Gavin Baker: SpaceX Might Be the Greatest Company of All Time

Doing great, man. How are you guys?

A round of applause for the banker.

20% pop almost to a tee.

I thought Goldman Sachs and Morgan Stanley did a very good job.

Well, I think the market is always, you know, quarters matter. The runners on X, they're these reply people on X for any topic or experts. But somebody once told me a marathon is 26 one-mile runs. And all the runners were like, 'No, that's not true at all.' Running a mile is different than running a marathon. But every quarter matters. At the same time, I do think if you look at how the market has looked at Tesla over the last five, six years, Amazon during the days when they were building out AWS and their retail distribution infrastructure when they'd go through these investment cycles, I do think the public market has a much greater tolerance for investment and a much longer time horizon than a lot of people in the venture ecosystem give it credit for.

Yes, he did. He did the work on SpaceX. So I thought you guys might want to talk about that. But I do think the story is not as far out as you made it seem. There are two variables that are going to matter a lot over the next year. The first is how quickly can they bring on terrestrial compute? It does seem like they monetize gigawatts at a higher rate than anyone else. We know from Jensen that they bring on data centers faster than anyone else per Jensen's words. Everyone in the ecosystem is really incentivized to get land and power in the hands of people who can energize it, because everybody starts making more money when the GPUs are energized. If they can energize two, three, four gigawatts in the next year, that's a lot of revenue. Prices may go up or down, but that's really going to matter.

Yeah, or Macro Hard, Macro Harder, Macro Hardest to that name. Yeah, Macro Hardest.

Yeah. We talked about it with Brad about how Composer 2.5 after three weeks of RL and supervised fine-tuning on Colossus 2 is kind of Pareto dominant. So what's going to happen when it's applied to a bigger, better base model? I do think Cursor being in half the Fortune 500 is interesting. So those two things—while I'm super excited for going to Mars, asteroid mining, a city and mass drivers on the moon—all that stuff is awesome and there's a decent chance it happens in my lifetime. But there are much more tangible near-term drivers here.

That's my compute.

A way to interpret that is that maybe the team at SpaceX is pretty confident in their ability to bring gigawatts online. The Altimeter figure is that they've ordered 20% of the Rubin chips, and Rubin is an epic chip. Like Blackwell, it was really hard to get it online. Rubin is kind of a more drop-in replacement. It's a really good chip. You'll have the Grock LPUs integrated at some point in the next six to nine months. So one way to interpret that is they're pretty confident that they can bring data centers online pretty quickly. But I also don't remember all the details. I do think there was an out in the Anthropic agreement.

Yeah.

My thought is that being a token factory is more than enough for the next 5 to 10 years, man. We need more token factories. But we'll see, and if they need that stuff, they will build it.

Well, I think there are many ways to go into the token factory business. If you're Meta, you could just say to someone like Fireworks, 'Here are 100,000 GPUs. We'll take a revenue split and see what happens.' I am increasingly looking at EV to net PP&E as a valuation metric, with the idea being installed atoms on Earth are going to appreciate in value. Meta's EV to net PP&E multiple is in an interesting place; it suggests the market has immense skepticism about their ability to monetize their own asset base. And I think that's warranted given that all we've got out of management is this concept of personal superintelligence. It would be different if Meta AI was at the top of the app store and everyone was using it. No one doubts that once Zuck has an AI product with a billion dollars of revenue, he can take it to tens of billions, but it's unclear what this API business will look like. We know they can use GPUs just on making the ad product better. But there are a lot of questions.

Oh yeah. So you're telling me there's a chance. The other quote from Talladega Nights is if you're not first, you're last. There are so many good quotes from that movie.

Yeah, 100%. But I'm pretty confident Mark Zuckerberg has proven to be a good entrepreneur over time and has been willing to pivot. It's so funny — right now, Meta tells investors there's no chance they will monetize GPUs externally. Well, 10 days before they announced all those layoffs, which was shortly after Brad Gerstner wrote that letter to Mark where he said, 'You're the king, you can do whatever you want, but this is what I might respectfully encourage you to do.' They went from saying we're not going to do that to super focused on OPEX in a couple of days. So when people meet with Meta and hear they're not going to monetize their GPUs externally, I don't know, check back in a few hours.

Yeah.

Yeah. A few things. People say 'retail' in a pejorative way, and I would just say stupid is as stupid does. We track indices of retail favorites, and man, they're up a lot in '24. So stupid is as stupid does. I think retail is probably outperforming the overwhelming majority of professional money managers, whether private equity, venture, or public equity managers. So I don't think it's pejorative. In the case of SpaceX, I'm not sure where this comes from, but I actually think after the IPO, some of that retail stock may have been allocated to people who were flipping it because there was more institutional buying. Retail is very sensitive to momentum, so maybe what's happening today changes that. Another underappreciated thing about the SpaceX IPO is that over 10,000 SpaceX employees bought in the IPO. There are all these lockup analyses, but employees own a lot of this. If you wanted to sell, you could have sold in the offering in the second half of last year. So supply-demand dynamics will be interesting. In general, retail has been a more powerful force in the market over the last three and a half years than at any other time in my career, including the year 2000 bubble, which this is nothing like. But yeah, they matter, and stupid is as stupid does.

Every CDN has this business. They signed a $1.8 billion deal with Anthropic. If you have all these points of presence, you can deliver really low-latency tokens to high-value users. What's interesting for all these CDNs is they're getting a massive premium for that latency. One lesson from Cerebras is people are willing to pay for speed. But my gut is that these CDNs are delivering less than 1% of tokens consumed on Earth, maybe less than 10 basis points. Most of the time, tokens happen internally and you just get the result. Would I say these guys are in the token path? I would say they have a path towards being in the token path. They have part of their business that's in it today, and they're trying to move it more and more there.

I mean, GPUs are going everywhere—going in your bed eventually.

Yeah, I do think the 'bottleneck bros' as they are called—this bottleneck trade is kind of nearing its end. The Wall Street Journal had a story about a Japanese company that makes something. Everybody got into it because they thought they were going to raise prices, and then they said we're not going to raise prices. I almost posted on X: welcome to Japan to all the bottleneck people. Everybody's had 'claw and run'—what's the next bottleneck? That was the game for the last year. The next game is what has enduring franchise value.

So I don't know that I would say land is in the token path. Maybe beachfront property, but that's about it.

Yep. Firmly in the token path. All that stuff is in the token path.

No. I think every country is going to want to have their own sovereign AI strategy, at a minimum for national defense. I saw today that the EU restricted exports of MRLs. I think that's a joke. People have gone so far on the MRL leash that they're saying it's like Mythos is distilled on it and breaking all the benchmarks. It's a good model, but I don't know that it's quite there yet. We can imagine we'll get there for sure. I think what sovereign AI will look like for essentially all countries other than the United States and China (and I think China will fall further and further behind) is just using one of these providers to do some reinforcement learning on your language, culture, and values. You have a system prompt, do some supervised fine-tuning, and run that on whatever the best open source model is. Then you run it in your own data centers so you feel it's safe for defense, intelligence, and policing activities. So there'll still be a big buildout for that, but sovereign AI at the frontier—I don't see it.

China's made a terrible mistake. It feels like the administration has been willing to let them buy H200s or P30s. They have this crazy belief that their own internal chips are good enough. They're not. What's making them think that is that Chinese labs are very, very good at distillation. Somebody told me it only took 160,000 reasoning traces from O1 and O3 to get the original DeepSeek. They're very clever at industrial scale distillation, running it through multiple APIs, pinging each API from iPhone farms. They've gotten really good at that, but all that goes away if people stop releasing these models at the frontier. Mythos is a sign of things to come there.

George, please.

Worried is the wrong word.

Well, I did own 15% of Nvidia when it was a sub-$2 billion market cap, and I owned 10% of Tesla when it was a sub-$2 billion market cap. But SpaceX certainly has been a special experience for me. I think there is a chance that it is one of the most important, iconic companies of all time. It already is. Maybe it becomes the most. I want to take a step back: all this trillionaire hate from the left. There was a great post from my friend Kevin Mafy saying, 'Oh, this is a Bond villain who's decarbonizing the world.' Last I checked, people on the left care about the environment, connecting poor people in low-income countries to the internet, and helping blind people see. That's a villain? There are so many parts of the SpaceX story—blue-collar workers who've gotten super wealthy. If you're opposed to data centers on Earth because of the environment, SpaceX is your solution. Another post said the left won and they don't even appreciate it. The world's first trillionaire has done more to solve the environment than everyone else combined. And there's the Elonsolves-world-hunger thing—if he sold every piece of stock, it would send the world into a global economic downturn, losing millions of jobs. Somebody said the World Bank could solve world hunger with $5 billion, and Elon said, 'If you prove it to me, I'll do it.' Turns out you can't solve world hunger for $5 billion. SpaceX, Nvidia, Tesla have all been very special to me in different ways. But there is nothing like a rocket launch. As a visceral experience, I highly recommend everyone listening—if you have kids, take them to a launch. You can go to a beach at Boca Chica and be near the viewing area. Take your kids to see a launch. It's super inspirational that humans are capable of this. It's much more visceral than you realize—the sound, the blast of hot air, a lot of people cry. I'm pretty skeptical there will be another company that delivers a visceral experience like that. If SpaceX puts a city on the moon and Mars and becomes the British East India Company of the solar system, a lot of hard engineering needs to happen, but this will probably be the most important company of my lifetime, maybe of all time. So from that sense, I don't think there will ever be anything else like it. I guess that's a little bit sad, but there's so much awesome stuff happening that it's hard for me to get too down.

Yeah, hard to be too down.

Haven't run up? Pros have taken everything under $2 billion above. Maybe not public companies, but in private areas, we're starting to see the knock-on effects of AI in bio, material science. Neuralink was sub-$2 billion for a long time. There are more not-in-theme but accelerated-by. There's loads of incredible venture startups under $2 billion. I talk to them several times a week. In public, until the bottleneck bros and AI came along, there wasn't much happening in small cap land. By the way, on all these small caps and bottlenecks, it feels a little weird that someone with a large following on social media can post about a $35 million market cap company and send it up a thousand percent. There's a lot of that going on. That person is smart, but there's a lot of shady stuff happening.

Yeah.

Thank you so much. Anytime. Let's get the whole team on too.

Let's do that with the whole team.

Awesome. I would love it. You got a conference room with a nice wide-angle camera, you can have the whole crew there. That's great.

Have a great rest of your day. Thank you so much for coming on.