The person with a peanut allergy is threatened by the mass agriculture of peanuts, but we inherently recognize that banning all peanuts would be unfair to the vast majority of people who do not possess such an allergy.

I'm sure there are interesting points regarding nutrition and agriculture that could be made about peanuts, but from a lay perspective what I find to be an overwhelmingly salient difference is that it's much easier to test for a peanut allergy, which of course does not cover the small minority of people who develop the allergy later in life (the more common version, pollen-food syndrome, is usually not seriously dangerous). That is not true about psychological issues, where yes you can look at psychosis (via much more intensive testing than peanut allergies) and put some people in the "definitely yes" category and most in the "almost certainly no (for now)" category, but there's a much hazier boundary, even focusing on these two groups in the distribution and not the people who (so far) fall between.

I also think even beyond the diagnosis issue the matter of self-policing on not killing yourself with an allergen is not that similar to self-policing on being especially vulnerable to having the lying flattery robot loosen your grip on reality, for reasons that I would vaguely compare to but not equivocate with self-policing on consuming addictive substances.

Thus "AI psychosis" should be educated about and guarded against as should other psychologically addictive behavior.

A lot of this depends on if we're talking about our present society or some imagined socialist one where there are regulations in place to prevent the product from being malicious. The trendy example is gambling. Obviously I don't think people should be forbade from playing poker with their friends or whatever, but I don't think mere education is enough of a guard to counterbalance the intensely predatory and ever-evolving practices of the industry, so we would need better substantiation on the "guarded against" part to take it seriously.

You can surely see how contemporary AI resembles this, where theoretically there are safeguards that could be put in place to make them much less dangerous, but we can't count on that happening and should have some idea of what our stance on how the present versions should be treated is.

The "classic" opinion on hexbear is most certainly not that all media is apolitical, what the fuck are you talking about.

I was being facetious. I apologize for the confusion. I am very familiar with lengthy discussions on here about which media is reactionary and why.

And to preempt the escalation that I know is coming next: obviously that logic doesn't hold if someone was harmed in the creation of the media itself, which is and should remain illegal

Contrary to your expectation, I at no point was going to bring up abuse materials, I understand that that's a different subject and its reasons do not apply here.

or if the class of media is such that its proliferation would make enforcing the previous rule impossible, which is a real and predictable harm to society by a certain class of generated content that AI psychosis does not represent.

I don't follow the allusion here, honestly, unless it's a different gen-AI thing.