Next steps for humanity in space: My conversation with Kevin Cannon

Welcome to Faster Please!—The Podcast, the audiovisual companion to my Faster Please! newsletter. I am your host, Jim Pethokoukis.

My guest this episode is Dr. Kevin Cannon. Kevin is the principal at Starbench Analysis, a platform that provides research and analysis on space exploration, emerging technologies, and the future of humanity in space. Previously, he was a professor at the Colorado School of Mines, where he taught planetary geology and space resources.

Today, we dive into the growing competition between SpaceX and Blue Origin, the possibility of AI data centers in orbit, the race back to the moon, and humanity’s prospects for settling Mars. Hope you enjoy it.

In this episode:

* Space after the midterms (0:00)

* Orbital data centers and the night sky (8:47)

* In search of a killer app (17:35)

* The coming launch crunch (20:18)

* The race back to the moon (26:53)

* Mars and bigger rockets (30:36)

* Planetary defense and asteroid mining (35:35)

A lightly edited transcript of our conversation will appear in my Week in Review issue on Saturday. (Another option is using the Substack auto transcript function.)

But here are some edited highlights from the chat:

On space after the midterms…

Space itself and NASA has enjoyed very good bipartisan support. It always has. And even under this administration where the OMB has tried to cut back on NASA programs, particularly the science programs, Congress has been very good about restoring that funding. So I think you’re going to see kind of continued broad support for NASA and for the Artemis program. I don’t think we’re going to see any kind of radical shift in direction going back to Mars or asteroids as we did in previous administrations.

But I think there are questions about kind of the enthusiasm, like the overall enthusiasm you talked about for kind of settlement, things like that. There’s always a question of how enthusiastic the president and the vice president are, and then also who they put in as administrator, right? So we’ve kind of seen NASA flourish under Jim Bridenstine and now under Jared Isaacman. And so that can really have a huge effect on just the workforce and the overall morale of the agency.

On orbital data centers…

The other thing though, I think the more interesting debate though is the societal acceptance of these data centers. And this is something that I haven’t seen too many people talking about. The one person who’s been doing really interesting work on this is this guy, Shane Ross, who’s a professor at Virginia Tech. And what he’s been doing is making these very accurate simulations of what these orbital data centers would actually look like to the naked eye in space. So if we take the SpaceX AI-1 satellite, that’s kind of their pathfinder. It’s a 250 kilowatt AI data center satellite. And one of those satellites would be as bright in the sky as Venus is.

Now it gets kind of worse from there. So if you want to put these satellites in orbit, there’s really only one type of orbit in low Earth orbit that makes sense, and it’s called a sun-synchronous orbit. So if you have a satellite orbiting around the Earth, it can have a different inclination. So the orbit can be inclined relative to the pole of the Earth. And if you get that inclination up to about kind of 97, maybe 100 degrees, you get into these sun-synchronous orbits that perfectly track the terminator, which is the dividing line between night and day. And they get basically 100% sunlight all the time, which is what you want if you’re doing compute in space.

On the search for a killer app…

The whole lunar thing. There’s a little bit there, but there’s still not really a lunar economy per se. So yeah, I think that’s why orbital data centers have attracted so much interest from the space industry. The appeal there is that you’re selling into an existing market, right? So there’s already an existing market for compute. We know how much people will pay for inference for AI models. And so if you can kind of get set up, you can just tap into that existing market.

That’s kind of the same argument that AstroForge just made for asteroid mining. So there were all these other people saying, well, we’re gonna go out and get water from these asteroids and bring them back and sell it into this kind of water-based orbital economy. And the problem there is that it’s kind of a chicken and egg. Like you don’t really have this kind of cislunar development where you could find a customer to sell water to in space. And so they say, we’re just gonna go get metals from asteroids. There’s already the London metals market. We can just sell our products into an existing market.

On the coming launch crunch…

So if you’re trying to do any of these other things, tourism, mining, whatever, right? If you don’t have your own rockets, you’re paying the price, and so you’re paying very healthy margins to SpaceX in this case because they don’t really have robust competition. So the price for that Falcon 9 might be 80, 90 million dollars, whereas the internal cost to SpaceX is only 10 or 15.

So this is really just a question of when is there going to be robust competition to get those prices down, kind of eat into those margins and start to commoditize launch. I think Blue Origin and ULA with their Vulcan, they’re kind of like the next best hope. Obviously Blue had that very tough incident on the pad. They’re scrambling to try to get back to flight. And then you do have a new slate. I think these high prices, these