Key Takeaways:
Elon Musk aims to build a self-sustaining city on the moon within a decade and launch the first AI satellites next year, laying the groundwork for space-based computing.
SpaceX plans to launch the first of as many as 100,000 AI satellites next year and send construction materials to the moon by year-end, as Elon Musk accelerates his timeline for space-based computing and off-world colonization.
"If you've got a really capable AI system, are you going to embed that in every robot? No, you're going to have centralized compute — that's what these satellites are for," Jim Cantrell, a founding team member at SpaceX and CEO of Phantom Space Corporation, said.
Musk said in a radio interview with Texas Gov. Greg Abbott that SpaceX will launch its first AI satellites in 2027, with large-scale deployment within two years. The company filed an application with the Federal Communications Commission for a constellation of as many as 100,000 solar-powered satellites that would operate as an orbital AI network, providing computing power for billions of devices on Earth and in space. SpaceX also launched 29 Starlink satellites into low Earth orbit Thursday using its Falcon 9 rocket, which has completed 36 return flights.
The plan represents a fundamental shift in computing infrastructure deployment. Instead of building more data centers on land — consuming power, water, and real estate — Musk envisions moving compute into orbit. "The way to expand compute without using up all the land on Earth is to do so in space," he said. The approach faces significant hurdles: a coalition of environmental and scientific groups has petitioned the FCC to pause licensing for orbital data centers, citing light pollution and risks to astronomy. There are about 15,000 satellites currently orbiting the planet, according to the European Space Agency.
Moon and Mars: The Two-Planet Strategy
Musk said he hopes to transport tens of thousands of people to a lunar base within 10 years, calling it "much faster to complete a moon city" than a Martian one. SpaceX plans to land an astronaut on the moon in two to three years, then expand rapidly. The first materials for building colonies are expected to launch before the end of 2026, with Tesla's Optimus robots doing much of the construction work before humans arrive.
"The robots build the settlement before the humans show up," Cantrell said. "And unless Musk does something stupid or somebody kills him, he'll live to see it all."
Mars remains the longer-term goal. Musk said he expects to send the first humans to the red planet in about five years, with thousands more over a 10- to 12-year timeframe. The first self-sustaining cities could be operational between 2045 and 2055, he forecast. The trip takes about six months, compared with three days to the moon, making Mars a far more complex logistical challenge.
NASA Administrator Jared Isaacman described the moon as a "proving ground" for deep space technology. "There will be a lunar economy someday," Isaacman said, adding that asteroid mining could follow. NASA's Artemis IV mission is targeting a 2028 moon landing, while China aims to land taikonauts by 2029, intensifying the space race.
Les Johnson, formerly the chief technology officer at NASA's Marshall Space Flight Center, said the moon offers a critical testing environment. "If something goes wrong, people can get home quickly and you can get spare parts quickly," Johnson said. "You go through all those learning processes on the moon so you don't have to worry when you get to Mars."
The Credibility Gap and the Cost
Musk has a long history of missed deadlines. He has predicted humans would reach Mars within 10 years at least 19 times since 2011, a New York Times review found. In 2017, he announced plans to fly private citizens around the moon by 2018 — a trip that never happened.
The economics are also daunting. SpaceX's S-1 filing shows Musk would receive 1 billion shares if the company reaches a $7.5 trillion market cap and builds a Mars colony with at least 1 million inhabitants. Analysts have questioned whether Starlink profits can sustain the cost of maintaining a colony. Musk briefly became the world's first trillionaire after SpaceX's IPO in June, but the company's falling share price has since pushed his net worth below $1 trillion.
SpaceX's public listing has made its ambitions a tradable thesis. The company's AI satellite program could compete with terrestrial data center operators and cloud providers, while its Starship program positions it for government contracts tied to NASA's Artemis missions and potential defense applications. Competitors including Blue Origin and Google are also exploring orbital data center concepts, though none have matched the scale of SpaceX's FCC filing.
This article is for informational purposes only and does not constitute investment advice.
