This is sort of missing the point-- people who dont deal with linguistics dont understand that there are multiple types of translation. There's word for word (which is what you're talking about) and sense for sense. If you let an LLM do all of your translation you're letting it interpret huge amounts of intent and context it doesnt (and probably cant) access. The ways in which this impacts the translation will forever be unknown to you and in the worst case lost forever.

So i guess in the end it just matters how important the work is.

Actually I was talking about tonally as well.

A raw "word for word" translation (which I also tried) made the story somewhat hard to follow and very dry, but just asking it to keep the same kind of jovial swashbuckling tone of the original made something pretty similar to Ellsworth's translation.

Again, before someone decides to "correct" me on this, I am aware that it's very likely that the Ellsworth translations are part of the training set so it's not directly a fair comparison.

If you let an LLM do all of your translation you're letting it interpret huge amounts of intent and context it doesnt (and probably cant) access.

What’s the intent and context that a human translator of a text is typically privy to that an LLM is not?

> If you let an LLM do all of your translation you're letting it interpret huge amounts of intent and context it doesnt (and probably cant) access.

Assuming lots of material local to the context one is wanting to translate is included, why couldn't it potentially access that additional context?