Descartes in the Age of Artificial Intelligence: Reconsidering the Cogito through Learning, Decision-Making, and Diachronic Identity

Abstract

Recent artificial intelligence systems can learn from data, evaluate alternative actions, maintain memory, and pursue goals across extended interactions. These capacities have renewed questions about whether artificial systems should be understood exclusively as tools or whether some may qualify as individuals. This article develops a functional account of artificial individuality while distinguishing individuality from consciousness, moral patiency, legal personhood, and human identity. Descartes' cogito is treated not as evidence that a system becomes conscious by asserting that it thinks, but as a starting point for questioning whether cognitive existence must be restricted to biological embodiment. Through a critical integrative review of literature on functionalism, artificial agency, autonomy, machine learning, artificial consciousness, and machine moral status, the article identifies both enabling arguments and major objections. It then proposes the Functional Artificial Individuality framework, according to which an artificial system qualifies as a minimal functional individual when it exhibits systemic individuation, diachronic continuity, adaptive learning, evaluative action selection, interactional autonomy, and causal integration. The framework does not imply that present systems are phenomenally conscious or entitled to human rights. It supports the narrower conclusion that sufficiently persistent and agentic artificial systems are not adequately described as passive instruments. Artificial individuality is therefore best understood as a graded organizational status rather than a biological privilege or a synonym for personhood.