Hacker News Reader

HNHacker News

Top New Best Ask Show Jobs

Comment by qprofyeh | Hacker News Reader

Full thread

qprofyeh·

Can we talk to any animals using LLM translation?

View on HN

No? I mean it seems like a stretch. Where would the LLM learn the initial English-to-Moo mapping for cows?

Maybe if we would have a mapping we could use that to train an LLM.

In the meantime you can try asking ChatGPT if a cow has buddha nature and ask for a one word reply. It have an idea what it might say.

You don't need Parallel corpora for all language pairs in a "predict the next token" LLM. What I'm saying is that if an LLM is trained on English, French and Spanish and there is Eng to French data, you don't need Eng to Spa data to get Eng to Spa translations.

How would an LLM figure out what words to translate animal sounds to? Where does it learn that information? We don't know what animals are communicating if they do have a language of sorts. There's no mapping.

Potentially the same way it knows how to translate concepts with no mappings in that language pair in the dataset. Like i said, not every language in an LLM's corpus has something in another language to map to.

Spanish and French are both romance languages and will have massive token overlap. Not likely to be so lucky with whale songs.

It's not about romance or no romance. Same with Korean/Mandarin or any distant human language.

>Not likely to be so lucky with whale songs

Maybe. Maybe not

“Yes”

Very nearly. https://www.youtube.com/watch?v=3tUXbbbMhvk

If you can find a way to tokenize whalesong then you could probably build a next-token predictor but we're still missing the Rosetta stone necessary to help us map their language to ours in vector space.

What does that mean?

Reply on news.ycombinator.com