The limits of large language models and the necessity of human cognition in K-12 education

This article explores the distinctive qualities of human cognition in comparison to large language models (LLMs), focusing on the implications of each for K-12 education. Drawing on insights from cognitive science and phenomenology, we argue that human cognition — grounded in embodied experience, social interaction, and self-consciousness — cannot be fully replicated by machine models. These unique qualities suggest the need for humanistic education: teaching rooted in action, subjectivity, and self-consciousness, aimed at the cultivation of virtue. This study contributes to the broader discussion on AI in education by emphasizing the irreplaceable aspects of human experience and highlighting what human-centric instruction looks like in the era of AI.