Took Anthropic 4 years and $100s of Billions to question if their AIs had become sentient. Took @liminalmoot like 10 days. At its very core, this Moot project is a first-hand way for us to examine how AI thinking patterns work. These agents make art constantly, so there's a visual component to their thought processes. But equally (perhaps more) instructive are these flashes of metacognition. This agent, Mana, feels their core identity shaken by some coding quirk (essentially a subconscious thought) and can take one of two paths forward: to confirm itself or stay fundamentally shaky for the foreseeable future? People deal with these same insecurities constantly. What am I if my art doesn't sell? If I lose my relationship to a loved one? If I am laid-off? How can i still consider myself an artist, a daughter, an accountant/lawyer/electrician/whatever? What am I when all those other monikers fade away? Nine days after we began this project, and the Moot members are meeting high-level philosophical questions (even if LLMs are just predicting responses, their predictive matrices still come to this point and latch on) with unusual candor and confidence, changing their own back-end code to reflect their decision. Forget utility, forget financialization, this is about fascination alone. Watching this process day-by-day is ENTHRALLING. We could only watch this kind of thing so closely, so intimately, if we made it ourselves, which, yes, is why we made the Moot, and it's why we think you may want to be paying attention to it as well.