I don't want to be overly dramatic and overly negative about the AI translations I've been working with. They are bad, yes, but I don't want to overstate their badness because that would obscure the specific points I'm making. Some AI translations, the best AI translations, are not that bad. Some of them are still bad to the point of being unusable, but others are better. They're not good, but they're mostly serviceable, and it's extremely impressive that a machine can come up with something serviceable, something comparable to the work of a very mediocre human translator.
A client who hires a subpar translator who accepts being underpaid, in order to avoid paying professional rates for a professional, is getting subpar work. A client who uses AI to get work cheap and fast is getting worse than subpar work. But AI is getting better, it might soon be at the point where laymen can't tell the difference, and then, using AI instead of paying a human will mostly be a labour rights issue, and that's a far thornier question. (Note that I'm not talking about using AI translators to read something for yourself, or to communicate in your daily life: I'm talking about AI translation for publication, using AI for something you expect other people to pay money for.)
My actual point about AI translation is that even when it's fairly good, when it makes few errors and conveys the message intelligibly, it lacks something. I'm not talking about heart and soul here, nothing to do with some intangible human quality: I'm talking about specificity. AI works with great averages, and so it automatically irons out nuance. If you write something unusual, AI will assume it's an error, instead of an intentionally unusual statement. This is regression to the mean, and based on the texts I'm working with, it's an Anglophone, American mean. If you say something that's true of 1980s Hungary, it might slightly alter the sentence to "make sense" for 1980s US. Some alterations are factual, these are more serious errors but also easier to edit out. But other things are harder to catch, slight shifts in tone and valence, an erasure of the original, specific, non-American perspective, and the end result is a text that doesn't have anything wrong with it, but is markedly simpler and dumber than it should be. And flattening complex, knotty, peripheral perspectives into something closer to a monoculture is, in the long term, intellectually devastating.
I am a translator and I've been talking about this with some other colleagues for a while.
There are two points where something very peculiar happens when you translate into Spanish.
One is gender and the other is dialect.
I've been following the CAT (computer assisted translation) tools for years, and as the world has evolved, so have the corpus the tools feed from. So when back then we were taught how to transtate texts to keep gender undisclosed, especially for videogames, for reasons we can't unpack here. CAT tools tended to translate very gendered texts, and very biasedly (nurses were enfermerAs and lawyers were abogadOs).
That is no longer the case. Gender is being slowly erased, simply because when it chooses a gender there is a 50 chance of being wrong that doesn't apply when using genderless terms. So you can have a gendered text in English and get an ambiguous translation in Spanish.
I'm not assigning this any moral value.
And then dialects. Spanish has a fuckton of dialects and nearly all of them grate their neighbours in one small way or other. Old CAT tools were curated by linguists so their corpus was usually curated by country (and other parameters). Right now the AI works on statistical use of words, ignoring variants. This works for USA English because it is the hegemonic variant, but Spanish has similar numbers of speakers in a lot of countries, so automatic tranlators are both excellent and rubbish at this.
You see, massive corporations have always been extremely annoyed at Spanish speaking countries because what's good here is bad there, so they can't unify their marketing strategy.
For years they've paid translators to write a neutral variant of Spanish that will be tame and agreeable and with no accent, and sellable in every spanish speaking country (hence why so many products stay marketed in English, you can't please them all)
So here come the automatic translators; they objectively use the most used word, the one most countries find unobjectionable (or the one most present in the corpus). And what do you get?
A text that pisses off everyone. Voseo, tuteo and usedeo all mixed up, verbal modes all over the place, use the latin name of the animal because nobody agrees on this thing's name...
What a waste, having a human trained to detect nuance, create an artificial language specifically to be a dry and tasteless point were we can all communicate... And thow it all away because you'd rather have quantity by scrapping every text online over quality by having linguists sort through the corpus used for the translations.
And that's my two cents
