science

14779 readers

49 users here now

A community to post scientific articles, news, and civil discussion.

rule #1: be kind

<--- rules currently under construction, see current pinned post.

2024-11-11

founded 1 year ago

MODERATORS

m3t00@lemmy.world

laverabe@lemmy.world

DeadPand@midwest.social

Joleee@lemmy.world

230

AI models fed AI-generated data quickly spew nonsense (www.nature.com)

submitted 3 months ago by ArcticDagger@feddit.dk to c/science@lemmy.world

52 comments fedilink hide all child comments

you are viewing a single comment's thread
view the rest of the comments

[–] MarauderIIC@dormi.zone 13 points 3 months ago (7 children)

I wonder if the speed at which it degrades can be used to detect AI-generated content.

[–] maniclucky@lemmy.world 6 points 3 months ago (5 children)

I wouldn't be surprised if someone is working on that as a PhD thesis right now.

[–] Ferris@infosec.pub 1 points 3 months ago (1 children)

how are you going to write a thesis on writing a FLAC to disc and ripping it over and over?

[–] maniclucky@lemmy.world 2 points 3 months ago (1 children)

By measuring how it does with real images vs generated ones to start. The goal would be to show a method to reliably detect ai images. Gotta prove that it works.

[–] KeenFlame@feddit.nu 1 points 3 months ago (1 children)

How would it detect, you would need the model and if you do you can already detect

[–] maniclucky@lemmy.world 1 points 3 months ago (1 children)

It's an issue with the machine learning technique, not the specific model. The hypothetical thesis would be how to use this knowledge in general.

Why are you so agitated by my off hand comment?

[–] KeenFlame@feddit.nu 1 points 3 months ago

Am I agitated? 😂 💜 You it's not with all models no

load more comments (3 replies)

load more comments (4 replies)