A new gound for VCs as well
Sam Altman calls it "ridiculous." Jim Chanos calls it "peak insanity." And yet — Starcloud Orbital, Axiom Space Sophia Space Kepler Communications Inc. AetherFlux-X have raised near $1B and filed plans, or put actual hardware in orbit to build AI @datacenters in space. SpaceX filed for 1,000,000 satellites. Blue Origin answered with 51,600. A Y Combinator startup hit a $1.1B valuation faster than almost anyone in YC history. Google ran its TPUs through a radiation beam and lived to publish the paper. So which is it — the next great infrastructure shift, or a capital trap dressed up in solar panels? In the new edition of the Physical AI Newsletter, I break down the whole field: → Key announced player — what they've raised, who's backing them, and how far they actually are from delivery → The physics that has to hold up (radiation, cooling in a vacuum, orbital congestion) → The one number the entire $trillion bet rides on: launch cost per kilogram → Why even Jeff Bezos thinks Elon's timeline is "a little ambitious" My take: it's no longer a question of *if*, but of *when cost parity arrives* — and that's a launch-economics question, not a chip one. Bull or bear on orbital compute reaching cost parity by 2030? 👇 #PhysicalAI #SpaceTech #AIInfrastructure #DataCenters #OrbitalCompute