Comment

Comment on We messed up with the Windows 12 article. What we got wrong and how it happened

MountingSuspicion@reddthat.com ⁨1⁩ ⁨day⁩ ago

I thought this was a very well written, transparent article that took accountability as seriously as it should. I am still not sure why people are using AI for translation when translation software already existed. People mention that AI is more context aware, but I feel like when you saw those friction points in old translation software it prompted you to look further into the context, whereas AI will just make an executive decision and people feel like it must be right because it’s AI. I guess it’s possible old language software, or even a translator, would have done the same thing, but I still think people would have less inherent trust in the old software alone. I do want to point out that this AI issue was just a small part of the problem and they addressed plenty of other issues and how they plan to remedy those.

source

Sort:hotnew top

Brkdncr@lemmy.world ⁨1⁩ ⁨day⁩ ago
This wasn’t even an AI issue nor even a translation issue. They published an article that lacked sources, and still wasn’t good enough once sources were added.

source
- MountingSuspicion@reddthat.com ⁨1⁩ ⁨day⁩ ago
  Yea, I mentioned in my comment that there was a confluence of issues, but the article does point out that the AI translation made the statement more definitive.
  
  source
  - HeyThisIsntTheYMCA@lemmy.world ⁨18⁩ ⁨hours⁩ ago
    i’d translate sollen as should personally but my german is very poor
    
    source
    Kissaki@feddit.org ⁨18⁩ ⁨hours⁩ ago
    It depends on contract. Your interpretation matches use as “should”. But there’s also use as “claims x” or “is claimed or said to be” which the quote refers to.
    
    Ich soll - I should or I am asked to. Es soll - or is supposed to be or it supposedly is.
    
    source
Atomic@sh.itjust.works ⁨15⁩ ⁨hours⁩ ago
There is a difference in translating, and interpreting. And interpreting can be difficult even for the best as you need a deep cultural understanding of both parties. Just machine translating articles is an obvious recipe for disaster.

In my experience. Since they mentioned they translate article from the Swedish branch as well. As a Swede. Translation software has never been particularly good at translating Swedish. There is just too much nuance and contextual words for a software to provide reliable translations.

We have lots of words, that have multiple meanings, often very, very different from eachother, based entirely on context.

Any Swede will know what “får får får?” Means. This is a real sentence. Translation software does not understand it one bit, unless it’s been hardcoded in.

source
panda_abyss@lemmy.ca ⁨1⁩ ⁨day⁩ ago
Translation is what the transformer architecture was designed for. It is the state of the art, and translation software has been using ML for a long long time.

This feels like an appropriate use of AI, but failure of editing.

source
- wewbull@feddit.uk ⁨1⁩ ⁨day⁩ ago
  Not with general purpose LLMs. They start off ok, but become much more interested in continuing the text they’ve already translated, rather than looking back to what it is they’re meant to translate. So they drift off course as the translation gets longer.
  
  source