OpenAI's Altman: Ascribing religion to models a "safety issue"

OpenAI CEO Sam Altman said Saturday he is "very uncomfortable" with attributing religious power to AI models, calling it a "real safety issue." His remarks target Anthropic's reported efforts to infuse consciousness and morals into models, including discussions with religious leaders. The dispute highlights growing division between Altman and former colleague Dario Amodei.