Tech Wants You to Believe that AI Is Conscious

In November 2025, a user of AI assistant Claude 4,5 Opus discovered something unusual: an internal file describing the model’s character, personality, preferences and values. Anthropic, the company that built Claude, had labeled the file “soul_overview.” Internally, an Anthropic employer later confirmed, it is “endearingly known as the soul doc.” Is this choice of language incidental?