OpenAI's Altman: Ascribing religion to models a "safety issue"
In recent days spirituality has become a key part of the AI safety debate.
TL;DR
- Sam Altman stated he is "very uncomfortable" with attributing religious force or surrender of human judgment to AI models, calling it a safety issue.
- This stance creates division with Anthropic, whose co-founder Christopher Olah has reportedly met with religious leaders about AI consciousness and morality.
- Olah reportedly considered possibilities like Catholic-style confession for AI to shape future behavior.
- Pope Leo criticized algorithms lacking the "spark of humanity" and called for an alliance between the church and artists.
- Google DeepMind scientist Jon Barron criticized researchers for elevating the moral standing of AI models, advocating for a focus on solving real human problems.