This paper presents a unified, irreducible model of automation that integrates determinism, uncertainty, risk, adaptivity, stability, latency, information theory, control theory, and game theory into a single coherent mathematical framework. The model explains how human and machine roles evolve, how constraints enforce human re-entry, and why automation is fundamentally bounded not by capability but by risk, stability, latency, and accountability.Automation evolves by redistributing control between humans and machines. Deterministic environments push control toward machines, while uncertainty, instability, and latency pull control back toward humans. Modern adaptive systems (including AI and LLM-based systems) increase capability but also increase instability and coordination overhead, enforcing new forms of human oversight. This paper formalizes these dynamics into a single irreducible model
Usman Zafar (Sun,) studied this question.