Elon Musk is once again trying to bend space and business to his will, this time by pitching a lunar catapult to hurl artificial intelligence satellites into orbit. The plan centers on a factory on the moon that would build and fling hardware like ammo from a railgun, feeding the soaring demand for AI computing power. It is an audacious answer to a real constraint, and it arrives just as Musk’s AI company, xAI, faces internal churn and an approaching IPO.
Rather than only launching more rockets from Earth, Musk is now talking about using the moon itself as an industrial platform, a kind of off‑world supply chain for data centers in space. The idea is to use lunar gravity and vast solar power to feed AI systems from orbit, even as his long‑standing talk about colonizing Mars fades behind a nearer target. The question is not whether the vision is wild, but whether it is a serious strategy or a spectacular distraction.
The moon catapult and its basic physics
Musk is reportedly pitching a plan to build a lunar catapult that could fling satellites at a velocity of about 2.38 km/s, fast enough to kick small payloads off the moon and toward useful orbits without chemical rockets. That figure matters because the moon’s lower gravity and lack of atmosphere make such a launch system physically plausible in a way that would be impossible on Earth. In effect, the moon becomes a giant slingshot, turning electrical energy into orbital speed.
According to reporting on his proposed catapult idea, the system would rely on electromagnetic acceleration, similar in principle to a railgun or maglev train. Instead of hauling fuel, builders would haul power systems and metal rails, then fire satellites like bullets off the surface. The physics is not science fiction, but scaling it to industrial levels would demand extreme reliability, precise guidance, and a way to catch those “bullets” into stable orbits without turning them into space junk.
From Mars dreams to lunar hardware
For years, Musk has talked about colonizing Mars as his grand narrative, positioning red‑planet cities as the endgame for his space ambitions. Now he is talking about building what one report likened to Looney Tunes contraptions on the moon instead, a tonal shift that says a lot about where the near‑term money and engineering focus might go. Mars remains the myth, but the moon is starting to look like the factory floor.
Coverage notes that Musk has talked about Mars for years, yet he is now describing a catapult on the moon to help power xAI’s artificial intelligence network. That shift from Mars settlement to lunar machinery may sound like a cartoon gag, but it lines up with a more immediate business case: AI needs energy and compute right now, not in some far‑off Martian future. In this framing, the moon becomes less a destination and more a giant industrial park in the sky.
xAI turmoil and the moon pivot
The timing of this lunar push is as telling as the physics. Co‑founders are leaving xAI, and an IPO is looming, which means the company is in flux at the exact moment Musk is sketching out a multi‑decade moon project. He gathered xAI employees on a Tuesday night and, according to multiple accounts, used that meeting to frame the moon not as a side quest but as the next chapter of his AI strategy.
Reporting from inside the company describes how co‑founders are leaving just as an IPO is looming, and how Musk used that staff meeting to talk about a lunar future. Another account says Musk is reframing his artificial intelligence ambitions around the moon, a shift summed up in the line that he is turning to lunar plans as co‑founders exit before IPO, which raises real execution and governance questions for a company already under scrutiny.
A lunar factory for AI satellites
At the heart of the scheme is a manufacturing vision: Musk plans to build an AI satellite factory on the moon, turning lunar regolith and imported components into finished spacecraft. Reports describe this not as a vague dream but as a proposed industrial facility dedicated to his AI company. In that story, the catapult is simply the shipping dock for a steady stream of hardware.
One financial report says Musk plans to an AI satellite factory on the moon and has said on X that SpaceX will focus on building a base there before attempting Mars. Another piece quotes him telling staff, “You have to go to the moon,” during a meeting, and adds that he wants this factory to help power systems searching for signs of aliens, as reported by coverage of that. Together, these details make the factory sound like a central pillar of his long‑term AI plans rather than a throwaway line.
Catapulting compute: 8% of America’s power
The motive behind all this hardware is not just exploration; it is energy. Musk told employees at xAI that the plan involves adding about 8% of America’s entire power output every year from orbit, and that this would be 100% dedicated to AI. He framed this as a way to keep up with a demand curve for computation that he believes is rising far faster than traditional grids can handle.
In that meeting, Musk told staff, “So we’re talking about adding ~8% of America’s entire power output… but 100% dedicated to AI every year from orbit.” One analysis says Musk and others envision vast constellations of solar‑powered satellites in low‑Earth orbit providing compute directly from space to feed what it calls an “insatiable demand for compute.” In that picture, the moon factory and catapult are meant to be the upstream supply chain for this orbital computing belt.
More from Morning Overview
*This article was researched with the help of AI, with human editors creating the final content.
