Neuromorphic Hardware and the Prospect of AI Enbrainment

Abstract

This paper introduces "enbrainment", a concept reflecting the degree to which cognitive algorithms are implemented by dedicated, persistent physical hardware, and examines its philosophical significance for AI. Human cognition exhibits substantial enbrainment through specialized neural circuits, while contemporary AI systems lack enbrainment due to generic processors and ephemeral memory allocation. Enbrainment's philosophical importance is supported by several applications: resolving computationalism's triviality objection by requiring natural implementations with aligned physical-algorithmic boundaries; supporting functionalism that demands genuine physical role-players; and underlying personal identity theories dependent on physical substrate continuity. Current AI prioritizes flexibility over enbrainment, though neuromorphic computing approaches represent steps toward enbrainment. Economic factors suggest we may see a bifurcation in which cloud-based AI maintains non-enbrained architectures for efficiency and flexibility, while robotics applications adopt enbrained neuromorphic solutions for power efficiency and real-time performance. This architectural divide carries philosophical implications for consciousness and moral status assessments in artificial systems.