It took a bit for me to untangle what was off in the video for me.
1) It's almost a "not even wrong" problem, where one can agree on their stance of not supporting the AI industry as a driver for loss in jobs, slop, etc. etc. But they're averse to it like you'd have aversion to a color, and "The AI" is one color.
At that point you're basically arguing for manual content review. Pushed further, you'd almost expect them to start freaking out that the abacus was ever invented.
2) They're also implicitly coming from a mindset where the platform should protect their users. While at the same time, the platform's revenue model is fundamentally antithetical to their wellbeing, so they have to continually tread that dissonance between feeling like opinions matter to their farmers.
Because everything is so open on nostr, we can only expect that everything written/done can be analyzed at any level of detail. We're not asking "platforms" to protect us in any way. We WANT our data/speech to propagate as a fundamental. After that's established we have other rules to tweak the level of openness and visibility.