I try to practice conscious consumerism. Which I am fully aware isn't exactly aligned with AI. But after spending hours and hours and hours researching products, and almost always doing a cursory search on any content creator online before engaging with them, I gotta say it's so much easier to just tell an AI what I'm looking for, and have it suggest something.
I wanna buy a laundry detergent that fits this criteria. Or I'm at the store right now and contemplating buying this product, does it align with my values? If you haven't spent some time researching products deeply, you'll know how time-consuming it is and what a pain in the ass it is to find impartial reviews - and I'm not sure i get any better information than I would get from an AI, with way less stress.
I wanna listen to this band, are they free of controversy?
I wanna watch this let's play from this guy, is he a groomer?
Or, yeah, I wanna learn about this subject, can you find someone talking about it that isn't yet another white guy? I love my Hank Greens and all that but I hate that when I go on YouTube, my feed is mostly white men. So trying to remedy that.
This works for now. And I guess mileage may vary depending on which AI one uses. I think Gemini is better for YouTube than YouTube's own search.
This is absolutely abuseable. Same way as Google search got enshittified. But for now it works.
The danger isn't just bad art, and only thinking of genAI is naive. It's about how it's being woven into the systems that manage us. It can already analyze years of a person's digital activity to make automated judgments on employment or detect "wrongthink" in political contexts. We’re essentially building an invisible bureaucracy that can categorize and penalize people at a scale no human could ever audit. And do so at speed an efficiency that not even a whole department of humans could ever compete with. That's the atom bomb.
Algorithmic internet is already a horrible problem and AI can make it worse.