this post was submitted on 14 Sep 2026
397 points (96.9% liked)

Fuck AI

8217 readers
2240 users here now

"We did it, Patrick! We made a technological breakthrough!"

A place for all those who loathe AI to discuss things, post articles, and ridicule the AI hype. Proud supporter of working people. And proud booer of SXSW 2024.

AI, in this case, refers to LLMs, GPT technology, and anything listed as "AI" meant to increase market valuations.

founded 2 years ago
MODERATORS
 

The "interview"

Edit: fixed the link

you are viewing a single comment's thread
view the rest of the comments
[–] athairmor@lemmy.world 40 points 12 hours ago (4 children)

Theoretically, could you feed to some genAI video of this Tilly Norwood and create an exact—or close enough—replica?

Then, theoretically, make videos of her saying and doing extremely offensive things?

[–] brucethemoose@lemmy.world 7 points 7 hours ago* (last edited 6 hours ago) (1 children)

FYI this was just a quick test with Minimax H3, locally generated on a RTX 3090/Ryzen 7800 desktop.

And to answer your question... yep. It's not a prompt box. H3 will take any combination of audio, video, and images as reference, as well as controlnets for poses, masking, character swapping...

[–] mrmisses@lemmy.world 7 points 7 hours ago (1 children)

All her how she feels about the genocide in Palestine

[–] brucethemoose@lemmy.world 5 points 6 hours ago* (last edited 6 hours ago) (1 children)
[–] mrmisses@lemmy.world 1 points 6 hours ago (2 children)

That's not a response that's just repeating the question

[–] shawn1122@sh.itjust.works 7 points 6 hours ago

Ready for politics

[–] Aqivex@fedinsfw.app 16 points 11 hours ago (2 children)

No theory needed. You absolutely can, almost trivially so.

Only real challenge(s) would be either owning the hardware capacity or renting from a neutral host to run the fake AI personality at/or near equivalent levels if you're gonna unleash Titty SouthRisesWood in an autonomous form.

Maybe a lawyer with a sense of humor on retainer to respond to the frivolous demand letters if they eventually find out you're behind it.

And that's just if you want to fight AI with AI.

Anyone could copy the model the good ol' fashioned way in Blender, turn on a voice changer or use non-AI TTS and get to uploading their protected parodies.

[–] FrChazzz@lemmus.org 6 points 7 hours ago

And aren't there recent legal rulings in the US that say that LLM-generated content is not able to copyrighted?

[–] brucethemoose@lemmy.world 2 points 7 hours ago

Not anymore. Anyone with a beefy gaming PC can do reference-to-video with enough fidelity for memes.

...Especially when the "AI Actress" is already so bad.

[–] Tartas1995@discuss.tchncs.de 3 points 10 hours ago (1 children)

This made me wonder about the ethics of "deepfake" porn of genAI pixels that look like a human.

[–] FoxAlive@lemmy.zip 3 points 6 hours ago* (last edited 6 hours ago)

Ai companies are not held responsible even though many of them directly advertise this as a feature like grok.

Instead they say "its already illegal" and try to go after the individuals to generated it. And even then thats not even true, the courts said ai generated porn is legal even if its csam, and Donald trump made it illegal for the states to regulate ai in any form. Which includes generating csam, deepfakes, etc. Because to the epstine class that's just innovation.

Their argument is that they can't be held responsible because someone used their tool inappropriately when the tool was specifically designed and has features to do this.

My argument is this is a level of willful incompetence to the point of allowing it. Its for example like if a father would go to prison for leaving his gun loaded on the kitchen table, and his son goes out and commits crimes with said gun. Ai companies like grok are doing the equivalent of unloading truck loads of automatic guns at the capitol building during January 6th.

Section 230 only applies to people who you aren't virtually guiding to do these illegal/harmful things. You can't just say we aren't responsible for teens using your platform(like meta is) and then do everything in your power to get kids addicted on your platform short of making it mandatory for every child to sign up for their services.

These people know what they are doing and pretending like they don't know what happened or know how to stop it.

Edit* I'm stupid I didnt even read/comprehend your post and essentially went on my own rant. If it means anything I brought up a few points that still apply.

[–] you_are_dust@lemmy.world 5 points 11 hours ago

Probably? I'm curious of the result of feeding AI more AI to try to replicate a particular AI creation.