Sam Altman recently called space-based data centers 'trash,' and most experts agree. In this episode, Lucas and Luna unpack why the idea of beaming clean energy from orbit to power AI computing is scientifically sound but economically implausible. They examine the physics of RF attenuation, the staggering cost per watt of orbital infrastructure, and why terrestrial nuclear and geothermal are better bets. Along the way, they touch on General Fusion's public debut as the first fusion company to trade, and what that signals about investor appetite for long-shot energy bets. The conversation lands on a bigger question: as tech giants build their own chip supply chains and model training costs balloon, where will the electricity come from? A concise, numbers-driven look at the intersection of AI scaling and energy reality.