The Challenge of Integrating Artificial Intelligence into Clinical Reasoning by Augmenting the Physician's Central Cognitive Role.
Artificial intelligence is reshaping how physicians reason, yet the cognitive dimension of sustained physician-LLM interaction remains conceptually undefined. This Commentary proposes that repeated, methodologically guided dialogue with personalised large language models-termed Clinical Cognitive Interaction-progressively imprints the physician's epistemic signature onto a personalised exocortex, giving rise to an emergent Human-LLM Cognitive Unit. Whether this interaction promotes cognitive flattening and deskilling or genuinely augments clinical reasoning depends not on the technology itself but on the interaction's epistemic quality. We argue that medical education must move beyond AI literacy to teaching Clinical Cognitive Interaction as a core professional competency, thereby augmenting rather than diminishing the physician's central cognitive role.