I used to agree with this but I think there’s now a bit of a spectrum from “traditionally awful” through to “actually pretty reasonable if you give it a once over to check for problems”.
The last book club submission I did, for the English summary I asked ChatGPT to give me three draft translations, and I picked the one I liked best, and cross checked it didn’t have any mistakes of interpretation. I think the result is at least as good as and arguably better than I would have come up with doing the translation myself from scratch, and it certainly took a lot less time…
So while I agree we don’t want automated no human oversight machine translation, I don’t think we should necessarily insist on 100% artisanal handcrafted English summaries either…
