2026 年 33 巻 1 号 p. 73-90
Humans make decisions and inferences that violate Bayesian (classical) probability, such as conjunction/disjunction fallacies in probabilistic judgment and order effects on judgments. This has led to the proposal of cognitive models using quantum probability theory. Quantum probability theory refers to rules for assigning probabilities to events based on quantum mechanics, without incorporating any elements of physics. However, no answer has been found to the question of why human cognition follows the laws of quantum probability. Living organisms are nonlinear, nonequilibrium, open many-body systems exhibiting duality of particle and wave properties. Biological entanglement, the principle of indeterminacy, context dependency, and path integrals are observed in life phenomena. We developed an organism mechanics theory to explain the quantum-like properties found in empirical observations through natural principles. This paper first discusses the quantum extension of this organism mechanics theory, then presents a cognitive model based on this quantum extension. It also demonstrates that this new cognitive model can model five functions currently lacking in AI: purposeful, spontaneous engagement with reality; sensory experience of the real world; a biological body; autonomous learning; and development that gradually matures through accumulated experience.