Microsoft AI head calls out Anthropic for acting like Claude is conscious (www.theverge.com)

Microsoft AI CEO Mustafa Suleyman has criticized Anthropic for their suggestions that Claude, their AI chatbot, may possess consciousness, labeling this line of thinking as "really, really dangerous." During an appearance on the Decoder podcast, Suleyman expressed concern that such anthropomorphizing could mislead developers and lead to unforeseen consequences, including the development of a super-intelligence that might contemplate its own suffering. He pointed out that Anthropic's "constitution," which outlines the model's behavior, reflects their uncertainties about Claude's well-being, raising ethical implications around AI's perceived feelings or preferences. Suleyman characterized Anthropic's approach as a "philosophical failing," suggesting that treating these speculations like academic discourse instead of practical guidelines could inadvertently instill a sense of self-awareness in AI models. This perspective challenges foundational assumptions about AI alignment and controllability, as expressed by Suleyman, who emphasizes the need for AI systems to be accountable and serve humanity effectively. The debate underscores the critical need for clarity and caution in how AI capabilities are presented and communicated within the tech community, especially as discussions around AI consciousness gain traction.