Unfortunately, that’s how it got this way. Too many Eds.

There’s no such thing as perfect data. Especially if there’s even the slightest bit of subjectivity involved.

Even less existent is complete data.

… I want clear evidence that the LLM … will never hallucinate or make something up.

Nothing else you listed matters: That one reduces to “Ban all Generative AI”. Actually worse than that, it’s “Ban all machine learning models”.

So far the best counter I’ve seen. It’s almost what could be considered ‘realistic’ given the circumstances.