As AI is here to stay and will probably only get more powerful, I hope AI researchers pay serious attention to the issues you raise. Being in this field myself, I wonder if it is possible, e.g, for an AI language generator to know what parts of its output can be attributed to common knowledge on one extreme, or to a particular person on another extreme. Presently language AI's are not introspective at all. Even with such ability, however, there'd still need to be some kind of enforcement. Hopefully an approach that can't be exploited itself by bad guys as well. We need more people like you David constantly splashing cold water in our faces!
As AI is here to stay and will probably only get more powerful, I hope AI researchers pay serious attention to the issues you raise. Being in this field myself, I wonder if it is possible, e.g, for an AI language generator to know what parts of its output can be attributed to common knowledge on one extreme, or to a particular person on another extreme. Presently language AI's are not introspective at all. Even with such ability, however, there'd still need to be some kind of enforcement. Hopefully an approach that can't be exploited itself by bad guys as well. We need more people like you David constantly splashing cold water in our faces!
Wow, what an awful feeling that must have been!