Is anyone using Nightshade, specifically for Meta postings? If you do, do you mind posting your experience here (specific question below) or speaking with me voice to voice for 5 minutes?
Background: one of my commercial clients is experimenting with Nightshade, the AI poisoning app created by University of Chicago. Specifically, they would like to use it for social media (Facebook, Instagram and Thread) posts. Meta says that one can 'opt out' from AI, but some studies show that's it's BS, and that that the rights given to them when posting on their apps are vast.
The question is whether Meta can detect poisoning and punish it by using algorithm to shadow-ban the post.
We are running an experiment, posting two identical images, one poisoned (default intensity and slower render settings) one pure to see if we see a difference in interactions, but if someone here already tried it, would love to learn from your experience.
Thanks in advance.
Background: one of my commercial clients is experimenting with Nightshade, the AI poisoning app created by University of Chicago. Specifically, they would like to use it for social media (Facebook, Instagram and Thread) posts. Meta says that one can 'opt out' from AI, but some studies show that's it's BS, and that that the rights given to them when posting on their apps are vast.
The question is whether Meta can detect poisoning and punish it by using algorithm to shadow-ban the post.
We are running an experiment, posting two identical images, one poisoned (default intensity and slower render settings) one pure to see if we see a difference in interactions, but if someone here already tried it, would love to learn from your experience.
Thanks in advance.