PSY 672 Cognition Beyond Computation: Rethinking AI

The computational theory of mind has served as the dominant framework in cognitive science for over half a century, grounding explanations of perception, memory, reasoning, and language in the logic of symbol manipulation and information processing. Contemporary AI systems, particularly large-scale neural architectures, inherit and extend this tradition while also exposing its conceptual limits in ways that demand rigorous theoretical and practical examination. This course provides a systematic analysis of computational models of cognition and their adequacy as accounts of mental phenomena. Core topics include the classical symbolic paradigm and its critics; connectionism and the debate over representational content; predictive processing and active inference as alternatives to feedforward computation; the explanatory gap between functional and phenomenal accounts of consciousness; embodied, enactive, and ecological approaches to cognition; and the epistemological status of transformer-based language models with respect to meaning, reference, and understanding. Additional topics include formal and empirical challenges to machine sentience claims and the ethical implications of deploying systems whose internal representations remain poorly understood.