Most of the surveillance stories we’ve looked at here lately have involved things you can at least see when you walk past them: cameras at the gate, sensors on the shelf, a label adjusting a price as you move through space.
While it surely can detect whether you’re doing something on your SSD, I’m highly doubtful that it could be used to accurately predict what that is. The training model was highly specialized to the target machine and with only certain application genres (e.g. iMac music, videos app store, etc), and only lists two websites (google.com, youtube.com) as identification. In addition, many websites and applications simply don’t access the disk.
Great. Just what we need.
https://hannesweissteiner.com/pdfs/frost.pdf
While it surely can detect whether you’re doing something on your SSD, I’m highly doubtful that it could be used to accurately predict what that is. The training model was highly specialized to the target machine and with only certain application genres (e.g. iMac music, videos app store, etc), and only lists two websites (google.com, youtube.com) as identification. In addition, many websites and applications simply don’t access the disk.