This is what the build out for a 300MW datacenter looks like. Multiply this visual by 1,000 then ask yourself, “Why do Elon, Sundar, Sam, Jensen, Schmidt and Bezos all agree and have come to the same conclusion that we need to repackage this compute and shoot it into space?” The answer is they all expect the real demand multiplier to be 1,000,000 not 1,000. Arguments against Space Data Centers revolve around limited op ex savings, depreciation schedules, etc which are all perfectly reasonable when you only care about the next 100 data centers or the next 18 months of IRR. Meanwhile the titans of the upcoming Industrial Revolution have all seen the writing on the wall and realize we simple cannot meet the demand for intelligence building project by project on the ground. You cannot copy and paste this 1M times in the next 10 years. Not when you need buy in from 1M local governments, interconnect 1M power hungry buildings, and appease NIMBY orgs 1M times. They’ve concluded it actually makes more sense to package up compute into modular boxes, shoot a lot of them into space and beam the results back. Betting against data centers in space is simply a bet against AI demand scaling beyond 100 GWs per year within 10 years. It’s not about radiators, ITU approvals, capex v opex, Nuclear energy scale up, etc. You could be right making that bet, but be honest with yourself about what bet you are making.