No Language Left Behind

#1

Post by pesenti » Wed, Jul 06, 2022, 7:52 PM UTC

Re: No Language Left Behind

#2

Post by btheshoe » Wed, Jul 06, 2022, 8:07 PM UTC

I'm not entirely sure why low resource languages are seen as such a high priority for AI research. It seems that by definition there's little payoff to solving translation for them.

Re: No Language Left Behind

#3

Post by dunefox » Wed, Jul 06, 2022, 8:13 PM UTC

btheshoe wrote: post #2

I'm not entirely sure why low resource languages are seen as such a high priority for AI research. It seems that by definition there's little payoff to solving translation for them.

Small data, big meaning is much more important than big data, little meaning. Much closer to real intelligence.

Re: No Language Left Behind

#4

Post by tehsauce » Wed, Jul 06, 2022, 8:21 PM UTC

btheshoe wrote: post #2

I'm not entirely sure why low resource languages are seen as such a high priority for AI research. It seems that by definition there's little payoff to solving translation for them.

I think the reason low resource languages are prioritized is to compensate for the fact that AI research normally has a tendency to marginalize these languages.

Re: No Language Left Behind

#5

Post by munificent » Wed, Jul 06, 2022, 8:27 PM UTC

btheshoe wrote: post #2

I'm not entirely sure why low resource languages are seen as such a high priority for AI research. It seems that by definition there's little payoff to solving translation for them.

Cynical answer: It's good PR.

Re: No Language Left Behind

#6

Post by btheshoe » Wed, Jul 06, 2022, 8:33 PM UTC

tehsauce wrote: post #4

btheshoe wrote: post #2

I'm not entirely sure why low resource languages are seen as such a high priority for AI research. It seems that by definition there's little payoff to solving translation for them.

I think the reason low resource languages are prioritized is to compensate for the fact that AI research normally has a tendency to marginalize these languages.

yes, but what principles justify the importance placed on low resource languages?

Re: No Language Left Behind

#7

Post by pesenti » Wed, Jul 06, 2022, 8:35 PM UTC

Blog post: https://ai.facebook.com/blog/nllb-200-high-quality-machine-t...

Paper: https://research.facebook.com/publications/no-language-left-...

Github: https://github.com/facebookresearch/fairseq/tree/nllb/

Re: No Language Left Behind

#8

Post by kwhitefoot » Wed, Jul 06, 2022, 8:39 PM UTC

What is a "low resource language"?

Re: No Language Left Behind

#9

Post by Jabbles » Wed, Jul 06, 2022, 8:41 PM UTC

btheshoe wrote: post #2

I'm not entirely sure why low resource languages are seen as such a high priority for AI research. It seems that by definition there's little payoff to solving translation for them.

Surely the fact that they did all the high-resource languages first and are only now getting round to the less-popular ones demonstrates that that is not, in fact, the case?

Re: No Language Left Behind

#10

Post by pesenti » Wed, Jul 06, 2022, 8:45 PM UTC

kwhitefoot wrote: post #8

What is a "low resource language"?

https://datascience.stackexchange.com/questions/62868/high-l...