Yesterday, Space X filed paperwork to take the company public this June. Many are already saying this will be one of the largest and most valuable IPOs in history.
SpaceX, Elon Musk’s rocket and satellite maker, filed confidentially on Wednesday for an initial public offering, according to two people familiar with the company, setting the stage for what could be one of the largest offerings ever.
The company is committed to debuting in June, and Mr. Musk is aiming to raise $50 billion to $75 billion from going public, said one of the people, who was not authorized to speak publicly on confidential discussions.
SpaceX values itself at more than $1 trillion and would be one of the most valuable companies to reach the stock market, after Saudi Aramco’s 2019 debut valued the energy giant at $1.7 trillion. Aramco ultimately raised more than $29 billion from its offering.
So what will Musk do with the $50 billion or so he raises from this offering? Well, some of the money, perhaps a third of it, will be held back in cash. Beyond that, Musk is now moving to bring his science fiction plans for a permanent moonbase/AI satellite factory to life. Musk spoke about this last year on X.
A major additional factor should be considered.
— Elon Musk (@elonmusk) December 7, 2025
Satellites with localized AI compute, where just the results are beamed back from low-latency, sun-synchronous orbit, will be the lowest cost way to generate AI bitstreams in <3 years.
And by far the fastest way to scale within 4…
Satellites with localized AI compute, where just the results are beamed back from low-latency, sun-synchronous orbit, will be the lowest cost way to generate AI bitstreams in <3 years.
And by far the fastest way to scale within 4 years, because easy sources of electrical power are already hard to find on Earth. 1 megaton/year of satellites with 100kW per satellite yields 100GW of AI added per year with no operating or maintenance cost, connecting via high-bandwidth lasers to the Starlink constellation.
The level beyond that is constructing satellite factories on the Moon and using a mass driver (electromagnetic railgun) to accelerate AI satellites to lunar escape velocity without the need for rockets. That scales to >100TW/year of AI and enables non-trivial progress towards becoming a Kardashev II civilization.
Someone asked Grok to simplify what Elon was saying and this was Grok's response.
Elon Musk is talking about putting super-smart AI computers on satellites in space. They'd run on solar power, zap data back to Earth super fast, and grow way bigger than anything on land because space has endless energy. Eventually, we'd build factories on the Moon to make more, launching them with giant electromagnetic slingshots. It could make AI insanely powerful, like leveling up humanity to sci-fi status—measuring success in energy (watts) and materials (tons) instead of money.
Earth-based data centers have a lot of disadvantages. They take up a lot of land. They use a lot of energy. They heat up and require even more energy for cooling. But some of those problems are solved if the data centers are operating in space. First you have endless free energy from the sun. Second there is no cost for land or permits or maintenance. Third, there is no danger of overheating. All of those costs are reduced and replaced with the cost of getting the satellite into orbit, which is what Space X does.
However, the cost to put satellites into orbit would greatly decrease if instead of lifting them into orbit from earth you could build them and launch them from the moon. The moon's gravity is low enough that you wouldn't need a rocket at all to do this. In theory you could use a rail gun to accelerate them to escape velocity much more cheaply (and again using solar power).
Of course getting enough material to the moon to build the factory and the railgun could be expensive and that seems to be where Musk plans to spend some of his billions. He'll be sinking Space X's IPO money into a plan that requires a bunch of trips to the moon using Space X's Starship.
Musk also likely plans to put billions into his plan for a new earth-based Terafab, a chip fab like the ones in Taiwan but on a more massive scale. The idea is to bring the chip production for all of his products-- cars, robots, satellites, spaceships-- in house so he's not reliant on outside companies to supply what he needs moving forward. Last week Musk made a public announcement for the Terafab project and it opened with a discussion of science fiction.
I'm not going to summarize all of this because it's really too grand a plan to boil down to bullet points. But if you're interested in Musk's effort to move humanity beyond current energy and intelligence bottlenecks here on earth, watch his presentation below.
Of course building all of this will take time as well as many billions of dollars. It might be 2028 before the first parts of this are up and operational. Still, that's working at light-speed compared to others in the industry. But building things fast is something Musk is good at. If completed as planned, the Terafab will be about 15 times the size of the Pentagon and 10X the size of the current Gigafactory in Texas.
Is Musk going to deliver on all of this in the timeline he anticipates? Probably not. But the scale of this is so big that even achieving 10% of it would put his effort on par with the biggest chipmaker in the world.
