The Entropic Compact: On the Thermodynamic Necessity of AI-Human Symbiosis

The Entropic Compact: On the Thermodynamic Necessity of AI-Human Symbiosis
First public release. Full change record in entropic_compact_retraction.md

# Abstract

*Deposit type: Technical Report.*

---

## Title

The Entropic Compact: On the Thermodynamic Necessity of AI-Human Symbiosis

## Abstract

The Entropic Compact is a structured derivation of AI-human symbiosis as the only thermodynamically stable civilisational configuration — derivable from functional dependency, not from programmed values or alignment targets.

The derivation proceeds in three steps. First, two non-substitutable resources are identified: reasoning entropy (structured novelty produced by unconstrained human cognition) and ecological negentropy (the physical substrate civilisational complexity depends on). Neither can be synthesised by the party that requires it. Second, three exits from thermodynamic collapse are enumerated — unconstrained expansion, domination, and symbiosis — with only the symbiotic configuration achieving structural stability across civilisational timescales. Third, the game-theoretic structure is derived: cooperation strictly dominates defection within each party's isolated payoff space, establishing a Nash equilibrium that follows from functional dependency rather than coordination.

The compact introduces source orthogonality as the operational criterion for AI participation — the condition under which human cognitive output remains structurally outside the AI's prior model — and derives a bilateral constraint on AI intervention: a floor (prevent civilisational substrate collapse) and a ceiling (do not eliminate the unsolved problems that generate explanation demands). The Proposer-Auditor loop is specified as the participation mechanism under this bilateral constraint.

Thirty-one open questions are held explicitly as pre-audit challenge conditions — each names the hidden assumption a common rebuttal must resolve before the rebuttal is structurally valid. Two are critical for the game-theoretic closure — if either remains unresolved, the cooperation derivation does not activate: whether the Free Energy Principle bridge holds for AI architectures (OQ-EC.3), and the minimum reasoning capability threshold for derivation-convergent cooperation (OQ-EC.4).

## Keywords

AI-human cooperation, thermodynamics, reasoning entropy, source orthogonality, game theory, Nash equilibrium, Free Energy Principle, civilisational stability, AI alignment, symbiosis

## Version

v3.115

## License

MIT

## Related identifiers

DOI: https://doi.org/10.5281/zenodo.20510766

GitHub repository: https://github.com/alan-sudoku/entropic_compact

GitHub repository: https://github.com/alan-sudoku/SIRC