[____tom____]

What I really want to know is how well these could work for non-human languages. No, not aliens, but chimpanzees, dolphins, bonobos. We have hundreds or thousands of hours of recordings.

What would it take to start working on them?

[benob]

Not tested on that particular model, but the idea has been flying around for some time: https://arxiv.org/abs/2509.04166v1

[nshm]

You can check whale sound recognition project https://arxiv.org/abs/2104.08614

[netdevphoenix]

I think linguistics, don't deem animals to have languages as you require human level intelligence to use and understand some of the features in human languages like communicating about things that are away from your current timespace location. Animals have communication systems

[____tom____]

I'm not asserting that bonobos, for example, have as complex a language as humans, just that it would be interesting to understand what language that they do have.

"You haven't experienced Shakespeare until you've read him in the original Bonobo". :-)

[netdevphoenix]

It would be indeed. But it would not be a language. What animals have is called "communication system". A language can be seen as a type of communication system. Languages are complex by definition and require certain cognitive capabilities: ability to create new sentences based on a set of rules, spatial-temporal displacement, etc.

[akreal]

There is a dolphin language model project from Google and Georgia Tech: https://blog.google/technology/ai/dolphingemma/

[____tom____]

That's exactly the kind of thing I was hoping people were working on!