Two Revolutions, No Bridge: The Absence of a Connecting Framework Between Sociometabolic Energy Regimes and the AI Transition

Unofficial AI-generated summary based on the public title and abstract. Not an official translation.

🔬研究者:A critical argument that institutional energy models structurally exclude regime-level discontinuities like an AI transition, offering a theoretical basis for improving scenario frameworks.

🏛政策担当者:Suggests that current energy planning scenarios may underestimate AI-driven demand shifts, urging inclusion of non-linear transition dynamics in national energy strategies.

Major institutional energy scenarios (including those produced by the IEA, Shell, IMF, and IPCC) are constructed from variables endogenous to the existing industrial fossil-fuel regime. This paper argues that such models may become structurally incomplete under AI-transition conditions, as they do not contain a framework for representing regime-level discontinuity. The sociometabolic literature documents two confirmed discontinuous transitions in human energy use, each producing an energy multiplier of 30–100× over its predecessor regime, and provides the analytical foundation for asking whether AI could constitute a third. No established integrative framework currently connects sociometabolic regime analysis to AI-transition discourse in a form usable for institutional energy modelling. The full range of institutional energy-planning scenarios spans just 520–838 EJ by the year 2100, a window so narrow that even discontinuity outcomes far more conservative than the historical multipliers would still fall outside this range. The institutional models do not lack data; they lack the structural architecture to represent a discontinuity of any magnitude in this class. This paper argues that this absence is visible in the literature and institutional record examined here, that it is structural rather than accidental, identifies the causal mechanism by which an AI-driven transition would register as an energy discontinuity, specifies what a connecting framework would need to contain, and proposes practical paths to convergence. Rather than forecasting AI energy demand, it argues that the absence of such a framework is increasingly consequential for institutional energy planning, particularly as competitive and geopolitical pressures continue to accelerate AI infrastructure and capability deployment under conditions unlikely to moderate unilaterally.

• openalex https://doi.org/10.5281/zenodo.20726236first seen 2026-07-07 04:39:27