4E cognition and the coevolution of human–AI interaction
Discover Artificial Intelligence November 13, 2025 Jörg Noller
This paper argues that large language models (LLMs) should be understood not as objects but as processual and relational phenomena that co-constitute human agency within a shared sociotechnical environment. Drawing on 4E cognition (embodied, embedded, enactive, and extended), the authors propose reframing AI as an interactive medium of enactive and extended human action. They offer a 4E-compatible connectionist account that views AI as a co-evolving component of the extended cognitive ecology, shaped by and shaping enactive practices, intentions, and norms. The paper also discusses AI's limits, addressing hallucination and model collapse from this perspective.