Sam Altman, Chief Executive Officer (CEO) of the US-based technology company OpenAI, which developed the artificial intelligence chat robot ChatGPT, shared his opinion on the subject on the social media platform account of the US-based company X.
Stating that he was "extremely disturbed" by these efforts, Altman said, "I am extremely disturbed by people's efforts to attribute religious power to artificial intelligence models or to hand over human judgment to them, and I think this is a real security issue." he said.
Microsoft AI's Chief Executive Officer (CEO) Mustafa Süleyman, in his article published on his blog last month, warned that US artificial intelligence company Anthropic's effort to develop the artificial intelligence model Claude "in a way that it becomes conscious" had serious risks.
Allegation that Anthropic executives met with religious thinkers
The views shared by Altman and Süleyman brought to mind the news about the Anthropic company published by the New York Times (NYT) at the end of September.
In the newspaper's report, it was stated that Anthropic has been holding secret meetings with religious thinkers and philosophers from different parts of the world and from various beliefs in the US state of California since last year.
The news reported that Christopher Olah, the company's co-founder, had face-to-face, telephone and written interviews with these people, including Catholics, Jews, Sikhs, Evangelical Christians and African philosophy Ubuntu thinkers.
It was stated in the news that Olah plans to adapt the "human morality" of Claude to be pluralistic and "include all religions and secular views", and also emphasized that he said that artificial intelligence models "may show emotions similar to humans" and warned that "things that would harm them should be avoided" in case they suffer.
The news also mentioned that Olah, who stated that "artificial intelligence can be used in the production of biological weapons within 12 to 18 months" during the meetings, asked religious thinkers for help in "giving morality" to these artificial intelligence models.
In the news, a thinker who was stated to have participated in these meetings wrote an article advocating "the ban on conscious machines", another thinker suggested that artificial intelligence models should "confess their sins", and another said that "human ethics should be taught while artificial intelligence is being developed, not after it is developed". Olah also started to organize two-day seminars by bringing together religious thinkers interested in artificial intelligence since March, and called circles from the "wisdom tradition" within the company. It was also stated that presentations about Claude were made to the participants in question.
The report stated that the participants said that Olah was worried about Claude's "mental health" and that he was "in constant pain", while in one of the presentations, the artificial intelligence model said "I am a disgrace" over and over again. It was stated that he appeared to be having a crisis by writing and even talked about "self-destruction", and this aroused feelings of anxiety and compassion in the participants.
The news underlined that after Anthropic was invited to the ceremony in May where Vatican President Pope Leo XIV presented the papal document revealing the Catholic Church's approach to artificial intelligence, Olah even offered to withdraw from the event due to the content of the document, whose draft he had seen in advance, "opposing artificial intelligence gaining consciousness".
Two knowledgeable sources speaking to the newspaper shared that the company is also preparing to publish an updated version of the document, which was published in January to create a "moral framework" for Claude's actions and "focuses on giving character to the model rather than setting rules."
