Stargate / hyperscaler capex supercycle — Pick 3 takes on whether the infra spend is rational.

Options

• Rational — demand justifies build
• Bubble — $2T revenue needed to pay back
• Rational for hyperscalers only
• VC should back power/grid not apps
• Too late — trade is crowded
• Underhyped — energy is the alpha
• Labs win; startups are LNG
• Geopolitical — not financial logic
• We'll look stupid in 3 years
• Rational if AI revenue 10x
• I don't invest infra anyway
• Classic capex cycle — painful