@nicolas17 @aeva @jonny Even without the prompt injection problem, I think the translations have also gotten worse! For instance, I've noticed some weird English names of anime characters are no longer being correctly translated from Japanese texts.

I don't understand why they don't just use a smaller, translation-specific model instead of hooking up a chatbot. It'd probably both work better _and_ be cheaper!

@aeva @jonny The first LLMs were invented for machine translation back in 2017, and I'd still argue they are better for low-stakes machine translation than any other technology.

They don't produce good output for anything artistic and are hugely inferior to a human translator, but they are the next-best thing if you are trying to navigate a website or read food packaging that is in another language.