SpaceX just laid out one of its most ambitious plans yet: building a factory on the Moon to manufacture AI satellites, then launching them into space using a giant electromagnetic catapult. The idea leans on the Moon’s much lower gravity to cut the enormous fuel cost of reaching orbit, part of a bigger push to move AI computing off Earth entirely, since space offers constant sunlight and, in theory, easier cooling.
It’s a genuinely fascinating engineering vision. It’s also enormously expensive. Independent analysts estimate the full satellite constellation could cost anywhere from $1.4 trillion to nearly $10 trillion once all the pieces, satellites, launches, ground infrastructure, are added up. SpaceX also just posted a $542 million loss in its launch segment and a $1.26 billion loss in its AI operations in its first quarter as a newly public company. This is not a cheap or simple bet.
A Nearer Alternative
Here’s a question worth asking: what if some of that same appetite for hard engineering was pointed at the ocean instead of orbit?
We already know how to build in deep water. Offshore oil platforms have operated at extreme depths for decades. Submarine engineering is a mature, proven field. None of it requires a new catapult, a new form of propulsion, or solving problems that have never been solved before.
And the biggest problem AI data centers face right now, heat, solves itself in the ocean. Seawater is an enormous, free heat sink. No trillion-dollar cooling innovation required.
Same Ambition, Different Direction
I’m not arguing against space. The Moon plan is bold and worth watching. But when I wrote Aquatopia, I built a story around a simpler idea: sometimes the most advanced answer isn’t the one that requires inventing something new. It’s the one that uses what’s already proven, just in a place people haven’t thought to look.
If you’re curious what a science-based, ocean-first alternative to space infrastructure looks like, Aquatopia is available now on Amazon.