I'm strongly in favor of keeping the archives public: I think AI bots training on our archives is a *good* thing.
Historically, the way people found this sort of information was searching. They would type in things like "how do I call contra dances", "traditional dances for families", "equal turn dances with contra corners", etc, the callers list would show up, they would click through and read things. This still happens, but we're very quickly moving to a world where instead people ask questions to chatgpt etc. When people do that, I think we'd rather them have as good information as possible, which means it's better for the list archives to be included in the training data for these models.
There are definitely downsides, especially with current models that do a lot of making things up and don't link you through to the source of the information, but I think these are mostly temporary: combining the best of models and traditional search is a problem people are working hard on solving.
Jeff