AI Infrastructure in Space? The market said yes like it’s already a reality

SpaceX didn’t just deliver the largest IPO in history by dollars raised. It also entered the public market at one of the richest valuations ever seen for a company of its size. At roughly US$1.75 trillion and US$18.7 billion of 2025 revenue, investors are paying close to 100 times annual sales. For comparison, even many of the world’s fastest-growing technology companies have historically listed or traded at a fraction of that multiple.

The IPO valuation implies that investors are already pricing in not just the success of Starlink or its launch business, but the emergence of entirely new industries that SpaceX could create over the coming decadeMost people value SpaceX as a combination of three businesses: rockets, Starlink and AI.

But Elon Musk appears to be thinking one level higher. His vision is that space itself becomes the next computing platform.

Today, AI is constrained by one thing above all else: computing power. Building data centers on Earth is becoming increasingly difficult because they require enormous amounts of electricity, land and cooling infrastructure. Permitting and grid capacity are becoming bottlenecks around the world.

Elon’s argument is that these constraints disappear in orbit.

A data center in space has access to continuous solar energy, doesn’t compete for land, and avoids many of the infrastructure limitations faced on Earth. Instead of transmitting internet from satellites like Starlink does today, future satellites could perform AI computation in orbit and beam the results back to Earth.

Even more interesting, Elon believes that once Starship reaches scale, running AI data centers in space could become cheaper than running them on Earth within two to three years.

If that happens, the opportunity becomes enormous. SpaceX would no longer simply be a launch company or an internet provider. It could become one of the world’s largest AI infrastructure companies.

Everything, however, depends on one assumption: Starship works.

Starship isn’t just another rocket. It is the technology that makes the entire vision economically possible. Dramatically lower launch costs would enable larger Starlink satellites, orbital manufacturing, space logistics and potentially AI data centers in orbit. Without that cost reduction, many of these businesses remain fascinating ideas rather than attractive commercial opportunities.

I absolutely believe SpaceX could build extraordinary businesses over the next decade. Elon has repeatedly achieved engineering breakthroughs that many considered impossible. It would be foolish to dismiss the possibility that he does it again.

But investing is about both quality and price.

Today’s valuation already assumes a future where many of these ambitions become reality. That future may well arrive, but there is little room for disappointment when expectations are already so high.