> Taylor has turned his AI chatbot memory off, so that he knows exactly what’s shaping how it responds to him. “The LLM [large language model] doesn’t know who you are, and therefore it won’t bias the results you get,” he says.

I found this amusing. Another example of expecting deterministic output from an LLM.

A lot of these stories always have a single thing in common: a single long-running conversation.