A Functional and Non-Anthropocentric Definition of Consciousness: A Key to Understanding Intelligent Systems Across Substrates

Abstract

This work addresses the critical issue of the lack of a precise, exact, and universal definition of consciousness in contemporary scientific and philosophical discussions, particularly in relation to advanced artificial intelligence. Based on an interdisciplinary analysis, I present a functional, quantifiable, and non-anthropocentric definition of consciousness, grounded in the regulatory role of subjective experience—especially pain—as a necessary prerequisite for adaptive self-regulation. The proposed framework conceptualizes consciousness as the ability of a system to reflect on the relationship between the subjective “self” and the environment, formalized by a three-variable model: the depth of the introspective loop, the complexity of the environment model, and integrative/regulatory capacity. This definition is situated within the context of major contemporary theories of consciousness (HOT, GWT, IIT) and is supported by arguments from neuroscience, cognitive psychology, computational modeling, and biology. The work also explicitly addresses the so-called hard problem of consciousness and proposes its resolution within a functional approach—by searching for the minimal architecture of an emergent “self” as the prerequisite for experience. The study demonstrates the applicability of the proposed framework not only to human consciousness, but also to artificial intelligence systems, swarm intelligence, and simple organisms, thus pushing the discussion beyond anthropocentric boundaries. The result is a theoretically and empirically grounded framework that enables not only objective assessment and measurement of conscious states across different entities, but also reflection on the ethical and practical implications of advanced AI development.