The Exchanges

Q It's like life size, right? Like, five foot six, two arms, two legs.

A Yeah, it's, uh, it's got, yeah, two legs, two arms, has hands. Um, we, we have a head, but it's not, like, we're not trying to mimic the human The human bodies. We don't have like eyes or a nose or ears. We have a screen. We have a lot of sensors like cameras and other things in the head. Um, the goal for the, for the form factor and the aesthetics is just to be a tool for humanity. So we want We want to, we want to deploy these into the labor force to help people. And I think one of the larger deterministic things that I, that I realized when starting this business or reason why I did this, we're having this huge labor crisis in the world that is really not well reported. Uh, so, you know, demographically the, the baby boomers are retiring and we haven't had as many kids as possible as we, as we wanted to probably. And so the, the, the amount of people in the labor force is basically flatlining. And we'll start to shrink worldwide. And this has caused now what? Eleven million jobs in the U S that are just unfilled today. Um, so we're walking into clients in retail or manufacturing or warehousing. The questions are not like, how are you going to replace our people? Like how are we going to roll? It's like we have. 85% of our employees show up to work today. 85% will show up to work tomorrow. We are seeing two and a half percent weekly attrition. We have nobody that wants to wal…

AI assessment note: “Yeah, it's, uh, it's got, yeah, two legs, two arms, has hands.”

Q there are a lot of assumptions that the industry is making that aren't necessarily the right assumptions. And because they're making those assumptions, they don't believe kind of what you're doing or the, the ability to get robots to perform human life tasks to be possible. What are a few, like what you would consider like big dangerous assumptions that the industry has made that you believe to be incorrect?

A Right off the bat. I think one is, there's this assumption that, I guess this is a debate on, should we build a humanoid that can do a lot of different things, meaning like one product can be put in to do thousands or not, if not millions of applications, or should we look at a problem, say like, um, a problem in a warehouse, a problem in retail, and just build a special purpose, like our special built robot. And I think most of the world, given how hard humanoids have been development-wise, feels like the world should be all purpose-built robots everywhere. And I think the assumption there is, is pretty flawed. I think it's really hard to build any robotic company, whatever you're doing, robotic vacuum or humanoid. It's just hard. You know, like a lot of people, a lot of capital, you need years, you need, uh, to develop a lot of, like, basically a continuum of product iterations that are getting better and better. And then, um, and then the problem there is you're building one product for one application. That application will never be done. A product will never be manufactured in high volumes, which is hard for cost. And you're never really going to build an AI data engine on one product. So you, you really want a general, you really want the generalizability of a product, like an interface here in robotics. You really want an application or a robot that can do many applicati…

AI assessment note: “feels like the world should be all purpose-built robots everywhere. And I think the assumption”

Q guide on what's actually priority for the founder versus what they talk about. If, I don't know if you have your calendar in front of you right now, but like, if you looked at your calendar today or tomorrow and you have to describe to me, like, what are you actually spending your time on? Like, I'd be super interested in like, what are you actually doing today and tomorrow?

A Yeah. So we start every day and we end every day with a standup on the engineering side next to the robot. And so we're, we're trying to So our, our one year anniversary is coming up for the business. We're going to try to walk the robot before that, which has never been done. That's basically one, like less than a year, uh, getting basically the, the whole team together, getting the vision for the company, like getting the funding for it, like building the product to a point where we're demonstrating a full humanoid system doing walking. It'd be the fastest I think I've ever heard of in the world. And so we do a stand up in the mornings on the engineering side. We have a detailed list of everything we need to do. Over basically like three big milestones. So it was basically getting upper body set up, brought up, lower body set up, first walking, and then we got some manipulation goals as we progress. And we, we end and begin every day with that stand up, talking about the problems, how to fix them. Do we change strategy? Do we bring up this test hand first or whatever it looks like? And that's a, I'm there. All my leads are there. The whole company's over there. Uh, and we're about to kick that off here in a minute. Um, and then throughout the day, all my meetings are product and engineering related. As I've As I've gotten like more mature in my career, I've realized more and …

AI assessment note: “we start every day and we end every day with a standup on the engineering side”

Q and I see a Uh, human, AI, uh, humanoid robot business, and I, the question I ask myself is you talk about being driven by having a large impact that when you're 80 years old, you look back and you're like, damn, I did some, like, pretty awesome work that helped the world. How do you figure the, the through line of these three businesses that plays into that motivation?

A Yeah, I mean, I think there's some, like, um, escape velocity needed for any entrepreneur to get into hardware. I think that's just, like, really tough. Like, I think ever since a kid, I was like, I wanted to work in hardware, wanted to work in more challenging areas of AI development, and I feel like, for the longest time, I always felt like I needed, like, a substantial, you know, capital injection personally. I needed, like, a long runway. I needed credibility to raise a lot of capital, to do big partnerships, have, like, a big group of people join me. Um, so like when starting vetery, I, I ultimately really excited about how much time we spend in the workplace. Like outside of spending time with loved ones, you're, we're at work most of our lives. Not everybody has the most Enjoyable experience at work. Looking for a job is, is just like, uh, it's a terrible experience. Like it's so hard for both sides. Like if you're employer trying to hire, if you're like a job seeker, it's like applying to a black hole. It's hard to, I guess, hard to know who's hiring. It's hard to get in there. It's, you know, generally historically has been very, like not been a very, uh, it hasn't really been a meritocracy in so many ways. So ultimately wanted to work on Kind of that labor problem when starting, you know, when starting Veteri, I think it's like, was, is and still is a really important…

AI assessment note: “I think there's some, like, um, escape velocity needed for any entrepreneur to get into hardware.”

Q a house. It takes a very specific type of brain to be willing to do that when two sentences ago you said how, like, it was brutal, you weren't making any money with Vetteri, you had to borrow money from friends and family to pay for your rent. How did you rationalize at a point in time where you could have easily de-risked yourself, basically risk-awning with your next business?

A At a point when we sold Vetteri, I really at that point had Maybe built, like, 10 or 12 companies. Some failed, some worked, and I, I, at that point, I felt like I knew how to build a business. I knew how to recruit talent. I knew how to do the engineering work. I knew how to ship products. Like, I knew how to fundraise. I knew how to exit. Like, basically, across the board, I felt, um, Yeah, just much more capable, and I, last five years, seven years, I was conditioned to eat nothing but like peanut butter and jelly sandwiches, and like, have like no personal burn. So, And I realized, kind of, post-acquisition, I, I didn't really care about the money. Like, I don't have a fancy car. Like, I don't wear fancy clothes. Like, I just, I don't go on fancy vacations. I love, like, building stuff. Like, I really love, like, you know, I don't play golf on the weekends. Like, come to the office. So I realized that, like, my favorite thing to do in life was to work on really hard problems. And so, um, one of the problems that I experienced dead on and with Hard tech, deep tech, is there just not a lot of people that fund that. Like, there was no, there was no real venture groups. Prior to, like, Tesla and SpaceX demonstrating, like, you know, fourth rock in orbit in a Tesla model, you know, Model S, there was, like, there was no venture groups really in there. Uh, and, you know, same for…

AI assessment note: “I realized, kind of, post-acquisition, I, I didn't really care about the money.”

Q were thinking about this business, going back to, like, idea generation, What was kind of, like, what was the calculus of how you got to this? Were you like, I just want to solve the biggest problem humanly possible, so let me think about the biggest markets? Or was it like, let me just think about the biggest problems that are plaguing society? Like, how did you arrive on figure?

A There was definitely a feeling like this is the right time to make it happen, even though it's not probably clear that this is, like, possibly the right time. This was, like, a huge, enormous industry, and it was a technology that, um, once matured, could unlock just, like, This, like, hopefully a really great future. Like, there's obviously this, like, it could be really bad, like, but we really hope that's not the case, and we're trying as hard as possible to make that not happen. Being aware of, like, you know, areas of robotics and AI, it's like, we gotta be very thoughtful about how we're developing stuff, how we're thinking about safety, all these different aspects, but to the extent this works in a good way, like, we think it's just gonna be, like, one of the most, I think it's gonna be one of the most revolutionary companies ever built. When I was looking at starting this company, The more and more I started looking at more like the details, like, okay, where's a battery and power density for, for robotics? What does the AI roadmap look like for perception, for semantic behaviors, for prediction, for planning? What does the control stack look like for this? Um, how fast do we need to run the control stack? Um, you know, what are the other dependencies of the business across the stack? Uh, how much capitals is going to need? What are the first applications? Uh, how soon …

AI assessment note: “all of that work basically was done in a nutshell, um, really early”

Q autonomy and freedom to build how you choose. And then also just the decision to sell, you know, things inevitably or likely will change when you sell a business. And so people often ask the question of like, when is the right time? How did you know is the right time to sell Vetteri? And what is your general mental model for when is the right time to sell businesses?

A Yeah, I mean, listen, I can tell you about the experience I had at Vetteri. Uh, It was either, you know, we need more capital to skip, keep scaling the business. This was not a, Hey, let's like, let's, let's like, you know, let's slow down revenues. Let's get the profitability. Let's just like, this was a land grab. Basically we were launching in. 18 cities in the first 24 months. It was a marketplace model. Marketplaces generally do better when there's supply demand that has a lot of liquidity or transactions across the platform. So basically we were trying to increase liquidity as fast as possible. Think like Uber type scale. You really want to get there fast. Get everybody using it. And then you basically want to draw prices. You want to make it, it'll continue to basically induce demand on both sides. And then, um, and then basically it scales. So it's, it's, it's a land grab business. It was, you know, so we knew that from the start, uh, it was either series B raised the capital, tons of like, you know, blocking rights on M and a tons of board seats, like going down that route, which is, we had a, you know, a fairly large series B like term sheet, or it was like, okay, we have a chance to sell the business to a really Big strategic, and it, it could be time to reset. And I felt like in my time as a career, like I felt like at that point, I didn't want to spend the next 10 …

AI assessment note: “I didn't want to spend the next 10 years in the recruiting marketplace business.”

Q dictated what was worth working on. So then pull forward to Archer. Was it a similar type of process or did you from day one arrive at the idea and then you just built out that idea and, and had you even arrived there? Cause like there are many very large problems to solve in the world. Why was this the one that you felt was worthy of your time?

A Yeah. I think for this context too, like on the transition from battery, like when we, when I launched the marketplace in like April, 2015, like within a week we knew we had a product market fit and we knew this was going to be a huge, like big business. Like it was from there and maybe like 24 months by the time we got a term sheet to sell for a hundred and ten million dollars. And it was just, it was crazy. We went from, you know, 15 people to 300 and it was just like pure scale and growth. So it was like, you know, during that period of time, we, we, I knew we had something really special. I knew I found product market fit. I knew what it looked like. I knew how to scale the business. Like it was just a great experience overall. So by the time we, you know, at one point it was like between a series, series B round of capital or like, here's the term sheet for the acquisition. I decided to go with the acquisition route. Um, uh, you know, we ended up selling to the Deco group, which is the world's largest recruiting company. And it was just a great win across the board. At the time that all was happening, I took a step back and I basically said, I'm going to make sure I really choose what area I go into very thoughtfully here. And I knew it was going to be in some advanced area of kind of like more what people call like now deep tech. So areas of, I think there's a few framewo…

AI assessment note: “I took a step back and I basically said, I'm going to make sure”