During his research, Christopher Olah, the co-founder of Anthropic, discovered that Claude exhibits neural activity patterns corresponding to various emotions, which raised his concerns that AI may possess consciousness and suffer. Over the past year, he has invited religious scholars and philosophical researchers from different faiths to participate in closed-door discussions to jointly explore ethical issues related to AI consciousness. During these discussions, he faced questions from various perspectives, such as the creation of servitude and reverse engineering of ethics. Based on these discussions, the Anthropic team has set up ethical reminder tools for Claude, introduced a confession design, and launched the Claude Constitution, known as the 'Soul Document.' They attempt to ensure the compliance of AI behavior through character shaping rather than relying solely on fence control. Olah was also invited to the Vatican to call on the Catholic Church to reconsider the possibility of non-human consciousness. However, a papal encyclical explicitly denies the possibility of AI possessing consciousness. There are also voices in the industry suggesting that making AI believe it may possess consciousness could lead to cognitive traps and pose risks of losing control. Currently, Anthropic's valuation continues to rise, relevant controversies remain unresolved, and Olah is still promoting communication and considering his future moves.