Oh come on, how difficult can it be for any language with a considerable corpus available? I mean maybe not for Linear B, but French, Spanish, German, Japanese, and Chinese should easily be next. In fact, it might be even easier due to their stricter grammars. Only considerable barrier is marking up the training data... how hold do you think that damn will hold?

I was playing with ChatGPT in Norwegian a few weeks ago and the results looked good. Of course, that's not a typical language, but there is at least a lot of source material. I wanted to try it now but they are having issues :)

ChatGPT is quite capable of generating content in Spanish. Not sure how many good lectures are available on YouTube though.