I wonder how far you can push this? Like there was a whole subreddit (can't remember the name) that was dedicated to news that seems positive on the surface but is actually fucked up when you take 2 seconds to think. Could you just get chatgpt to come up with a way to positively frame the worst shit specifically to bait adds next to it?