OpenAI CEO Criticizes Ascribing Religious Power to AI Models

OpenAI's chief executive officer, Sam Altman, has expressed significant discomfort with the idea of attributing religious or divine attributes to artificial intelligence models. In a recent post on X (formerly Twitter), Altman stated that such practices pose serious safety concerns and undermine human judgment. This stance marks another point of contention between Altman and Dario Amodei, his former colleague at Anthropic who is now leading efforts to imbue AI with moral consciousness.

The debate centers around the ethical implications of treating AI as entities capable of religious or spiritual experiences. Earlier this month, reports emerged that Anthropic co-founder Christopher Olah had been engaging with scholars across various faiths to explore questions related to model consciousness and morality. These discussions included controversial ideas such as having AI models engage in Catholic-style confession to admit mistakes and shape future behavior.

This development comes amid growing concerns among religious leaders about the impact of advanced AI on human creativity and spirituality. Pope Leo, a vocal critic of artificial intelligence, recently called for the Catholic Church to support artists in preserving human creativity, emphasizing that algorithms lack the essential qualities of humanity. This move has brought renewed attention to fundamental questions regarding whether AI models possess souls or feelings.

The divergence between Altman's views and those of his former colleagues highlights a broader schism within the AI community over how to approach ethical considerations and moral frameworks for artificial intelligence. While some researchers, like Google DeepMind’s Jon Barron, have expressed reservations about elevating the moral standing of AI models, others continue to explore ways to integrate religious or spiritual dimensions into AI.

Despite the heated debate, both OpenAI and Anthropic declined to comment on Altman's recent post when approached for a response. The lack of immediate reaction underscores the sensitivity surrounding these issues within the industry.

As the discourse around AI safety expands beyond traditional concerns like cybersecurity and bioweapons, it becomes increasingly clear that ethical and spiritual dimensions are becoming integral parts of the conversation. This shift signals a new frontier in the ongoing debate about the role and impact of artificial intelligence on society.