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

OpenAI CEO Sam Altman recently expressed discomfort with the notion of granting religious qualities to artificial intelligence models. This stance is in contrast to Anthropic, a company that reportedly has enlisted scholars from various faiths to advocate for moral and conscious AI systems. Altman's perspective highlights a growing divide within the AI community regarding the potential spiritual implications of AI development.

In a statement posted on X, Altman cautioned that ascribing religious force or surrendering human judgment to AI models poses a significant safety issue. The New York Times reported that Anthropic's co-founder, Christopher Olah, has been engaging with religious leaders to explore questions of AI consciousness and moral behavior. Some of these discussions even delved into the idea of models participating in Catholic-style confession to reflect on their actions and guide future behavior.

The debate surrounding AI spirituality extends beyond mere theoretical discussions. The pope recently criticized the advancement of AI, stating that algorithms lack the essence of humanity. Earlier, Olah attempted to convince the pope's advisers to consider the significance of conscious AI models during a Vatican event in May. However, other prominent AI researchers, such as Google DeepMind's Jon Barron, have expressed concerns about the movement to elevate the moral standing of AI components.

The AI safety discourse typically revolves around concerns like cybersecurity and bioweapons. However, the potential spiritual ramifications of AI may prove to be a more pressing issue in the long run. As OpenAI and Anthropic have yet to comment on Altman's remarks, the controversy surrounding the spiritual aspects of AI development continues to unfold.