I believe that the *reader* should have the power to choose the contexts that they consider AI-generated content to be acceptable. I personally don't want to read AI Substack posts, but if somebody else has no issue with it then I won't stop them from reading or posting.

Pointing out that an essay or post is AI-generated is not policing. It is simply readers expressing their preference. IMO if your first instinct is to tell people to change their preference, I would ask for a bit more empathy.

This is the reason I take issue with your argument that a tool like Pangram shouldn't exist, or shouldn't be broadly accessible. If you look at the trendlines for AI content, 80-90% of the internet is going to be AI-generated in the next couple of years. We need AI content filters for the same reason we need spam filters, and I think it's irresponsible to pretend that AI content isn't a problem because some people put a lot of work into their Codex prompts.