An interesting, relatively short discussion about using AI chatbots as tools for revitalizing endangered languages. The upshot is that, if it's going to be done at all, it needs to be done very, very carefully and only by or in close collaboration with the community that actually uses a given language. But there are an awful lot of details to work through, and this discussion is really tip-of-the-iceberg level. (Personally, I'm deeply skeptical of the value of "AI" tools for projects like these, and I'm not really convinced that the payoff is likely to be worth the effort. That's without even mentioning the ecological cost of "AI". Still, the discussion is interesting even if just in terms of the ethics involved and the particular complications of working with endangered languages.)




