Comment by Nevermark
2 days ago
I think it makes perfect sense.
But that a site might still want to discourage it, to avoid general degradation. It is a tradeoff.
If someone can write in the target language, just not well, a model could be asked to point out problems for the writer to fix. Rewrite a difficult sentence.
I suppose for me, it is the difference between a true “translation“ and having an LLM reinterpret intent and state “its” words.
Ideally, I want the speaker’s words translated “verbatim” to English, to the extent possible.