This seems to imply that customers assume by default that the LLM remembers their past chats? I feel like the UI makes it incredibly obvious it’s a clean slate every time? But then again people ask ridiculous meta questions all the time to these chatbots expecting a correct answer.