“Join xAI if the idea of ​​mass drivers on the Moon appeals to you,” CEO Elon Musk proclaimed yesterday after a restructuring that saw a stream of former executives except for the AI ​​lab.

This is an interesting recruitment strategy after the merger of Musk’s rocket maker company, SpaceX, and the expected IPO of the combined company. You might think that xAI employees should be drawn to the achievement of AGI, using deep learning models to disrupt traditional software companies, or bad wordplay like “Macrohard.” But instead, Elon is going to the moon.

After outlining the plans of building AI data centers in orbitthe main synergy between the two companies, Musk took the idea further. “What if you want to exceed one terawatt per year?” Musk asked. “To do that, you have to go to the moon…I’d really like to see a mass driver on the moon that shoots AI satellites into deep space.”

According to Musk, the step beyond Earth-orbiting data centers is larger computers in deep space. And in addition, Musk said that the best way to achieve that is to build a city on the moon to make computers in space and drop them into the solar system using a large maglev train.

If that all feels a little off, veteran Musk watchers know there’s a clue where the discussion can be found in a video of a meeting with all hands xAI shared with the public. The slide depicting the moon base comes at the end of the presentation deck, where, during SpaceX pep talks, Musk usually shares renderings of SpaceX rockets that landed on Mars and waxes rhapsodic about the future of multi-planetary humanity.

Notably, the moon base comes after SpaceX publicly backed away from its long-held goal of colonizing Mars. Now, with xAI in the corporate fold, Musk needs a new science fiction metaphor for the future: In this case, the Kardashev Scalea theoretical scale of galactic civilizations created by the eponymous Soviet astronomer in the 1960s. The idea is to scale up the use of energy – the first civilizations thought about how to use all the sources of electricity on their planets, and then (hypothetically) go into space and build infrastructure to capture the energy of the sun.

With the moon base, Musk says the company will be able to “maybe even use a few percent of the sun’s energy” to train and operate AI models. “It’s hard to imagine what an intelligence on that scale would think,” he told his staff, “but it’s exciting to see it happen.”

In the nine years since Musk revealed his plan for Martian exploration and colonization, the vision has been an effective hiring tool for SpaceX: The founding story of Musk’s interest in the Red Planet offers a long-term vision that unites the company’s various development efforts, and marks the company’s ambition among other space contractors settling for incremental government jobs. “Occupy Mars” t-shirts offer a visible symbol of SpaceX’s aspirations.

That’s where the hypothetical moon base fits in — part of a long history of Musk wrapping his companies in a powerful narrative. It’s a million people living on Mars, but now preparing for a future where AI is the most interesting thing. Martian mission creep has become less visible at Musk’s May 2025 Starship updatewhen the presentation ended with a now-cancelled vision of Tesla Optimus robots clomping across the Red Planet.

There’s only one problem with SpaceX and Mars: No one wants to pay them to get there. Plans announced in 2016 to repurpose the company’s Dragon spacecraft as a Mars lander were abandoned the following year after technical challenges proved too costly. And since Musk exposed the vehicle that will be Starship in 2016, its capabilities, which were initially intended for the colonization of Mars, were scaled back to focus on two additional paid tasks: launching satellites for the Starlink comms network and $4 billion worth of contracts to land astronauts on the moon for NASA.

Unlike a multi-planetary civilization, there may be some logic in SpaceX’s purchase of an AI that burns money and social media to build data centers in Earth’s orbit, especially if the predictions of rising demand and costs on Earth come true. Experts suggest that this is possible in the 2030s.

Hypothetically, building satellites on the moon would require more than Musk’s other dreams to come true first. Scientists and startups are experimenting building chips and other space precision components. But the creation of tons of advanced computers on the moon means that we live in a universe where it is very cheap to go to space, which is the essential requirement for technologies, to get all the raw materials for such an effort on the moon, and everything else that is needed for a “self-sustaining city.”

In a sense, that’s the point: It’s the, uh, spinning goal. If meme-happy retail investors buy the argument, they could turn SpaceX shares into the next Tesla. The engineers, AI or aerospace, that Musk needs to achieve his goals may find change. But the vision is a way to explain what xAI is, apart from an LLM that is probably best known as a pervert. As one of the departing company executives SAYS on his way out the door, “every AI lab is building the exact same thing, and it’s boring.”

Building a solar system-scale supercomputer on the moon is many things (I get emails for not using the word “crazy”), but it’s not the same thing, and it’s not boring.