From SpaceX to Founders Fund to Solving America's Nuclear Fuel Problem (Scott Nolan, General Matter)

Key points

• Career framework: 'do something useful, something that won't get solved otherwise.' SpaceX at 30 people because incumbent aerospace was stuck on cost-plus contracts with subcontractors dozens deep. Same pattern later at Founders Fund, same pattern now at General Matter.
• Peter Thiel's two core lessons for Nolan: (1) avoid the herd — think from first principles with layers of abstraction; (2) avoid competition on two axes — company-level (perfect competition compresses profits) and investor-level (crowded deals get priced up).
• Stagnation heuristic: 'cost-plus + oligopoly + no progress incentive' is the signature of industries ripe for disruption. Space launch, defense (Anduril), infrastructure (Boring Company), civilian nuclear fuel all share the DNA.
• Airbnb reinvestment story: when guests trashed a listing, the company took the insurance stance and professionalised. Founders Fund did cohort analysis across every market — market share was increasing globally. 'They're winning, it's over, they're going to win.'
• On hardware vertical integration: don't outsource engineering. First-of-kind manufacturing must be co-located with design. Subcontractor interface requirements calcify the architecture and prevent cross-layer optimisation.
• Energy-per-capita vs GDP-per-capita: R² > 0.8 across every country. Energy is the ultimate proxy for prosperity. US grid has been roughly flat since the 1990s while China has grown 3x the US total. 'If we want to continue to be economically relevant, we need more energy.'
• The 90s-to-2020 consensus — 'the market will bring energy online when needed' — has broken under AI-data-centre demand. Stranded power (West Texas wind, North Dakota flare gas) is gone; new capacity takes years; nuclear takes even longer.
• Why nuclear specifically: safest, cleanest baseload; lowest on particulates and carbon; on first principles should be among the cheapest (a pellet of fuel ≈ a ton of coal). The cost has drifted high because we've built so little, not because the physics demands it.
• Advanced-reactor taxonomy: (1) gigawatt-scale for the grid; (2) 100-300 MW SMRs — the natural fit for behind-the-meter data-centre islands over the next 5-10 years; (3) 1 MW microreactors for stranded communities currently on diesel.
• **'Bring Your Own Energy' (BYOE):** hyperscalers site data centres with co-located generation; at even a 10% over-provision on a 1GW facility, 100 MW of extra power feeds the local community grid. Speed-to-power matters more than small cost increases for AI hyperscalers.
• What kills the advanced-reactor renaissance: (1) no fuel — which is why General Matter exists, to rebuild US uranium enrichment capacity — and (2) cost blow-outs. The reactor engineering companies can handle the physics; the supply-chain side has been hollowed out.
• US has zero operational civilian uranium enrichment capacity today. All outsourced overseas for decades. General Matter's mission: restore that capability domestically as the upstream of the reactor boom.

Notable quotes

Themes

• Cost-plus industries and the repeating Thiel playbook
• Energy as the ultimate civilisational input
• The advanced-reactor renaissance's real bottleneck is fuel
• Bring Your Own Energy: hyperscalers as grid-builders