News for hackers. Stuff that matters.
Re:Seamless: Meta's New Speech Models by diob on Friday December 01, 2023 @02:47PM (#38491305)
The problem is you need a full sentence, plus surrounding sentences to properly translate a lot of things (aka context matters).

So no matter what, conversations in your native speech would have to be delayed before translation.

Re:Seamless: Meta's New Speech Models by ItsMattyG on Friday December 01, 2023 @04:18PM (#38492563)
My understanding is that they trained a separate model to specifically estimate when they have enough context to begin translating, as a skilled translator would.
Re:Seamless: Meta's New Speech Models by scherlock on Friday December 01, 2023 @06:52PM (#38494227)
My mom used to do English/French translation. Her favorite example was the word "file". That word has multiple translation in French depending on the context, and that context may simply be implied by who is speaking. You may not be able to figure it based on the conversation alone.
Re:Seamless: Meta's New Speech Models by ChuckMcM on Friday December 01, 2023 @03:05PM (#38491556)
I think I could adapt to that. But it would be an interesting experiment.
Re:Seamless: Meta's New Speech Models by DigiDigiorno on Friday December 01, 2023 @05:08PM (#38493174)
Even the native original version needs the proper context. Sometimes you need the entire sentence to figure out what the sentence was really about.

I'm reminded of Mark Twain complaining about verbs arriving at the very end of sentencess in German (among a myriad of other complaints)

"The Awful German Language* -Mark Twain https://faculty.georgetown.edu/jod/texts/twain.german.html

Re:Seamless: Meta's New Speech Models by scotty79 on Friday December 01, 2023 @05:57PM (#38493738)
Sometimes you even need a second sentence of even a few to understand what the first sentence was about.
Re:Seamless: Meta's New Speech Models by sexy_seedbox on Friday December 01, 2023 @05:55PM (#38493713)
So then we need something like neuralink to get the whole thought from one's brain first, then the sentences are processed properly for the context, then translated before the speech is delivered.
Re:Seamless: Meta's New Speech Models by kabouseng on Friday December 01, 2023 @11:11PM (#38495804)
Most thoughts are in a language. There is no one underlying universal machine language for the brain.
Re:Seamless: Meta's New Speech Models by plastic3169 on Saturday December 02, 2023 @01:34AM (#38496469)
Are most thoughts in language? This doesn’t reflect my experience. Language floats on top, but there is layers under there. You can also feel it when you end up thinking in another language. It does not go through the first one but is a thing of it’s own.

Pretty sure there is nothing universal there though as you say.

Re:Seamless: Meta's New Speech Models by sexy_seedbox on Monday December 04, 2023 @06:22AM (#38515964)
What about blind deaf people, do they think in a language?