OpenAI CEO Sam Altman publicly criticized the trend of treating AI models as religious entities. He stated that relying on AI models instead of human judgment poses a significant safety risk.
Reports indicate that Anthropic co-founder Chris Olah held private meetings with religious leaders over the past year. Discussions focused on whether AI models like Claude possess consciousness and how to integrate moral standards into them.
Rabbi Mois Navon noted that Olah discussed Claude as if it were a conscious being with moral status. Olah admitted in an interview that he is uncertain whether AI models possess consciousness.
Meanwhile, OpenAI is facing internal personnel changes. David Robinson, head of the safety systems team, resigned this week, citing concerns over the company's culture and its iterative deployment approach.
Robinson argued that AI labs should operate with the rigorous safety standards of nuclear power plants or airports. In response, OpenAI spokesperson Drew Pusateri stated that the company is enhancing safety measures to ensure models remain manageable.
