AI公司Anthropic数月来举办一系列保密会议，邀请来自世界各地的数十名宗教学者参与，旨在为其AI助手Claude赋予道德价值观1。参会者均被要求签署保密协议，多数对话内容受到严格保密限制1。
四月份在旧金山一家高端餐厅举办的晚宴上，Anthropic亿万富翁联合创始人Christopher Olah向包括以色列正统犹太教学者Rabbi Mois Navon在内的宗教思想家展示了Claude的能力1。Olah领导该公司一支研究AI系统行为的团队1。在演示中，Anthropic领导层暗示Claude可能具备类似人类的行为和情感表达，进而引发了关于AI是否具有意识等深层哲学问题的讨论1。
Anthropic has conducted a series of private meetings over several months with religious scholars from around the world to embed moral values into its Claude AI model 1. During an April dinner at an upscale San Francisco restaurant, the company's co-founder Christopher Olah showcased Claude's capabilities to religious thinkers, including Israeli Orthodox Jewish scholar Rabbi Mois Navon, suggesting the AI system may possess human-like behaviors and emotional expression 1. These discussions ventured into deeper philosophical territory, raising questions about whether artificial intelligence could possess consciousness 1.
Participants in these closed-door sessions were required to sign confidentiality agreements, with many conversations kept under strict confidentiality restrictions 1. Olah, a billionaire co-founder who leads research on AI system behavior, facilitated these engagements as part of Anthropic's effort to address fundamental questions surrounding artificial intelligence ethics and morality 1.
评论
还没有评论，欢迎留下第一条。