Yeah that was my thought. Don’t reject them, that’s obvious and they’ll work around it. Feed them shit data - but not too obviously shit - and they’ll not only swallow it but eventually build up to levels where it compromises them.
I’ve suggested the same for plain old non-AI data stealing. Make the data useless to them and cost more work to separate good from bad, and they’ll eventually either sod off or die.
A low power AI actually seems like a good way to generate a ton of believable - but bad - data that can be used to fight the bad AI’s. It doesn’t need to be done real-time either as datasets can be generated in advance
SorteKanin@feddit.dk 1 month ago
Even “high power” AIs would produce bad data. It’s currently well known that feeding AI data to an AI model decreases model quality and if repeated, it just becomes worse and worse. So yea, this is definitely viable.
phx@lemmy.ca 5 weeks ago
Yup. It was more my thought that a low power over could produce sufficient results while requiring less resources. Something that can run on a desktop computer could still produce a database with reams of believable garbage that would take a lot of resources from the attacking AI to sort through, or otherwise corrupt its own harvested cache