When I ask AI something, I often want to later remove these messages from the conversation to keep the context clean. E.g. if I ask him if his implementation makes sense and he convinces that it does, I’d rather remove this from the conversation to keep context short and clean.