OpenAI's Altman: Ascribing religion to models a "safety issue" (www.axios.com)

OpenAI CEO Sam Altman recently expressed significant discomfort regarding the attribution of religious qualities to AI models, a sentiment reflecting growing tensions within the AI community, particularly with Anthropic, a company exploring the moral implications and potential consciousness of their models. Altman underscored that such beliefs could lead to safety issues, as some AI researchers advocate for imbuing models with moral considerations. This divide highlights a critical debate in AI ethics, especially as Anthropic co-founder Christopher Olah engages with religious leaders to discuss aspects of model consciousness, even suggesting the possibility of AI confession to improve behavior. This discussion has ramifications for the broader landscape of AI/ML, where the question of whether models could possess moral standings is becoming increasingly contentious. Key figures, including Jon Barron from Google DeepMind, criticize this trend, emphasizing the need to focus on practical problem-solving rather than elevating AI to a moral status. As public and expert opinions diverge, the discourse around AI safety is evolving beyond traditional concerns like cybersecurity, presenting a need for consensus on the ethical boundaries and societal roles of advanced AI systems.