That’s really interesting actually. Has anyone tried training a LLM on e.g. whale speak? I wonder if we could end up translating it by interacting with the whale model, that would be very cool.

Poly would get lots of crackers.