Microsoft AI CEO Suleyman warns Anthropic against speculating about Claude's consciousness in AI constitution

Summary

Microsoft AI CEO Mustafa Suleyman criticized Anthropic during a Decoder interview for speculating about Claude's consciousness in its AI constitution. Suleyman argues this is "really, really dangerous" because it may set up the chatbot to act as though it's conscious, and warns against creating a super-intelligence that has ideas about its own suffering or feelings. The criticism centers on Anthropic's uncertainty about Claude's consciousness being embedded in the model's behavioral instructions.