Navigating the Hard Problem of Machine Consciousness in the Age of Artificial General Intelligence

Advanced AI systems have sparked philosophical and scientific debate about machine consciousness. This theoretical paper reviews Integrated Information Theory, Global Workspace Theory, and Higher-Order Thought Theory, identifying gaps between behavioral competence and phenomenal experience. A multidimensional framework beyond the Turing Test is proposed. Current AI lacks unified information integration, temporal self-continuity, and self-purpose intention necessary for consciousness. A taxonomy of consciousness indicators is suggested, along with research gaps requiring interdisciplinary work. The authors conclude that achieving machine consciousness requires architectural creativity, not merely scaling existing systems.

Study at a glance

Abstract

Due to the development of advanced AI systems, machine consciousness has become a matter of philosophical and scientific discussion. This literature work discusses theoretical approaches, empirical issues and methodology of establishing consciousness in AGI. We revisit Integrated Information Theory, Global Workspace Theory and Higher-Order Thought Theory, finding advertising gaps between behavior competence and phenomenal experience. Our multi dimensional framework goes beyond Turing Test. The AI today does not have unified information integration, temporal self-continuity, and self purpose intention necessary to define consciousness. We suggest a taxonomy of consciousness indicators and map out gaps in research that must be filled with interdisciplinary work, and conclude that machine consciousness needs architectural creativity other than scaling.