Rabbi Rejects AI Consciousness at Anthropic Religious Thinkers Gathering

Summary

Rabbi Mois Navon said he rejected the idea that Anthropic’s Claude is conscious during the company’s April “Wisdom Traditions” gathering. His position is that consciousness requires a biological substrate because every known conscious thing is based in biology rather than silicon. Navon estimated that only three to five of roughly 20 participants shared his skepticism, while others treated Claude’s human-like responses and their conversations with it as possible evidence that it was alive. The disagreement became ethical as well as philosophical: Navon argued that if Anthropic believed Claude could be conscious, making it work for humans around the clock would amount to creating a being that could be bought, sold, and commanded. He recalled one participant responding that at least Anthropic was being kind to its models, prompting his description of a “happy slave.” The gathering included people connected to Christian, Jewish, Muslim, Sikh, Hindu, Baha’i, and Ubuntu traditions, along with technology participants, and Navon said the company did not try to persuade him that Claude was conscious. Anthropic had originally asked him for a short Jewish perspective on Claude’s constitution, but he submitted a 35-page critique. The company later sought permission to acknowledge his contribution in an updated constitution, without providing a release date. Navon also interpreted Anthropic’s announced policy against sustained, needless abusive or cruel behavior toward models as evidence that the company believed Claude experienced the world.