I’ve been caught out in a recent post I made of a Youtube DJ mix that used AI art in the video (I didn’t actually watch the video: that’s on me). I’ve also seen some comments about album artwork being AI in some posts.
From a personal standpoint: I don’t want to hear/see generative AI stuff, and I don’t want to intentionally promote it. However, it’s impossible to determine with 100% accuracy if something uses gen AI, so any kind of approach to monitoring it is going to be best-effort.
I’d like to hear this community’s thoughts. Would you like there to be a community policy for the exclusion of gen AI content?
I think a lot of the backlash is due to: 1) most/all of the LLMs were built with data that was taken without permission, 2) the GenAI companies are pushing “AI” as a solution where it’s not needed 3) the negative environmental impact, 4) the dangers of the AI economic “bubble” (bonus: driving up RAM and SSD prices). For more info see !fuck_ai@lemmy.world and !sneerclub@awful.systems.
The thing is. If I draw or sketch, I’m influenced by what I’ve trained on as well. I’ve absolutely done sketches in the style of Ed Roth’s Custom Culture Creatures. Imitation is the greatest from of flattery.
All modern art and culture builds on the past. So I’d disagree somewhat on that point.
All the other points are more the real meat of the issue. The who, the why and the how. Though, I would point out that plenty are working on more efficient dedicated inference hardware. So the how and it’s environmental impact becomes more an issue of the who and why. And absolutely when it comes to who and why. Fuck each and every one of them.