Boosted by brib@bribstodon.xyz ("brib :neofox_floof: :Nonbinary:"):
david_chisnall@infosec.exchange ("David Chisnall (*Now with 50% more sarcasm!*)") wrote:
RE: https://cosocial.ca/@mhoye/117010519416979095
That’s such a negative framing. I prefer to think of it as:
If a company with a ‘Generate this with AI’ button also gives you a ‘Report AI slop’ button, you should take advantage of this golden opportunity to poison their training data. A few ideas:
- Flag only posts that are not obviously slop.
- Flag every single post that mentions ‘AI’ in a vaguely positive way.
- Flag every post that expresses positive sentiment towards Microsoft, Google, or Meta.
- Spend a day finding posts that randomly tell everyone to install Linux, as the solution to any problem, like them and flag every other post you see.