Will the AI Boom Price the Rest of the Economy Out of Capital?
Brett Winton (ARK Invest) and Elon Musk argue that AI infrastructure's high IRR and quick payback periods could absorb massive capital, driving up economy-wide interest rates and crowding out traditional businesses The mechanism is economically sound: if AI projects yield 20% returns, aggressive borrowing by AI companies would compete with other sectors for finite capital, pushing neutral interest rates higher Federal Reserve officials (Philip Jefferson, Lisa Cook) and the Bank for International
Analysis
TL;DR
- Brett Winton (ARK Invest) and Elon Musk argue that AI infrastructure's high IRR and quick payback periods could absorb massive capital, driving up economy-wide interest rates and crowding out traditional businesses
- The mechanism is economically sound: if AI projects yield 20% returns, aggressive borrowing by AI companies would compete with other sectors for finite capital, pushing neutral interest rates higher
- Federal Reserve officials (Philip Jefferson, Lisa Cook) and the Bank for International Settlements have acknowledged this dynamic, with over $220 billion in AI-related debt issued in 2026
- The author's key counterargument: extraordinary AI returns attract competition, which historically erodes excess profits through increased capacity, falling prices, and compressed margins
- Evidence of this self-correcting mechanism: AI inference costs fell 280-fold between November 2022 and October 2024, suggesting high returns are unlikely to persist indefinitely
Why It Matters
This debate directly affects how AI practitioners, investors, and policymakers should think about the macroeconomic implications of the AI boom—specifically whether current investment enthusiasm is sustainable or whether competitive forces will normalize returns. For AI companies and infrastructure investors, understanding whether high returns will persist or compress is critical for capital allocation and valuation decisions. For traditional businesses and financial institutions, the question determines whether AI-driven rate increases could materially impact borrowing costs and investment viability outside the AI sector.
Technical Details
- Capital competition mechanism: AI companies with projected 20% returns on infrastructure would borrow aggressively, competing with governments, manufacturers, utilities, and property developers for finite savings, forcing bond yields higher until equilibrium is restored
- Neutral rate dynamics: Fed Vice Chair Philip Jefferson noted that AI-driven productivity gains could simultaneously increase business investment demand and reduce household savings (as workers anticipate higher future incomes), pushing the equilibrium real interest rate upward
- Debt market scale: Over $220 billion in AI-related debt was issued by late August 2026, contributing to total U.S. corporate bond issuance of approximately $1.68 trillion, signaling a shift from cash-flow financing to debt-market financing for AI capex
- Competitive erosion evidence: Stanford's AI Index documents that GPT-3.5-level inference costs dropped from ~$20 per million tokens (Nov 2022) to $0.07 (Oct 2024)—a 280-fold decline—while hardware price-performance improved ~30% annually, illustrating how competition destroys scarcity premiums
- Regulatory monitoring: The Bank of England's July 2026 Financial Stability Report found little evidence of crowding out so far but explicitly warned the situation could deteriorate as AI financing requirements expand
Industry Insight
- AI infrastructure investors should plan for margin compression over time rather than assuming perpetual extraordinary returns; the competitive dynamics in model development, cloud computing, chip manufacturing, and data center operations suggest a race toward commoditization
- Traditional businesses should monitor interest rate trajectories and bond market conditions as leading indicators of AI capital competition, even if their operations are unrelated to AI, since economy-wide cost of capital could rise independently of direct AI disruption
- Policymakers and central bankers should track AI debt issuance and neutral rate estimates closely, as the scale of AI investment ($220B+ in a single year) is now large enough to influence monetary policy parameters and financial stability assessments
Disclaimer: The above content is generated by AI and is for reference only.