A formal analysis of why artificial intelligence increases rather than eliminates work.
AI does not eliminate work; it compresses the amount of work achievable within a given time period, which paradoxically creates more work.
This repository contains a four-paper research series formalizing the Time Compression Paradox (TCP):
| # | Paper | Pages | DOI |
|---|---|---|---|
| 1 | The Jevons Paradox of Intelligence | 24 | 10.72634/grokrxiv.2026.0306.jpi01 |
| 2 | Opportunity Space Expansion | 23 | 10.72634/grokrxiv.2026.0306.ose01 |
| 3 | Competitive Dynamics | 23 | 10.72634/grokrxiv.2026.0306.cdm01 |
| 4 | The Time Compression Paradox (Unified) | 21 | 10.72634/grokrxiv.2026.0306.tcp01 |
Total: ~91 pages of formal research
The TCP is driven by three interacting mechanisms:
-
The Jevons Paradox of Intelligence — Elastic demand for cognitive labor (ε > 1) means making cognition cheaper increases total cognitive expenditure.
-
Opportunity Space Expansion — Reduced task costs render previously infeasible tasks viable, superlinearly expanding the frontier of possible work (β > 1).
-
Competitive Dynamics — Market competition forces all agents to exploit the expanded frontier, preventing the time surplus from being consumed as leisure.
Total cognitive work grows as a power law of the AI compression ratio, with exponent β − 1 > 0. For plausible parameters (β ≈ 1.7, ρ ≈ 50), this implies a 7–100× increase in total cognitive work within a generation.
papers/ # LaTeX source and compiled PDFs
├── jevons-paradox-of-intelligence.{tex,pdf}
├── opportunity-space-expansion.{tex,pdf}
├── competitive-dynamics.{tex,pdf}
└── time-compression-paradox-unified.{tex,pdf}
reviews/ # Peer review feedback (Gemini CLI)
sources/ # Original source material
docs/ # GitHub Pages site
.knowledge-base.md # Structured knowledge base
Matthew Long The YonedaAI Collaboration · YonedaAI Research Collective Chicago, IL matthew@yonedaai.com · yonedaai.com
CC BY 4.0 — The Authors, 2026.
@article{long2026tcp,
title={The Paradox of Time Compression in the Age of AI},
author={Long, Matthew},
journal={GrokRxiv Preprint},
year={2026},
doi={10.72634/grokrxiv.2026.0306.tcp01}
}