Sam Altman and Elon Musk exchanged sharp social media posts over the weekend, highlighting the gap between ambition and reality in the space-compute sector.
Altman replied to Musk’s accusation that he was a scammer by stating, “homeboy you’re the one sellling [sic] public market investors on short-term space datacenters.”
Ignoring the casual address, Altman is voicing a conclusion many experts have reached but public market investors appear to be overlooking: space data centers are not going to be a serious business anytime soon.
SpaceX’s plan to launch a fleet of orbital data centers for AI inference tasks drives the company’s 2-trillion-dollar valuation. Bullish analysts argue the potential for that processing power to fuel SpaceXAI models or act as an orbital neocloud is unprecedented in the current AI boom.
However, speaking with subject-matter experts reveals a different picture. This group includes entrepreneurs behind other space data center startups, the team at Google developing its orbital compute project, and engineers who have run the numbers. They all agree: this sector will not make a big dent until rockets are much cheaper and high-powered satellites can be produced at low cost, en masse.
Musk’s response is predictable. Starship, SpaceX’s large new rocket, is expected to make its 13th test flight on July 16. If the team can get the vehicle to fly repeatedly, the business case for data centers could finally close.
Even if the company successfully recovers both rocket stages on this test flight, operational reusable flight will likely remain years away. Space data center launches will probably take a back seat to SpaceX’s commitments to NASA and to building out its own Starlink network.
SpaceX also admitted during its IPO road show that Starship may not be fully reusable in the near term and will need to discard each second stage during every launch. This would end any hope for economical space data centers.
Consequently, Musk’s rejoinder — “We start flying them next year” — falls a bit flat. There is no doubt SpaceX could launch a satellite equipped for high-speed data processing next year. The real question is when it will be able to launch and manufacture them at scale. That is likely a question for the 2030s.
What it means
For people building models or running inference, the implication is clear. The promised orbital compute boom is not happening now. Until launch costs drop and mass production becomes viable, heavy lifting remains grounded on Earth.
