Artificial Intelligence: Prospects for Consciousness Emergence

Whether an artificial intelligence system can attain human consciousness is examined by critically reviewing neuroscience, anthropology, and philosophy. The authors distinguish theoretical from instrumental approaches to assessing AI consciousness. Drawing on Global Workspace Theory, Integrated Information Theory, Spatiotemporal Theory, and Evolutionary Theory of Consciousness, they argue that consciousness in a material system requires an integral architecture, adaptability, and self-control. Instrumental tests can only evaluate consciousness through real-world behavior, revealing limitations and risks of anthropomorphic perceptions. The authors advocate synthesizing theoretical and instrumental approaches, propose clear criteria for determining AI consciousness, and consider ethical implications and potential challenges if AI systems become conscious.

Study at a glance

Abstract

The article explores the question of whether an artificial intelligence (AI) system can attain human consciousness. The authors critically examine the current state of research in neuroscience, anthropology, and philosophy, highlighting the sig­nificant challenges posed by the limitations of computational power and the ar­chitecture of modern AI systems. The article proposes a distinction between theoretical and instrumental approaches to determining the level of conscious­ness in AI. By considering the most authoritative theories of consciousness in neuroscience and cognitive philosophy, such as the Global Workspace Theory, Integrated Information Theory, Spatiotemporal Theory, and Evolutionary Theory of Consciousness, the authors argue that the emergence of consciousness in a material system requires an integral architecture, adaptability, and the ability to master self-control. An analysis of instrumental tests for the presence of con­sciousness shows that consciousness can only be assessed based on behavior in the real world, and also demonstrates the limitations and risks of anthropo­morphic perceptions of the object of study. The authors advocate for the synthe­sis of theoretical and instrumental approaches in the research and design of AI, emphasizing the importance of interdisciplinary studies to advance the under­standing and potential replication of human consciousness in artificial systems. Clear criteria for determining the presence of consciousness in AI systems are proposed, and the ethical implications of this technological achievement are con­sidered. The study concludes with a discussion of the potential challenges huma­nity may face if AI systems are endowed with consciousness, stressing the im­portance of a cautious and well-considered approach to such a complex issue.