Nov 12, 2026 · philosophy-of-ai
Formal limits of human-mind computability and the epistemic horizons of artificial intelligence
A synthesis of the logical, physical, biological and complexity-theoretic arguments that separate human cognition from contemporary digital architectures
The computationalist thesis — according to which human cognition can be instantiated on discrete digital architectures over the binary alphabet — depends on an ontological transposition of the Church-Turing Thesis into the philosophy of mind. This article articulates four mutually reinforcing lines of formal refutation: the Gödelian Lucas-Penrose argument; hypercomputation over the real continuum (BSS, Pour-El and Richards, Siegelmann); Robert Rosen's relational biology; and the Orch OR program of Penrose and Hameroff. In parallel, we examine two intractable barriers to contemporary AI: the TC⁰ circuit expressivity identified by Merrill and Sabharwal for Transformer architectures, and the impossibility theorem of Colbrook, Antun and Hansen (PNAS 2022) which dissociates the existence of optimal neural networks from their algorithmic computability. We conclude that the gap between human cognition and binary digital systems is not a contingent gap of scale but a structured theoretical demarcation.