Leopold Aschenbrenner reportedly lost 67% in a month and remained up on the year. The consensus read is that his AI thesis was right and his leverage was wrong. This report argues the deeper error is shared by the entire market: AI can transform production without rescuing consumption. A 22-chapter structural analysis of the Intelligence–Demand Paradox — why near-infinite productive capacity may collide with structurally weak demand, and what could prevent it.
Executive summary: seven findings
- 1
AI expands productive capacity far more directly and far more quickly than it expands consumer demand. The two are not the same variable, and markets are modelling only one of them.
- 2
Corporate AI savings do not automatically become wages, lower prices or new investment. Where they land depends on bargaining power, market structure, debt and ownership — not on the technology.
- 3
Simultaneous automation is self-cancelling at the level of the system: one company's removed cost is another household's removed income and another firm's removed revenue.
- 4
Debt obligations are fixed in nominal terms while labour income is becoming more variable. A technology that makes earnings less predictable, inside economies carrying record debt, is a financial-stability question, not merely a productivity question.
- 5
AI investment concentrates spending into a narrow cluster — chips, compute, power, land, cooling — while compressing revenue across the diversified supplier ecosystems that employ most people.
- 6
The binding constraint on growth may be shifting from production to customers: purchasing power, attention and creditworthiness are the genuinely scarce inputs of an AI economy.
- 7
The greatest economic risk from artificial intelligence is not technical failure. It is successful deployment inside an unequal, overleveraged and demand-constrained system.
Read the full intelligence report — free
Enter your email to unlock the complete analysis, forecasts and sources, and download the PDF.
- 01The $16 Billion Warning
- 02Leopold Was Not Wrong About AI
- 03The Intelligence–Demand Paradox
- 04The £100,000 Saving That Never Reaches the Economy
- 05The Automation Fallacy of Composition
- 06The Missing Customer
- 07Productivity Without Wages
- 08The Fixed-Debt, Variable-Income Trap
- 09The AI Capital Funnel
- 10Lower Costs Do Not Mean Lower Prices
- 11Sector by Sector: Where the Paradox Bites First
- 12Abundance Meets Scarce Attention
- 13Capitalism's Worker–Consumer Contradiction
- 14Lessons From Previous Technological Revolutions
- 15The Geopolitics of Artificial Abundance
- 16The Two-Speed AI Inflation
- 17The Capital Concentration Loop
- 18The Optimistic Case
- 19What Could Prevent a Demand Crisis?
- 20The Crisis After the Productivity Boom
- 21What Would Falsify This Thesis
- 22Methodology, limitations and glossary
Already a member? Sign in to read everything instantly.