Comment Correction: (Score 1) 68
>>If you bothered to read, they are not blaming the victim, they fixed the problem and just point to where it came from.
They fixed one manifestation of the problem. The real source of the problem is that the AI can't seem to evaluate the status of it's training data; even at the most basic level of fiction vs non-fiction, to say nothing of more subtle differences like plausible vs implausible, joking vs serious, fact vs opinion, etc.