Work

The Compute-to-Cognition Asymmetry: Thermodynamic Limits, Value-Chain Hollowing, and Physical Compute Governance in the Post-Subsidy AI Economy

Year: 2026
Type: preprint
Abstract: The provided research literature examines the Compute-to-Cognition Asymmetry, a structural macroeconomic and thermodynamic paradox wherein deploying energy-intensive silicon architectures to automate low-dimensional human perceptual tasks generates diminishing economic returns. By modeling the Probabilistic Verification Tax, the study demonstrates how non-deterministic error rates and mandatory human auditing impose hidden operational costs that outweigh initial labor savings once venture-backed Token Subsidy Cliffs expire. Furthermore, unrestricted perceptual automation erodes strategic firm moats through Bertrand Commoditization and suppresses aggregate demand by hollowing out middle-class wage shares. To counteract these systemic failures, the author proposes Physical Compute Governance via an international treaty framework that caps national data center facilities and megawatt quotas, redirecting scarce computational resources into Domain-Chartered Compute Specialization across high-dimensional scientific and sustainable development goals.
Author: Rai, Ayush
Language: English
Open Access status: green
OpenAlex ID: W7234538103
Cites:
Cited by:
Related to: 10