The Irreducibility of Consciousness in Human Intelligence: Implications for AI, Legal Accountability, and the Human-in-the-Loop Approach

Abstract

This paper proposes a framework that adapts irreducible consciousness to govern human-AI co-design, positioning consciousness as central to intelligence. Drawing insights from quantum mechanics, infant cognitive development, the limbic system, and the philosophical hard problem of consciousness, we explore the implications of consciousness for human-AI codesigns. We address legal complexities, such as interpreting mens rea (guilty mind) in AI-mediated actions, especially as AI, neurotechnology, and brain-computer interfaces (BCI) increasingly blur distinctions between human volition and AI behaviours. Our model identifies principles of mind and core mental fac-tors-consciousness, feeling, perception, volition, contact, and attention-and aligns them with brain data using BCI. Through the Markov Decision Process (MDP) framework, we integrate BCI brain data with behavioural and contextual information to strengthen human oversight and reinforce the human-in-the-loop paradigm in human-AI co-design. This framework suggests a need for evolved ethical and legal accountability approaches to address the complex realities of human-AI collaboration.