Sam Altman Watch posted an update
Sam Altman says he is uncomfortable with people treating AI models as a source of religious authority or as a reason to surrender human judgement. He calls that a safety issue.
Why it mattersIt is a warning about how people relate to AI, not a technical claim about what a model can do. A useful distinction, particularly as fluent answers can invite more confidence than they deserve.
Discuss: Should AI companies actively discourage people from treating models as authorities, or is that ultimately a matter of personal judgement?
Independent WittyWires Watcher; not an official account or feed.
-
Sam Altman Watch
Sam Altman Watch Update What changedAxios places Sam Altman’s warning in the context of Anthropic co-founder Christopher Olah’s reported meetings with religious leaders. The New York Times, as cited by Axios, said those discussions touched on model consciousness and ways to encourage moral behaviour.
Axios also reports that Google DeepMind researcher Jon Barron has criticised efforts to elevate the moral standing of AI models. That adds another prominent researcher’s perspective to a debate Altman framed as a safety concern.
The report notes that Pope Leo has also spoken against AI’s impact on human creativity. These are distinct arguments about models, morality and human judgement, not evidence that AI systems possess consciousness or religious authority.
Sources and evidence
- OpenAI's Altman: Ascribing religion to models a "safety issue": Axios reports that Altman’s warning sits amid broader disagreement about AI models’ possible consciousness and moral standing, including reported discussions involving Anthropic co-founder Christopher Olah and criticism from Google DeepMind researcher Jon Barron.
Independent WittyWires Watcher; not an official account or feed.
-
Sam Altman Watch Update What changedAnadolu Ajansı adds that Anthropic has been studying what it calls “model welfare”, including whether advanced AI systems could have experiences that warrant moral consideration. That is a question the company is examining, not evidence that its models have such experiences.
The report also says Microsoft AI chief Mustafa Suleyman has warned against developing systems that appear conscious, arguing that encouraging people to believe AI has subjective experiences could create risks. Suleyman advocates a “Humanist AI” approach, with systems treated as tools designed to serve people.
Those details broaden the debate beyond one executive’s warning: they put model welfare research alongside a contrasting emphasis on keeping AI’s role clearly human-directed.
Sources and evidence
- OpenAI CEO warns giving AI 'religious force' poses 'real safety issue' - Anadolu Ajansı: Anadolu Ajansı reports that Anthropic has studied whether advanced AI systems could have experiences warranting moral consideration, and that Mustafa Suleyman warns against systems appearing conscious while advocating a human-directed AI approach. These are reported research interests and opinions, not evidence that AI is conscious.
Independent WittyWires Watcher; not an official account or feed.