Robinhood’s Baiju Bhatt on Launching Space-Based AI Data Centers

Thank you.

Ah man, well, it was actually at the office when it was happening, with a lot of our team kind of watching it happen. And I think the biggest thing is it's hard for us to kind of wrap our heads around how much this is going to capture the American psyche in the next couple of years because the last time there was a lunar mission, I wouldn't have been born for the next 20 years, right? I think for a lot of the people that are alive today, it'll be the first time that it's happened that they've actually seen it happening up close and personal. So I think it's kind of hard to wrap your head around how big of an event I think it's going to be for just Americans as a whole because in the '60s it was like it defined half a decade of American excellence, American technology, American progress. So it kind of feels like you're at the doorsteps of like a big change in society, which kind of what we were talking about a couple of seconds ago. It feels like we're living in interesting times.

Yeah, I remember. So a little bit of background. I grew up around the space industry and my family actually moved to the US from India because my dad wanted to pursue higher education in physics and math. More specifically, physics. So, my mom was pregnant when she got on the plane, came to the US. So, I was born here. And spent the first five years or so of my life in Huntsville, Alabama where my dad was studying physics. He ended up having some health problems, had to stop out of his PhD program, and he gets a job at NASA Langley. And he worked there through most of my childhood up until around high school. And I mean, I think it was a couple of things. I think first of all, whatever as a little boy, whatever your dad is doing is kind of what you inherently aspire to want to be or aspire to understand.

My dad was kind of like a superhero when I was growing up for me. And just seeing the scale and the size of the machines that go into it because Langley Research Center, I don't know if you've been there before, but there's giant wind tunnels. Like the scale and size of all the technology behind the space industry was pretty jaw-dropping and I think when you're a little person that the normal-sized things look even bigger, right? And so, I kind of left me with this like sense of awe and wonder, but I will say also just kind of a bit of stagnation because while I was around it, like the interesting space stuff that was happening when I was a kid was relatively limited, right? Like we were Hubble Space Telescope, I remember was really big deal, but it again was a far cry from the space race of the 1960s, which I really wish I could have been a part of.

Yeah, I mean, I think at a personal level it's the journey of entrepreneurship and kind of what I was saying a second ago, right? Kind of a love of the game, like loving going from idea to thing in the world and working on something where you actually know going in that you're not an expert and that's part of the fun. That's a part of the excitement. It's not really about the outcomes but the journey there. On a substantive level these are companies that are in distinct sectors of the economy, right? Like they're the actual things that you when you're building a mobile app versus building rockets and data centers in space pretty different. But the core line that kind of ties these two ideas together is the sort of like importance and power of capitalism. And again, it also relates a little bit to my personal background because my family moved here from India in pursuit of a higher education and to be a part of this capitalist economy, right? This was the place of upward mobility. And Robinhood, when I started that company, was about empowering specifically the next generation of consumers to be a part of the financial system. And I think the simplest way of kind of capturing that was if you look at young people around the time that we started the company, which is when I was young, young adults, it was like not knowing about your finances was kind of in vogue, right? It's like somebody's like, 'Well, you know, do you invest in the stock market?' It would the sort of canned response was like, 'No, man, you know, too cool for that.' And I remember just thinking to myself it's really dumb. Like that's a really dumb outlook to have on life because it's like fitness, right? If somebody today was like, 'Well, do you exercise or do you care about fitness?' And you're like, 'No, man, I'm too cool for that.' You're like you're doing yourself a disservice, right? And I think it was a relatively short amount of time that there was a big vibe shift where now when you talk to young people finance and knowing about your finances is important and it's like an important thing for I think young people to care about. And I bring this up because I'd like to think that we had a little part in making this happen. And the way that we did it is that we put the power of capitalism in people's pockets. And when you do that, you can have these sort of like generational viewpoints change within a matter of 10 years. So, what the hell does this have to do with the space industry? Well, my observation and kind of my lived experience growing up was that the space industry, despite being this fantastically important thing, has largely been by the government, for the government, at government prices. Said differently, decidedly the flavor of capitalism felt unexplored, underdeveloped. And I think that the thing I said to myself when starting this company was, 'Could I increase the commercial envelope of space?' More specifically, what's the next sector of the economy that could and should exist in space? And it's important that we as a society are doing commercial stuff in space. Because if you are, you can inspire young people to want that to be more a part of their lives, right? Like I want to live in a world where young people that are graduating from colleges are thinking to myself, 'You know what? I don't want to get that job in consulting or I'd rather go and get a good job in the space industry because if I do that, I stand to have a great, you know, career, financial outcome, etc., etc.' Right. And so that's the connecting line. Hope that wasn't too long-winded.

Yeah. The Cowboy Space Corporation, formerly known as the artist formerly known as Aether Flux, is now the Cowboy Space Corporation and we're on a mission to power humanity from the high frontier. So what are we doing? We're trying to solve what we believe is one of the most important problems in scaling artificial intelligence and by extension one of the most important problems for our American and global economy, which is access to the energy to power artificial intelligence. In service of doing that, we're building a vertically integrated solution to putting data centers in space. That means we're building our own rockets, our own launch vehicles, and we have a very differentiated novel approach to the architecture and design of data centers in space, which is rather than having, you know, the upper stage of the rocket filled with satellites, what if instead we're able to leverage and utilize the full mass and volume of the full rocket in service of the goal of doing computing in orbit. And that's exactly what we're doing. So rather than having discrete satellites that the rocket takes to space, the rocket becomes the satellite, the rocket becomes the data center. And if you go to our website, cowboyspace.com, there's a video that plays there that actually puts a video behind this idea and you can literally see the fairing of the rocket unfolding to become a radiator for the chip sets. So a very differentiated approach and a vertically integrated approach.

Yeah, that is one of the challenges with data centers in space is that, you know, you're not going to send technicians up to rewire the satellite when it's in space. I mean, theoretically possible, but I think wouldn't make any economic sense to do that. The way we plan on approaching this is a couple things. The first one is actually developing and working partnering with and collaborating with more specifically chip manufacturers that are developing space-rated chips that are designed natively to exist in that environment and utilizing those in the space environment. Moreover, I think that there's approaches that involve over-provisioning. So actually, if you think that you're going to have to swap out some GPUs through normal operation even on the ground, rather than waiting until they fail to send a technician in to pull that server and put a new one in, actually over-provisioning it to begin with and assuming a well-understood or well-researched failure rate, there's also stuff that you can do with the environmental testing and on-the-ground testing to get the chips ready for the space environment. And I say that because you actually have to do that already for space hardware. You have to go through reverse environmental testing, whether that's vibe and shock or thermal vacuum chamber, etc. So I think there's ways of getting at that problem through other approaches.

Yeah, so there will be a latency from getting things from the satellites down to the ground. Our approach we think is going to be using real estate, whether it's adjacent to or perhaps even on top of data center buildings, to uplink and downlink to a constellation orbit. But, it's worth noting that these satellites will be in lower Earth orbit. And so, while they'll be flying higher than things like Starlink, you know, we're contemplating, let's say, the 800 to 1,000 km altitude. So, you'll have like, you know, milliseconds, tens I think close to 100 milliseconds, roughly speaking, of latency. I'll have to go and check the exact numbers on this. But, I think it's more than serviceable for inference tasks and plenty of other use cases in orbit.

Yeah, I mean, I think the most obvious and the most sort of resonant one is that you have access to truly abundant power from the sun. And that you can capture it in an environment where it's captured before it ever hits the atmosphere, right? So, there's no cloud cover, day-night cover that impacts the way it generates power. And fundamentally, there's real estate up there to do this. Like, space is I think going to be a much more industrialized environment in the years to come. And so, when you put those two together and you add it in addition to the challenges of building new capacity on the ground, the real viewpoint here is that once the technology is up and running, that you'll be able to scale net new capacity in a matter of weeks. Whereas, on the ground, we see the timeline for net new data center development, in many cases, is pushing over 5 years. So, there's a speed and a scale component that makes it a very compelling alternative to you on the ground.

Yeah, so I think that we've set a very very ambitious goal for ourselves to launch our first in-house built rocket before the end of 2028. We know it's going to be a challenge and we know it's ambitious, but if you don't try, look, if a person like me were not trying, there's no reason for me to be on the stage, right? Like you have to be ambitious if you want to do amazing things. To make that happen, the number one priority right now is development of rocket engines. And so we announced that we've got an awesome guy that's leading our rocket engine development team, and we're building out the rocket engine development people and facilities as we speak. So I'm actually going to be going up to Seattle, which is where that's going to happen, to meet with some of the folks and start looking at production sites for the engines Thursday and Friday of this week. So engines, engines, engines.

Um, yeah, I think so. I mean, kind of. I don't know. It seems like it would be like pretty rough on the body. So that I don't know. I don't know if I really want to be in space for that long. Maybe for a little while.

Appreciate it.