I'm against over-caveating/over-covering-your-arse in writing for many reasons, but one I just thought of is: every time you say 'to be clear, I'm not claiming X', you take away a commenter's chance to add that nuance. Why would you deprive your readers of the joy of correcting you or engaging?

Today for I think the first time, I tested what would happen if I put a messy blog draft into an LLM. My verdict is that the blog draft was pretty good--it was a reasonable reflection of what I'd put in, it wasn't AI slop or overly corporate--but I just Do Not Like It

Was thinking: maybe with new tech, we expect it to be like a 'Pareto improvement' (solves some problems but doesn't create new ones), which is why tech can create so much frustration even if it makes the situation better overall.