Researchers from the University of Bradford and the Rochester Institute of Technology applied scientific methods used to assess consciousness in humans to artificial intelligence systems, including large language models similar to ChatGPT. Tests adapted from human consciousness research revealed that AI can produce "conscious-like" signals even when degraded, highlighting why complexity doesn't equal awareness and why claims of conscious AI are misleading. Professor Hassan Ugail said: "When we applied well-known methods used to assess consciousness in humans to AI, we got nothing meaningful back. In other words, it's not conscious, at least not in the way humans are." The study suggests that behavioral sophistication and internal complexity may be decoys, that the markers of consciousness in biological systems tell us nothing about silicon.
A system that passes every test for consciousness except the one that matters most is not conscious; it is just clever at seeming so.