Space data centers without in space manufacturing & resource mining are of course stupid. But in space infra is what is necessary for humans to move beyond Earth & that's the critical bit, that enables everything else in the long term.
Amazon and Google are also pursuing the same thing. Either all three of these companies are full of it or they believe they have solved the blocking problems.
No they don't. They invest in R&D as they in general do.
There is a aanalysis of google engineers regarding the effort for having DC consteliations in space but its clealry research and clearly shows the difficulty of it.
Musk is the only one who needs this to keep the evaluation of Space-X.
Local protests didn't stop Colossus 1 or Colossus 2.
Local protests also didn't stop Space-X. People around his DCs or Space-X still suffer today.
He had to come up with some magic story. The Payload increased only due to his Starlink. But even then, the payload into space is basically non existend.
2025 was the year with the most payload and its only 5000t.
And for us as human beings, a DC in space is the worst case scenario. This will create a lot of stress on our atmosphere (potential, reentry poisning of our atmosphere with lithium and aluminium), co2 usage and the loose of real resources.
He will send metals into space to burn them later into our atmopshere. Limited resources we as a planet have.
And for what? For a DC? A DC which you can put in any dessert on our planet for cheap energy and not having any neighbours.
Only Edge DCs need low latency, your training clusters don't need low latency to end user, plenty of inferencing jobs don't need low latency either.