Despite public assurances that AI wouldn’t touch creative content, fans spot ChatGPT in Crunchyroll’s latest anime release. Oops, ChatGPT Said That Out Loud Crunchyroll has accidentally confirmed …
I don’t see what the problem is with using AI for translations. if the translations are good enough and cheap enough, they should be used. If they’re not good enough, another business can offer better translations as a differentiator.
I don’t see what the problem is with using AI for translations. if the translations are good enough and cheap enough, they should be used.
Because machine translations for any large chunk of text are consistently awful: they don’t get references right, they often miss the point of the original utterance, they ignore cultural context, so goes on. It’s like wiping your arse with an old sock - sure, you could do it in a pinch, but you definitively don’t want to do it regularly!
Verbose example, using Portuguese to English
I’ll give you an example, using PT→EN because I don’t speak JP. Let’s say Alice tells Bob “ma’ tu é uma nota de três pila, né?” (literally: “bu[t] you’re a three bucks bill, isn’t it?”) . A human translator will immediately notice a few things:
It’s an informal and regional register. If Alice typically uses this register, it’s part of her characterisation; else, it register shift is noteworthy. Either way, it’s meaningful.
There’s an idiom there; “nota de três pila” (three bucks bill). It conveys some[thing/one] is blatantly false.
There’s a rhetorical question, worded like an accusation. The scene dictates how it should be interpreted.
So depending on the context, the translator might translate this as “ain’t ya full of shit…”, or perhaps “wow, you’re as fake as Monopoly money, arentcha?”. Now, check how chatbots do it:
GPT-4o mini: “But you’re a three-buck note, right?”
Llama 4 Scout: “But you are a three-dollar bill, aren’t you?”; or “You’re a three-dollar bill, right?” (it offers both alternatives)
Both miss the mark. If you talk about three dollar bills in English, lots of people associate it with gay people, creating an association that simply does not exist in the original. The extremely informal and regional register is gone, as well as the accusatory tone.
With Claude shitting this pile of idiocy, that I had to screenshot because otherwise people wouldn’t believe me:
[This is wrong on so many levels I don’t… I don’t even…]
This is what you get for AI translations between two IE languages in the same Sprachbund, that’ll often do things in a similar way. It gets way worse for Japanese → English - because they’re languages from different families, different cultures, that didn’t historically interact that much. It’s like the dumb shit above, multiplied by ten.
If they’re not good enough, another business can offer better translations as a differentiator.
That “business” is called watching pirated anime with fan subs, made by people who genuinely enjoy anime and want others to enjoy it too.
Even if it could it wouldn’t matter, what makes art so valuable is the amount of work put into it. It’s the highest effort form of communication you can imagine, it can express so much. An LLM chat bot could never do that.
A good human translator is always the best solution.
But if the choice would be between crappy google translate or a LLM I would take the LLM translation.
There’s no excuse for a big studio, they should hire translators. But for indie creators without a budget it can be the best way to get their creation to more people.
I don’t know why it’s so difficult for these Ayn Rand readers to understand that almost no market is perfectly free, efficient, contestable or optimal. When it comes to licensing IP in particular, the overhead of trying to get permission for a show that’s already licensed by a competitor is an insurmountable barrier to entry for anyone who isn’t already a huge media conglomerate.
The only realistic alternative to shitty subtitles is piracy.
If you need translation for just getting facts and information for say math equation and its annotation translated, there’s little margin in variety, what you need is database - that’s mostly fine.
Pieces need translation are usually not like that. They have cultural context, pun, wordplay in rhymes, structural parallel, underlying tone, a lot of things only work in the language originally written. Translation is always a (nearly impossible) challenge for the translator to reconstruct all of them in target language.
I did game translation for a while. Translation is a field where AI hit first and honestly I’ve seen people lowering standards. The criteria of “good enough”, “passable” is not the same compared to pre-AI days, and will keep changing. I’m almost sure this trend is happening in every industry the same way, and “just translation” is a slippery slope.
I don’t see what the problem is with using AI for translations. if the translations are good enough and cheap enough, they should be used. If they’re not good enough, another business can offer better translations as a differentiator.
Because machine translations for any large chunk of text are consistently awful: they don’t get references right, they often miss the point of the original utterance, they ignore cultural context, so goes on. It’s like wiping your arse with an old sock - sure, you could do it in a pinch, but you definitively don’t want to do it regularly!
Verbose example, using Portuguese to English
I’ll give you an example, using PT→EN because I don’t speak JP. Let’s say Alice tells Bob “ma’ tu é uma nota de três pila, né?” (literally: “bu[t] you’re a three bucks bill, isn’t it?”) . A human translator will immediately notice a few things:
So depending on the context, the translator might translate this as “ain’t ya full of shit…”, or perhaps “wow, you’re as fake as Monopoly money, arentcha?”. Now, check how chatbots do it:
Both miss the mark. If you talk about three dollar bills in English, lots of people associate it with gay people, creating an association that simply does not exist in the original. The extremely informal and regional register is gone, as well as the accusatory tone.
With Claude shitting this pile of idiocy, that I had to screenshot because otherwise people wouldn’t believe me:
[This is wrong on so many levels I don’t… I don’t even…]
This is what you get for AI translations between two IE languages in the same Sprachbund, that’ll often do things in a similar way. It gets way worse for Japanese → English - because they’re languages from different families, different cultures, that didn’t historically interact that much. It’s like the dumb shit above, multiplied by ten.
That “business” is called watching pirated anime with fan subs, made by people who genuinely enjoy anime and want others to enjoy it too.
That’s a really awesome explanation.
Also, r/suddenlycaralho.
This is such a great response. I learned something from you today, thank you.
Translating art is still art. It requires cultural knowledge and contextual understanding to be done well, and no soulless bot can do that.
Even if it could it wouldn’t matter, what makes art so valuable is the amount of work put into it. It’s the highest effort form of communication you can imagine, it can express so much. An LLM chat bot could never do that.
This. There’s a huge different between AI’s shallow transliteration and an actual translation that respects the culture and context.
A good human translator is always the best solution.
But if the choice would be between crappy google translate or a LLM I would take the LLM translation.
There’s no excuse for a big studio, they should hire translators. But for indie creators without a budget it can be the best way to get their creation to more people.
I don’t know why it’s so difficult for these Ayn Rand readers to understand that almost no market is perfectly free, efficient, contestable or optimal. When it comes to licensing IP in particular, the overhead of trying to get permission for a show that’s already licensed by a competitor is an insurmountable barrier to entry for anyone who isn’t already a huge media conglomerate.
The only realistic alternative to shitty subtitles is piracy.
The issue is that they’re almost certainly going to be cheap, and therefore companies are going to use them even if they’re complete dogshit.
But I think the bigger issue is saying “no ai” and then… using AI.
If you need translation for just getting facts and information for say math equation and its annotation translated, there’s little margin in variety, what you need is database - that’s mostly fine.
Pieces need translation are usually not like that. They have cultural context, pun, wordplay in rhymes, structural parallel, underlying tone, a lot of things only work in the language originally written. Translation is always a (nearly impossible) challenge for the translator to reconstruct all of them in target language.
I did game translation for a while. Translation is a field where AI hit first and honestly I’ve seen people lowering standards. The criteria of “good enough”, “passable” is not the same compared to pre-AI days, and will keep changing. I’m almost sure this trend is happening in every industry the same way, and “just translation” is a slippery slope.
They’re fine if no official translation exists and they’re free. If you’re paying for a service, you are owed much better.