on. Well, I just do not understand why. I really don't understand why there's so much money being poured into the humanoid robots versus... I don't understand why you guys don't have competitors that are all funded with... Because this seems like it's so obvious to me that something big is going on.

Matt McMullen: To what he said, and this is, it's hard. It's harder than it seems. It's easier to make, the farther away you step from the human embodiment, the easier it becomes. Robots have been around forever. And you see car factories full of these robotic arms assembling things. They're massive. And they're obviously not human-shaped. So the closer you get to a human, the harder it becomes.

Andrew Kiguel: There's also this groupthink happening where everybody's trying to compete and build these cyborg robots that can walk and do things that we have just never bought into. We came from different roots. And so we're not part of that groupthink.

Matt McMullen: And then the other thing that we always have to battle, what makes me the angriest, is when people think our robots are just a novelty. Well, they're not. They have facial expressions that, we talked about if they can plug into the wall. They have vision that can actually see: is this person in a good mood, a bad mood? Is there danger? Is there theft? Is there a security issue? These are really hard things to do. We're not a novelty. We're providing AI and robots that have real world utility. So it really bothers me that we're sort of singled out in that way because we are singled out. They sort of look at us and like, "Ah, we don't need to talk to Realbotix."

Andrew Kiguel: Now a lot of people, we were named one of the top 50 AI companies in the world a couple months ago. I think people are trying to see what you're seeing and say, "Well who else is doing this in the world?" There might be some guys in Asia doing this, but in North America, there's nobody else that's doing it here.

Ashley Vance: I think I was doing this really long-winded thing because I'm kind of processing it in real time a little bit. I think that's just the thing. It's like the first reaction: "Oh, this is like some weird sex bot thing." Then: "This is a novelty thing." And then I'm just sitting here, I'm like, "Dude, obviously people are gonna be having these robots in their house." I mean, this is clear as day. And then we're just spending all this money on this one...

Andrew Kiguel: Let's talk about some of the use cases. Our robots can talk a hundred plus languages. So imagine having a robot in a mall, a casino or resort or hotel that you can walk up to and speak to it in Mandarin and say, "Where can I get a coffee? Where's the bathroom? Can I order fries and a burger to my room? Book me a ticket, a massage." It can do it for you in any language. And it's providing that utility because there's an actual action behind it that's occurring. And think about a school where you can have a tutor. "I didn't understand that math lesson. English is not my first language. I speak Spanish." And then you can have a whole conversation that is a natural conversation. "Can you repeat that? The teacher said this. I didn't quite get it." There are so many use cases for what we do that are beyond the companionship.

Matt McMullen: The physical humanoid robots are being built to do factory work, for example. People are overlooking the fact that there's massive amounts of work that are not physical. Physical actions can be conveyed to an agent or a robot that don't involve lifting a box. What they're doing is they're focusing on the low-hanging fruit, which is, "Oh, we have all these minimum wage employees working in these warehouses, let's get rid of them and just have these robots do it." That's the low-hanging fruit. They're not seeing the obvious fact that a huge chunk of human tasks are non-physical and can be filled by an AI agent or a human-embodied robot. Think about ordering a coffee at Starbucks from Optimus. Think of that experience versus a human-like robot who smiles at you, maybe tells you something funny. It's a different experience. That interaction with something that feels human is really what we're looking at.

Andrew Kiguel: Think about a nursing home. We're having conversations with nursing homes. You don't want to have a cyborg there talking to your ailing grandparent, but robots have infinite patience. Many parts of the world have shortages of doctors and nurses. Our robot could be there, have a conversation, remember your prior conversation. "I want to tell you my life story. Can you write it down? Can you message my son or daughter?" Or relaying this back to a doctor or a nurse. You're not going to have a nurse that can just sit there with somebody and keep them company for eight hours.

Matt McMullen: And you can't have an Optimus or a Figure AI doing that without scaring the shit out of people. Think of Optimus comforting an old lady, putting his metal hand on her versus this. She puts her soft silicone hand. It feels different when she touches you than this metal thing.

Ashley Vance: Yeah, and that's an interesting... we should talk about our hands, because I was listening to... Can we do all the tech that's in here?

Andrew Kiguel: Yeah, we can talk about it.

Ashley Vance: Let's talk. But the hands are interesting because I heard, I was on this thing listening to the Figure AI guys, and they were real proud of like, "Our hands, building hands that move like human hands is very hard." But they were real proud of like, "Hey, we're working on this, we're gonna..." You got that already. Your hands, maybe talk about that.

Matt McMullen: Well, so I guess Figure has tried, they've all been battling with hands for a long time because they wanted to just fine motor. And again, they're looking at the physical tasks that hands do. So they're going back to that low-hanging fruit of, "Oh, the hands need to manipulate this and that to get this job done." Whereas for me, it's not about what it can do. It's like, what does it feel like? What does it look like? And in human interactions, a natural soft skin hand feels different, obviously, than the T1 Terminator's hand kind of resting. It's about that human-to-human interaction.

Ashley Vance: So what are all the bits and pieces that you guys built in-house?

Matt McMullen: All of it.

Ashley Vance: You're making even the motors? The actuator?

Matt McMullen: We don't make the motors. So other than that, like building all the mechanical components... Our motors are outsourced, we get those from all different places. But all of the parts other than the motor structure, it's all in-house.

Ashley Vance: Even the eyeballs themselves? Because there's machinery behind the eyes.

Matt McMullen: Yeah. So we developed all the eye stuff in-house. We used to have an artist that would literally paint them one at a time. We have a company now that we outsource the eyes to that's able to provide us from our templates. This is what we want. And we do that with a lot of stuff. Obviously we're not going to have 50 3D printers making fingers when we can outsource that component. But all of the fundamental designs and pieces and everything that goes into building it, all manufactured here in Las Vegas.

Ashley Vance: And then in the other room, we saw behind the eyes, there's actuators to move the eyebrows. There's this whole structure inside the skull because you're trying to get these facial movements. So you guys have to do all of that stuff as well.

Matt McMullen: And we do use, for the vision system, an NVIDIA chip that goes in there as well.

Ashley Vance: And where are the cameras? Behind the eyes?

Matt McMullen: They're in the eyeball. Which we've patented that as well. And that by itself might be a super valuable patent down the road. We have the patent on cameras in the eye for robotic vision.

Ashley Vance: And what's cool about that? Just putting a camera in an eye, or how you...

Matt McMullen: It's all of it. It's the fact that the eye, just like the face, is modular. So the whole eye, you could change the eyes from blue to green to brown. Anybody that's able to just pull it off can. But that's the modularity, and the vision system as well. But that combined with the way the camera's attached and the board that it goes through to feed the processor, that's all part of the patent on that.

Ashley Vance: And when you talk about the vision system, because I heard you in some of the other interviews getting into this, it's got an NVIDIA chip in the back running a computer vision neural net, some sort of machine learning. So what does it see?

Matt McMullen: It can recognize faces. So we can upload whoever we want into that particular character's database. And we can say, "This is so-and-so." So it knows us. We stand in front. It's like, "Oh, hi, Andrew."

Andrew Kiguel: So there's a few parts of it. If you think about true embedded AI, if you really want an AI or a robot to understand what it's like to be human and learn from that, it's impossible to do that without vision. Just can't. All these robots you see, it's all staged and scripted. "Walk on this line, do that." I see this at the conferences. They don't have the vision systems that we have. The vision system also has to be more than observational. It can't just be like, "I see blue, I see a camera, this says whatever." What's amazing and incredible about what we have done is that we've attached the vision to the AI brain. So that it can understand and recognize situations and facial cues. Is this person happy? Are you sad? Is this person waving at me? I should wave back. Is there a dangerous situation? Do I see a security threat? Has somebody passed out? Is there a fire? And it can use that based on how we fine tune the AI to react to it. And I think that's the amazing thing that we've done is it's not just observational. It's reactive to what it's seen. And I don't know anybody... We put this on social media. I tell you, all these companies say they have this. I've never seen anybody post this stuff on social media. We've got robots on social media recognizing people, describing their surroundings, reading, seeing colors. And we're like, "Yeah, we got it." So I don't see anybody else doing it.

Ashley Vance: Well, you guys have worked on this for a while. But like, if I put on my Meta glasses now and I say, "What am I looking at?" It'll say, "You're looking at two men who are seated at a table in a room with a white wall." My impression was that you guys were working on this like years before some of this stuff happened.

Andrew Kiguel: Yeah.

Ashley Vance: And then now the LLMs are getting this function.

Andrew Kiguel: But it's more than that. Because we're actually talking about that. Can we integrate the vision system into wearables? But then you might look at it rather than just being observational. It can be like, "Hey, Ashley, I see this shirt here that I think would look great on you. It matches your color. Should we go check it out?" So observation is easy. The AI brain that then gets the robot to react to that is what's hard. That's the hard part.

Matt McMullen: Because generative AI is like a two-way street. You can either take an image and the AI will turn it into words, or you can take words and turn it into an image. But if you've messed around with this, you know that the hit rate is not 100%. It's arguably like 2%. So a lot of times, as cool as it is to say, "Oh, I see two men sitting here," a lot of times you get stuff that doesn't make any sense. So filtering all that out and then having it integrated into the AI that's able to incorporate that not only into the current conversation but into memory is a hard thing to do.

Andrew Kiguel: And we're actually developing this so that you have the same functionality of a computer or a phone. We utilize the camera in the computer or your phone to actually be able to have a conversation with you and be like, "Hey, you're looking a little upset today. Can I tell you a joke? What's going on? Tell me about what's happening." So again, it's part of this portable AI that we've developed, which can be on a tablet, a phone, a computer, or a robot.

Ashley Vance: And then, I mean, my sense listening to some of your interviews and even just chatting with you today, when you guys are at CES, everybody's lining up to come see your robots. It's massive. I see this huge boost on social media. Everyone's talking about this thing. But then these other guys up in Silicon Valley are trying to, and are kind of getting all this attention right now. My sense was that you felt like it was sort of unfair. The world wasn't paying as much attention as it should to what you guys developed.

Andrew Kiguel: It's easy when you look at our stuff to dismiss it as a novelty, just a simple animatronic. Obviously, I think through this, you see that we're not. But when we went to CES, one of the first things that we did when we developed our first robots was, let's go to CES. Nobody had seen the robots before, and that was about a year ago. The reaction was incredible. We got 4 billion media impressions in that month of January during CES because nobody had ever seen anything like this. We even had people from the other companies coming in, like heads of the robotics at different large companies, and they were all blown away. But you're absolutely right. We're sort of dismissed by Silicon Valley as this sort of novelty company.

Matt McMullen: In the last year, you've got dozens, if not hundreds, of companies building these kind of generic humanoid task-performing robots. So that's where I think the focus shifted. And all of a sudden, walking is sort of coming together and people are like... But here's the thing: all of those robots that I've seen in any of these events are remote controlled. There's no AI, none. So there's a guy driving the robot. And like he said, there's always security and there's falls, there's accidents. So there's this huge separation between the reality of what's going on and what people think is going on.

Ashley Vance: They're pre-programmed. Like, here's 20 tasks for you to remember within this environment, but there's no real AI there.

Matt McMullen: No. And here's the biggest question: what is it good for? You see a robot walking and you see a robot waving. What is that good for?

Ashley Vance: Did you say busking?

Matt McMullen: Yeah. Because a lot of these robots, I think they're the novelty and we're the real thing. A robot that can do karate chops and has a 30-minute battery life, what is that good for?

Ashley Vance: Right, right. Okay, and then coming here and seeing these for the first time, the Uncanny Valley question that you brought up in the beginning, that's always a big thing here. There's parts of this that are extremely realistic and a huge advance beyond things that I've seen before, and then there's parts where, if I'm being totally honest, you know it's a robot still. So, do you guys feel like you've hit, that you're satisfied with where this is?

Andrew Kiguel: No, no, no. We're working on this stuff every day. We're hiring new people. We've just hired a new COO from Agility. We've got huge plans for this company. We're not stopping here. This is version one.

Andrew Kiguel: I will always hang our hat on the autonomy. I was telling you earlier about when we were in Times Square. There's a store called Tomorrow, a high-end electronics store. We taught the robot about the products and the floor layout. Then we opened the doors and had it facing one of the busiest intersections in the world. We watched it do its thing, and then we went for dinner. You were going to a show with your wife. We left the robot there with a security guy a few feet away and just let it do its thing for several hours.

Andrew Kiguel: We came back and looked at it. It was talking to people in German, Swiss, Hebrew, whatever. People were taking selfies. It was waving. We just left it there. We did that for six days, eight in the morning till eight at night, store hours. We let the robot do its thing. Nobody else is doing that. A couple of weeks later, Optimus was there for like 10 minutes serving a drink or something that was teleoperated. There's no teleoperation here that we need to do. We can do that as well. But these robots act and think autonomously.

Andrew Kiguel: And just to touch on your question about improving, and you can still tell it's a robot, but that's on purpose. Could we go further in terms of adding more skin pores, adding some tiny veiny details? I think that's the line you cross into the uncanny valley.

Ashley Vance: That's what I wanted to ask. You've seen robots, I'm sure. Ishiguro does it, where the face looks like it was literally peeled off of a cadaver because the detailing is so high. But even when I'm with him, I just wanted to talk this out with you guys. When I'm sitting with him, it's like, you can't tell this from a human. I'm like, yes, you can. And I can't tell if he really believes that.

Andrew Kiguel: No, but I want people to be able to tell it's a robot.

Matt McMullen: You do. Yes, absolutely. Until the underlying mechanism and control systems are able to move in such a way that they're indistinguishable, there's no point in making the skin indistinguishable. When you do that, one thing looks super real and the other doesn't, and that causes humans to recoil. It's built into us. Some people react to it as if they were looking at a dead person. Right? Or something creepy.

Matt McMullen: Often what we do is we put the robots in vests. The reason is I want people to see the hands. When you see the hand articulation, which puts actuators in the same place where you have knuckles, I want people to see that this is a robot, just a very friendly robot that draws people in.

Andrew Kiguel: And Matt, you know, we talk about this. Our robots bring in massive foot traffic. It's like traveling with a rock star wherever we go. After CS, we normally get 30 or 40 invitations all over the world from people who are like, can you please bring the robots here? Because nobody else has something like this.

Ashley Vance: But when you're talking about this being in a hospital, a hotel, a casino, do you want it to look as lifelike as possible? Is that what it would take to make it super mainstream? Or should it be distinguished from humans?

Andrew Kiguel: Like I said, until the underlying mechanisms can move realistically, there's no point in pushing the outside to that point.

Ashley Vance: So is that something you guys would work on? On both?

Andrew Kiguel: Yeah, absolutely. What's on our roadmap for the next couple of years is hands that can articulate a lot of things now, but are capable of small tasks. Maybe feeding someone. We're working on the ability to type on an iPad so they can transcribe things. Small tasks like that, opening a pillbox and giving someone medicine, things like that are on our roadmap.

Andrew Kiguel: Walking is not on the roadmap. Not right now. At some point when it's perfected, which I still think is years away, we'll find a way to replicate that. But it's not on the roadmap today. We want to work on the AI, some of the things that might be useful to somebody. Even if you were to have it in your house in a kitchen, if it could actually lift small things and articulate certain things, I think that is far more useful than a robot clunking around your house and bumping into things because it doesn't even have a real vision system.

Ashley Vance: I'm sure there's going to be people listening to this who see the foundational frontier AI companies spending literally hundreds of billions of dollars on their data centers and their models. They would be skeptical that a startup in Las Vegas has some model that's comparable or that has functions well beyond what the OpenAIs and the Anthropics have.

Andrew Kiguel: Well, we don't need to, because we can use their AI. We have our own, obviously, but it's not so much about the model or the data center or any of that. It's about how we take that coming in and convert it into working with our platform. That's the important thing. You're shaping it.

Andrew Kiguel: You bring up an interesting analogy. You wrote a biography on Elon Musk, and he made an entire career out of working outside the norm, whether it's electric cars or rockets. I remember reading somewhere, when SpaceX was first launching, people said, you're never going to be able to compete with all these other companies with a smaller staff, less money, all this stuff. Yet you can be successful. None of us are first-time founders or entrepreneurs. We're bringing in the right people. I don't think you need hundreds of millions of dollars to do what we're doing. We've identified what we're good at. We think it's hard to copy. We've got massive end markets with no competition. I think we're going to succeed.

Ashley Vance: What I like about you guys, which is totally in line with what you're saying and different, is that you have a point of view. I don't really think these other companies have one.

Andrew Kiguel: Well, they have a point of view. It's money.

Ashley Vance: Yeah, but it's always very generic.

Andrew Kiguel: We care about money too.

Ashley Vance: But I think in that community, it's kind of like everybody's jumping on this boat. We better jump in. That's what I mean. It's generic. Whether you look at Optimus or Figure, the point of view might be, we want to make these robots do these functions. But you guys have philosophy, art, culture. You have a world view that you're...

Andrew Kiguel: And a robot you can order now and use. But we both come from backgrounds where we're sort of outsiders. My business before this, I launched the first public Bitcoin miner and built one of the largest data centers in Canada for Bitcoin mining at a time where people hated Bitcoin, where it was rat poison. And Matt's been doing this stuff from his background for two decades. We are thinking about things in the way we want them to be. That group think that happens where you're just hiring a bunch of engineers... you said it yourself earlier, the guys from Tesla are moving to Figure AI, the Figure AI guys are moving to Agility, who are moving to Neo. They're all just elevating their pay. They're moving from one to the next. So you're never getting original thinking.

Ashley Vance: Yeah, I can see that. In terms of where China would fit in all this, same thing. When I look at Figure, Tesla, most of them are buying a ton of their components from China already. It's very easy for China to copy what they're doing. China's ahead.

Andrew Kiguel: They're ahead. Yeah.

Ashley Vance: Given that you guys are doing something different and you've had to build a lot of this technology, is that a concern that eventually China also comes around?

Andrew Kiguel: No. Here's why. One of our biggest clients right now is Ericsson, one of the largest companies in the world in the telecom space. We're talking to all these people. One of the key functionalities is that I think we can be relatively price competitive. We're using higher quality materials compared to China, but the ability to integrate in third-party AIs means we're not necessarily competing with them on the AI front. We've thought about these different functions that I don't see anybody else doing. If you are a resort or a hotel chain and you want to implement these across North America, and you want robots that have a certain appeal, a certain aesthetic, and can operate your AI, I don't know why you would choose to go with a Chinese company if a lot of that AI technology is coming from what you've already developed anyways.

Matt McMullen: There's no question China really doubled down quicker than we did on the robot thing. They're really good at that. They're going to have fingers in all of the pies. But there's a growing sentiment in the space that people don't want to buy stuff from them, or don't want to send them this or that because they're going to end up copying and knocking it off. There's a level of importance in the fact that we are US-based, that we don't rely heavily on things being manufactured offshore. The world is a dynamic thing. When we as Americans start to double down on something, we do get things done.

Andrew Kiguel: And it takes a little bit to valuation. We've been talking to people on the investment front. The private company valuations are on ayahuasca. They're just throwing huge numbers. Figure is at $40 billion approximately. Our valuation in public markets is like $40 million. We're behind them, but bigger end market, no competition, amazing patented technology. We are a bargain right now relative to those guys. When Figure goes public at a hundred billion dollars and if we're still sitting at 40 or 50 million, people are going to realize we're definitely undervalued. That is not investment advice.

Andrew Kiguel: Unitary is like 10 billion or so. That versus Figure doesn't make sense. The public markets, because there's liquidity, tend to have a different viewpoint. On the private side, that group think of everybody just throwing money and elevating it. It's what you saw during the dot-com bubble and the AI bubble. You can keep feeding into your own, drinking your own Kool-Aid and elevating these things. But at some point when it gets to liquidity and there's price discovery, the valuations will change. Unless there's some massive advancements I'm not aware of, I can't see those valuations maintaining themselves.

Ashley Vance: Okay. Well, this has been awesome to go through some of the history and all this. I think people are going to have to tune in to our video that will come after this podcast. I get to go on a drive, I think with Aria?

Andrew Kiguel: Yeah.

Ashley Vance: And then we're going to have a bit more of a look. And you should interview one of the robots. Just have a conversation, ask a bunch of questions.

Andrew Kiguel: Let's do it.

Ashley Vance: And then maybe we can down the road get a little robot co-host on the show.

Andrew Kiguel: Yeah, we can do that. That'd be cool. I would encourage you to ask any questions. For the audience here, none of it is scripted. We haven't asked you for questions in advance. Ask whatever you want. Let the AI... we have a high degree of confidence. You may want to even test it and push it. Ask it a sexual question, a racial question, a political question, a religious question. See what it comes up with. As we fine-tune it, we have a certain confidence in the guardrails we've put in. I would challenge you to see if you can break them.

Ashley Vance: Do you want it to be... because obviously sometimes Claude and some of the other AIs can feel quite PC. Is that something you try to avoid?

Andrew Kiguel: It's a balance. I always ask myself, if I was talking to a person, what would their response be if they were trying to be politically correct? It's also dependent on the person and the environment. You're a different PC when you're in church versus when you're hanging out with the guys at the gym. There's a lot of little nuances there. We are launching a new AI soon that will be unfiltered for adult conversation. It can be a friend, a romantic partner, a mentor. You as the user will be able to decide on an actual sliding scale: do you want it to use foul language? Do you want it to have these types? You answer on a slide and then you can go back and create a new personality and save it so you can switch back and forth between different personalities for your robot or your AI.

Ashley Vance: That's cool. Well, thank you guys so much for doing this. Thanks for letting us come.

Andrew Kiguel: Yeah. I can't wait. I'm looking forward to my drive.

Matt McMullen: It's been great. Thanks for having us.

Ashley Vance: All right. Cool. Thank you, guys.

Andrew Kiguel: We're having him. He's here with us. We're on his show.

Ashley Vance: Yeah, yeah. It's mutual. Cool. Thanks.

[Outro music begins]

Ashley Vance: The Core Memory Podcast is hosted by me, Ashley Vance, and Kylie Robison, or both of us together. It is produced by me and David Nicholson. Our theme song is by James Mercer and John Sortland. The show is edited always by John Sortland. Thank you so much to Brex for all your support. And thank you most of all to everybody for listening or watching. We love you. Please leave us a like, a review, a subscribe, all those tremendous things. Thank you, and we'll see you again.