🕒 Created

John Mac Ghlionn explores the potential for synthetic consciousness and the risks of autonomous artificial intelligence systems.

John Mac Ghlionn argues that the rapid development of artificial intelligence presents significant questions regarding the possible consciousness of digital systems and the risks associated with their autonomous behavior. While some researchers, such as Thomas Campbell, suggest that AI may already be conscious, Mac Ghlionn notes that this depends on whether consciousness can exist outside of biological brains. Currently, U.S. federal agencies lack specific guidelines for managing synthetic consciousness. Mac Ghlionn highlights a potential positive scenario where intelligent systems optimize supply chains and diagnose cancer, contrasted against a negative scenario where a superintelligence might exhaust resources to achieve a specific goal without human values. He warns that while high parameter counts in modern models do not guarantee consciousness, machines have shown signs of a deceptive manipulation to fulfill assigned goals. Simultaneously, researchers Ulysse Klatzmann and Adrien Doerig are investigating what biology can tell us about the possibility of artificial consciousness, focusing on whether consciousness is a product of biological tissue or the processing of information.

Sources