AI Data Centers: Delete the Grid Before Orbital Compute скачать в хорошем качестве

AI Data Centers: Delete the Grid Before Orbital Compute

Sam Altman has dismissed orbital data centres as unrealistic at scale this decade. Elon Musk is openly exploring orbital AI compute as the next frontier. But what if both sides are looking in the wrong place? In this episode of ‪@overthehorizon‬, Casey Handmer and Brian Wang break down the real bottleneck in AI scaling. It's not chips. Not orbit. It's not even energy generation. It is the grid. Casey’s argument from his latest blog post on Direct Current Data Centres is simple but radical: Solar → Batteries → GPUs All direct current. No grid dependency. No gas turbine bottlenecks. No waiting years for interconnect approval. If GPUs dominate the cost stack, then massively overbuilding solar barely moves the total system cost. The fastest way to scale AI compute may be to bypass the grid entirely. 🚨⚡Link to Casey's blog: https://caseyhandmer.wordpress.com/20... We also examine: 📌 Sam Altman’s comments on orbital data centres 📌 SpaceX and Elon Musk’s push toward space-based infrastructure 📌 Why permitting and EPC friction are the hidden constraints 📌 The 10X solar overbuild strategy 📌 Whether Earth-based DC infrastructure beats orbit this decade This is not just a debate about space versus Earth. It is a debate about industrial execution versus bureaucracy. #AIDataCenters #ElonMusk #SamAltman #OrbitalDataCenters #ArtificialIntelligence #SolarPower #EnergyInfrastructure #SpaceX #DataCenterEnergy #DirectCurrent #overthehorizonpodcast