Boy my experience with small chunks of translation between languages I know well is not the same at all. When prompted properly the translation quality is unbelievable and can absolutely catch nuances, puns, and add footnotes.

That said, I use it with pretty direct prompting, and I strongly prefer the "AI Partners with a Human" model.

I'm sorry but I simply do not believe you that it does handle things like puns and things requiring footnotes, my experience is that LLMs are miserable at this even when directly "instructed" to.

But for what it concerns my previous comment: It doesn't really matter what the "state of the art" AI is because companies simply do not use that. They just pipe it through the easiest & cheapest models, human review (that does not actually get the time to do meaningful review) optional.