> A new language now has to clear the ever growing hurdle of not being in the LLM training data.

I found this to be far less of a prblen than I thought it would be.

Do you have practical experience?

Not with a new language but novel approaches within a framework or paradigm feel outside the LLM wheelhouse.

Even if the LLM is adept at novel languages the developer still has to learn it and learning new languages now when most programming is done via a prompt-review loop feels like it has lower ROI.