- cross-posted to:
- [email protected]
- [email protected]
- cross-posted to:
- [email protected]
- [email protected]
Only halfway in and it’s already very interesting and too much for me to summarize but I do love the idea of subtly releasing music to poison models.
The demonstration at 15:07 is a good preview but the entire video is worthwhile.
Or 22:25 for a more malicious use of this tech.
Edit: fixed timestamped links.
The point of this method is that the music it self contains data that cannot be heard by our ears, but that causes the ai to encode the data completely wrong.
Like others have said. Watch the video, I swear it’s interesting and most likely worth your time. :)