msgraves

joined 1 year ago
[–] msgraves@lemmy.dbzer0.com 18 points 7 months ago (2 children)

One of the worst parts of this boom in LLM models is the fact that they can "invade" online spaces and control a narrative. For an example, just go on twitter and scroll to the comments on any tagesschau (german news site) post- it's all rightwing bots and crap. LLMs do have uses, but the big problem is that a bad actor can basically control any narrative with the amount of sheer crap they can output. And OpenAI does nothing- even though they are the biggest provider. It earns them money, after all.

I also can't really think of a good way to combat this. If you would verify people using an ID, you basically nuke all semblance of online anonymity. If you have some sort of captcha, it will probably be easily bypassed- it doesn't even need to be tricked. Just pay some human in a country with extremely cheap labour that will solve it for your bot. It really sucks.

[–] msgraves@lemmy.dbzer0.com 0 points 7 months ago (3 children)

same energy (and impact) as "X formerly known as Twitter"

[–] msgraves@lemmy.dbzer0.com 7 points 7 months ago

Exactly, this isn't about any sort of AI, this is the old playbook of trying to digitally track images, just with the current label slapped on. Regardless of your opinion on AI, this is a terrible way to solve this.

[–] msgraves@lemmy.dbzer0.com -2 points 7 months ago* (last edited 7 months ago)

I never said we dONT nEED aLL tHeSE aRTiSTS, I gave you a pretty convincing refutal to the argument that AI uses more energy than human artists. Not anything else.

[–] msgraves@lemmy.dbzer0.com 1 points 7 months ago (2 children)

that’s the argument OOP was making. I didn’t make it, I never said we should remove human artists. I think human creativity/creation will and should become even more valued in the coming years. You’re putting words in my mouth, that’s not very nice.

[–] msgraves@lemmy.dbzer0.com 1 points 7 months ago (4 children)
[–] msgraves@lemmy.dbzer0.com 1 points 9 months ago

actual example please not like your other friend Luddite on the other comment

[–] msgraves@lemmy.dbzer0.com 3 points 9 months ago

disregarding the fact that the model learns and extrapolates from the training data, not copying,

have fun figuring out which model made the image in the first place!

[–] msgraves@lemmy.dbzer0.com 10 points 9 months ago (4 children)

you’re gonna have a bad time restricting software

[–] msgraves@lemmy.dbzer0.com 5 points 9 months ago

I wish that would stop Nintendo.

[–] msgraves@lemmy.dbzer0.com 2 points 9 months ago

yeah- mistral and llama are the ones you want to looks at- grok is too big to run even on enterprise cards (and sucks worse than a model my pi can run)

[–] msgraves@lemmy.dbzer0.com 5 points 9 months ago

don’t let this turn into r/196

view more: ‹ prev next ›