Predictive Emotional Selfhood in Artificial Minds: Toward Emotion-Cognition Integrated AI

This paper proposes PESAM (Predictive Emotional Selfhood in Artificial Minds), a novel AI framework integrating predictive processing, emotional simulation, and dynamic self-modeling. By combining neuroscientific theories such as the somatic marker hypothesis and active inference, PESAM aims to create artificial agents capable of emotion-cognition integration and embodied selfhood, offering implications for artificial general intelligence and human–AI interaction.