I want an AI that will tell me when I have asked a stupid question. They all fail at this with no signs of improvement.
I think it's more believable that the holodeck is ran from separate models that just run inference on the same compute and the ship AI just spins up the containers, it's not literally the ship AI doing that acting itself. Otherwise I have... questions on why starfleet added that functionality beforehand lol.
What we have built in terms of LLMs barely qualifies as a VI, and not a particularly reliable one. I think we should begin treating and designing them as such, emphasizing responding to queries and carrying out commands accurately over friendliness. (The "friendly" in "user-friendly" has done too much anthropomorphization work. User-friendly non-AI software makes user choices, and the results of such choices, clear and responds unambiguously to commands.)
I don't actually think being told that I have asked a stupid question is valuable. One of the primary values, I think, of LLM is that it is endlessly patient with stupid questions. I would prefer if it did not comment on the value of my questions at all, good or bad.