Heuristic Compression Model of Emotion (HCME): A Paradigm for Emotion-Informed AI Architecture

This paper introduces the Heuristic Compression Model of Emotion (HCME), a paradigm-shifting proposal for affective cognitive architecture grounded in information theory. Unlike traditional AI systems, which prioritize abstract symbolic reasoning, HCME models emotion as a lossy compression layer — a biologically inspired affective substrate that heuristically ranks and filters decision pathways under resource constraints. Drawing from evolutionary neuroanatomy, HCME is positioned as the architectural analog of the amygdala within artificial systems: a pre-rational affective core that informs, compresses, and accelerates neocortical-like reasoning modules. This paper outlines HCME’s computational role in cognitive prioritization, offers concrete strategies for engineering implementation (including affect-tagged inference, compression-based attention allocation, and introspective feedback loops), and situates the model as a practical solution to bounded rationality. By bridging emotion theory and AI design, HCME offers a foundational scaffold for building synthetic minds capable of prioritizing meaningfully, not merely computing exhaustively. As such, HCME may provide a missing architectural substrate necessary for constructing general-purpose, adaptive AI systems.