Naveen Rao: 4D Computing, AI's Energy Wall & Beating Biology

Unconventional AI claims a working non-von-Neumann chip that could cut AI compute energy use by 1,000x within 3.5 years.

• Energy is the real bottleneck: Google alone serves 3.2 quadrillion tokens/month; at his assumed 10 joules per token that works out to roughly 12 gigawatts, out of ~40GW total US data-center capacity, implying an energy shortfall within roughly three years.
• Token cost breakdown: About 50% of the cost of serving a ChatGPT token is energy; the rest is hardware CapEx and floor space.

Themes, sentiment and reasons are AI-written paraphrases, checked so they never quote an episode. They may contain errors. Not investment advice. Episodes remain the property of their creators.