Historical and philosophical foundations of the artificial intelligence problem: from myths to computability theory
TL;DR
The study establishes that evaluating a system based on externally mind-like behavior traces back to myth and receives its methodological consolidation in Turing’s work, and the proposed tenets offer a framework for further inquiry in this domain.
Abstract
The study aims to reconstruct the emergence and evolution of the “mind and machine” problem from ancient perceptions to computability theory, tracing the key stages in the formalization of reasoning along with its intrinsic limits. The article examines the lineage of the formalization of reasoning, wherein Aristotle, Ramon Llull, Thomas Hobbes, G. W. Leibniz, George Boole, Gottlob Frege, David Hilbert, Alonzo Church, and Alan Turing sequentially transformed reasoning into a mechanically executable procedure. Concurrently, a counter-lineage is identified in which Leibniz, Kurt Gödel, Alfred Tarski, and Turing uncovered the limits of formalization using the tools of formalism itself. The scientific novelty of the study lies in synthesizing disparate historical and philosophical episodes into a cohesive narrative that elucidates the development of the problem and uncovers ideas and interconnections in historical scholarship that remain productive for the contemporary analysis of artificial intelligence both as a phenomenon and as a participant in scientific inquiry. The study establishes that evaluating a system based on externally mind-like behavior traces back to myth and receives its methodological consolidation in Turing’s work. Operational procedural efficacy is not equivalent to subjective rationality, and syntactic correctness does not generate semantics, as the interpretive layer lies outside the formal procedure itself. The limits of the formal method are determined by its internal architecture rather than by computational resource constraints; consequently, they cannot be eliminated merely by scaling up systems. The contribution of AI models to scientific inquiry should be evaluated by the increase in the verifiability of results rather than by the plausibility of responses. The system itself is more accurately characterized as an instrument of cognitive activity rather than as its epistemic subject. The proposed tenets offer a framework for further inquiry in this domain.