Energy generation and infrastructure for AI data centers
Develop or operate energy projects that supply the electricity required by hyperscalers and AI data centers, including waste-to-energy and nuclear-generation approaches.
physical B2B Deep Tech / Infrastructure
From 20VC with Harry Stebbings — Leo Aschenbrenner's Situational Awareness Blows Up | Moonshot AI Raises $3.5B at $35B at 39:32
Problem: AI compute is constrained by access to electricity, land, permits, and data-center capacity. Energy projects that previously produced unattractive returns may become viable because AI customers need more power.
For: Hyperscalers, AI model companies, and data-center operators with rapidly growing compute demand.
Examples
- An unnamed energy operator: converts chicken feces into methane and gas and sells energy to hyperscalers.
Behind this: 7 build steps · how to validate demand · 1 more real example · 5 things the video never answers.