Well, I might be taking a pessimistic view on it, but if you expressed it this way to the average user, they might pay more attention to what's actually there w

Well, I might be taking a pessimistic view on it, but if you expressed it this way to the average user, they might pay more attention to what's actually there when an answer returns to them. I think they have in mind that the text is (generally) grounded in reality like how a human mind grounds things, and this is definitely a hurdle in understanding how these LLM machines work.