Meta has announced that it will begin training its own AI using data from European users. The company claims that it has a legitimate interest in this practice. Users can object to the use of their data; however, there have been complaints regarding the procedure for submitting these objections.
To object, users must fill out a form, detailing how the data
The only exception is private messages, and some users have reported difficulty opting out.
What if we poison all the data with garbage?
Honestly thats my plan, subtly adjust all my buffalo buffalo buffalo buffalo comments to induce stupid behaviour in any AI thats trained on them
Like fake news, aggressively inaccurate emotional stances on basic issues and general dickery? Done.
Facebook essential oil and bleach enema boomers are way ahead of you on that one.
Truly a miracle that LLaMa or Mistral ever output anything coherent.