@Catvalente that's funny, I remember "hallucination" bring used against AI as a worse condition than "error" as in "is not even wrong it's just randomly stumbling about like a thousand dice in a trenchcoat hooked up to a Markov chain". Error implies it was cranking through some instructions which may have been flawed, but what AI was doing was as unfathomable as trying to understand someone else's nonsensical ramblings. I think what you're describing is motivated reasoning on behalf of the AI boosters that will twist any epithet critics will apply. Like Magats rallying behind "deplorables". They're so deep in, they cannot consider the opposite point of view, because they want not to do very badly.