I agree, but I'm not sure that's sufficient. ChatGPT is pretty good and say that the next version actually works well enough to handle most non-close social interactions. Then I run into you in person and realize you have _no_ idea of what “we” have been talking about — I'm not sure there is a good way to repair that.

Ideally, we'd require labels but there's no easy way to enforce that without something like random auditing of AI implementers.