Beyond Alignment: AI as Hormē-Enhancement Tools in a Thermodynamic Framework

AbstractThe discourse on Artificial Intelligence is paralyzed by the “agency mistake”: the assumption thatcomplex, goal-directed behavior implies agency, leading to intractable pseudoproblems like valuealignment and control. This paper reframes the debate from the ground up. First, it establishesfrom computer science and physics that AI systems are deterministic state machines, executingscripts that are causally closed and semantically empty. Second, drawing on the Neo-Pre-PlatonicNaturalism (NPN) framework, it defines agency via Hormē: the thermodynamic, constitutivestriving of a far-from-equilibrium system to persist. AI fails the Hormē test; it is a tool, notan agent. The real danger is not misaligned AI agency, but the obfuscated amplification ofhuman Hormē. We propose the Hormē-Enhancement Paradigm: the ethical purpose of AI is toaugment human capacities to navigate reality. This dissolves the pseudoproblems, redirects focusto accountability and tool safety, and offers a clear, productive future for the ethical developmentof intelligent technology.