Text settings
Story text
SmallStandardLargeStandardWideStandardOrange
* Subscribers only
Learn more
SpaceX has formally acquired another one of Elon Musk’s companies, xAi, the space company announced on Monday afternoon.
“SpaceX has acquired xAI to form the most ambitious, vertically-integrated innovation engine on (and off) Earth, with AI, rockets, space-based internet, direct-to-mobile device communications and the world’s foremost real-time information and free speech platform,” the company said. “This marks not just the next chapter, but the next book in SpaceX and xAI’s mission: scaling to make a sentient sun to understand the Universe and extend the light of consciousness to the stars!”
The merging of what is arguably Musk’s most successful company, SpaceX, with the more speculative xAI venture is a risk. Founded in 2023, xAI’s main products are the generative AI chatbot, Grok, and the social media site X, formerly known as Twitter. The company aims to compete with OpenAI and other artificial intelligence firms. However, Grok has been controversial, including the sexualization of women and children through AI-generated images, as has Musk’s management of Twitter.
Vertically integrating AI and space
There can be no question that the merger of SpaceX—the world’s premiere spaceflight company—and the artificial intelligence firm offer potential strategic advances. Musk strongly believes that artificial intelligence is central to humanity’s future and wants to be among those leading in its development. With this merger, he plans to use SpaceX’s deep expertise in rapid launch and satellite manufacturing and management to deploy a constellation of up to 1 million orbital data centers. This will provide the backbone of computing power needed to support xAI’s operations.
Musk’s plan for the merged companies is predicated on several assumptions, including that AI is not a bubble, but rather a technology that will be fully embraced in the future; that orbital data centers are cost-competitive compared to ground-based data centers; and that compute is the essential roadblock that must be solved for widespread adoption of AI by society.
If these assumptions are true, the merged SpaceX-xAI company holds a powerful position. It could potentially own a full stack of capabilities, from launch to orbital bandwidth to frontier AI models, and with Starlink Internet, it could provide AI on demand, anywhere in the world, to any mobile device.
SpaceX already has the world’s workhorse reusable rocket with the Falcon 9. It can presently deliver about 20 tons to low-Earth orbit for an internal cost of $15 million, compared to more than four or five times that on the open market. Moreover, SpaceX is working toward fully reusable super heavy lift rocket with its Starship vehicle.
The privately held company also operates more satellites, about 9,600, than any other country or company in the world by a factor of 10. It has extensive operations not just in deploying but also in operating this constellation over the last decade. This is not a simple capability.
“I would say there have been as many engineering advancements in orbital safety and collision prevention in the last 10 years as there have been advances in rocketry, and that may have gone unnoticed,” said Brian Weeden, director of civil and commercial policy at The Aerospace Corporation, in an interview.
“Happening really fast”
In an email to SpaceX employees on Monday, Musk said Starship will begin launching V3 Starlink satellites into orbit this year, as well as the next generation of direct-to-mobile satellites. The launches, he said, will be a “forcing function” to improve the performance of Starship, making it more rapidly reusable for data center deployment.
“The sheer number of satellites that will be needed for space-based data centers will push Starship to even greater heights,” Musk wrote. “With launches every hour carrying 200 tons per flight, Starship will deliver millions of tons to orbit and beyond per year, enabling an exciting future where humanity is out exploring amongst the stars.”
Musk told employees that launching 1 million tons per year of satellites, generating 100 kW of compute power per ton, would add 100 gigawatts of AI compute capacity annually, “with no ongoing operational or maintenance needs.” Ultimately, Musk believes there is a path to launching 1 TW/year from Earth.
“My estimate is that within 2 to 3 years, the lowest cost way to generate AI compute will be in space,” Musk wrote. “This cost-efficiency alone will enable innovative companies to forge ahead in training their AI models and processing data at unprecedented speeds and scales, accelerating breakthroughs in our understanding of physics and invention of technologies to benefit humanity.”
Musk is clearly bullish on the future of AI and on space’s potential to address the voracious power needs of AI data centers. Many people in the AI industry speculate that artificial intelligence is ly to go through serious and sustained growing pains, or doubt that space-based data centers can compete with operations built on the ground. But Musk, more than anyone, has the means to press forward the bull case for space-based AI, and he is going for it.
Monday’s merger s an ultra-ambitious filing on Friday with the Federal Communications Commission in which SpaceX sought permission to launch 1 million satellites that will operate as “orbital data centers.” The company said it would deploy the satellites to orbits with an altitude between 500 and 2,000km, and 30-degree and Sun-synchronous inclinations.
SpaceX also recently announced its plans to deploy a space situational awareness system, called Stargaze, that will use star trackers to provide data on potential conjunctions between satellites in orbit. The goal is to help de-conflict satellite trajectories and avoid collisions in low-Earth orbit.
“This is all happening really fast,” said Victoria Samson, chief director of space security and stability for Secure World Foundation, in an interview.
Crowded orbits
Samson said that, at present, satellites have a fairly large “bubble” of space around them when it comes to collision detection. This is because of uncertainties in the precise location and movement of vehicles. If you improve space situational awareness, such as what SpaceX seeks to do with Stargaze, those bubbles could be shrunk to reduce the number of potential collision warnings. But that will come with risks.
“There’s a lot of room in space, of course,” Samson said. “But the question is, how much risk do you want to take?”
A technical expert at The Aerospace Corporation, Marlon Sorge, told Ars that many unanswered questions about SpaceX’s proposed megaconstellation for orbital data centers make it difficult to assess the risks of collision. This includes their size (they will require very large solar arrays to collect sunlight) and precisely where the satellites will be deployed. There is already a lot of debris at around 800 to 1,000 km above Earth from previous collisions, including from an infamous Chinese anti-satellite missile test in 2007, which created more than 3,000 pieces of golf-ball-sized or larger debris.
Above that altitude, there is less debris, Sorge said. But objects at that altitude take centuries to deorbit naturally, due to the very limited atmosphere.
“The big challenge at those altitudes is the stuff that’s up there stays up there,” Sorge said. “If you generate more debris, if you have problems, it won’t go away, so you’re stuck with it.”
SpaceX sought to address these concerns in its regulatory filing, noting that each satellite would have “redundant maneuverability capabilities” in order to deorbit into Earth’s atmosphere. The filing also appears to recognize emerging science that indicates that aluminum burning up from reentering satellites is harmful to ozone levels. To address this, SpaceX is considering moving aging satellites into “high altitude Earth orbits or heliocentric orbits.”
However, Sorge noted that the amount of energy, or delta-V, needed to move a satellite from low-Earth orbit into a heliocentric orbit is “non-trivial.”
Has SpaceX lost its way?
One of the many questions raised by the new merger is whether SpaceX has lost its way. Musk founded the company in 2002 with the singular purpose of settling Mars, an audacious if not impossible goal at the time. In the decades since, SpaceX has made credible progress toward Mars, and with Starship, humanity has for the first time a transportation system potentially capable of landing humans on the red planet.
But acquiring an AI company and putting so much effort into orbital data centers? Is this consistent with the Mars mission? Musk clearly thinks it is.
“While launching AI satellites from Earth is the immediate focus, Starship’s capabilities will also enable operations on other worlds,” he wrote. “Thanks to advancements in-space propellant transfer, Starship will be capable of landing massive amounts of cargo on the Moon. Once there, it will be possible to establish a permanent presence for scientific and manufacturing pursuits. Factories on the Moon can take advantage of lunar resources to manufacture satellites and deploy them further into space.”
And from there, he said, Mars will be firmly on the horizon.
“The capabilities we unlock by making space-based data centers a reality will fund and enable self-growing bases on the Moon, an entire civilization on Mars and ultimately expansion to the Universe,” he wrote.
That’s the vision, at least.
Eric Berger Senior Space Editor
Eric Berger is the senior space editor at Ars Technica, covering everything from astronomy to private space to NASA policy, and author of two books: Liftoff, about the rise of SpaceX; and Reentry, on the development of the Falcon 9 rocket and Dragon. A certified meteorologist, Eric lives in Houston.
