A formal analysis of why artificial intelligence increases rather than eliminates work. Four papers, ninety-one pages, one unified theory.
The paradox is driven by three interacting economic mechanisms that ensure total work increases monotonically with AI capability.
Elastic demand for cognitive labor (ε > 1) means that making cognition cheaper increases total cognitive expenditure. Just as cheaper coal led to more coal consumption, cheaper intelligence leads to more intelligence consumption.
Reduced task costs render previously infeasible tasks viable, superlinearly expanding the frontier of possible work. The opportunity space grows as ρβ with β > 1 due to the combinatorial nature of newly feasible task combinations.
Market competition forces all agents to exploit the expanded frontier, preventing the time surplus from being consumed as leisure. The competitive ratchet and expectation escalation ensure work expands to fill compressed time.
Each paper can be read online or downloaded as PDF.
Elastic demand and the paradox of cognitive abundance. Formalizes intelligence as an economic resource subject to the Jevons Paradox, proving that total cognitive expenditure increases as AI reduces cognitive price.
A superlinear theory of work frontier growth. Proves that the opportunity space of feasible tasks grows as ρβ with β > 1, deriving the work multiplier bound MW ≥ ρβ−1.
Why markets prevent the AI leisure dividend. Game-theoretic analysis proving that Nash equilibrium forces universal AI adoption, with output scaling linearly in the compression ratio.
The capstone. Synthesizes all three mechanisms into a complete dynamical system, proves monotonicity of total work, and presents thirty-year numerical simulations predicting a 10–100× increase in cognitive labor.
Every major productivity revolution has followed the same pattern: efficiency gains lead to increased total consumption, not decreased.
| Technology | Efficiency Gain | Total Use | Implied β |
|---|---|---|---|
| Printing Press (1440) | Copy time ↓↓ | Books ↑↑↑ | — |
| Steam / Coal (1830–1900) | 3× | 10× | 2.1 |
| Electricity (1920–1970) | 5× | 50× | 2.4 |
| Computing (1970–2010) | 106× | 109× | 1.5 |
| Telecom (1990–2020) | 104× | 107× | 1.75 |
| AI (2020–) | Cognition cost ↓↓↓ | Cognitive work ↑↑↑↑ | ~1.7 |
Interactive visualization of the TCP dynamical system, Jevons backfire effect, and parameter sensitivity analysis. Built with Julia + DifferentialEquations.jl.
View Simulation Results →