4 ms·
This argument is utterly incoherent. Of course they don't include into the false positive rate images that are intentionally designed to generate positive hits,
by saynay 5y ago
This argument is utterly incoherent. Of course they don't include into the false positive rate images that are intentionally designed to generate positive hits, what would be the point of that? The only interesting metric is the false positive rate for normal images.
- saithound 5y agoIt's not an argument. It's an explanation of what the number means. > The only interesting metric is the false positive rate for normal images. Wrong. The only interesting metric is the false positive rate _in practice_: i.e. how likely are false positives to affect innocents. Indeed, Apple is presenting their "one in a trillion" number as reassurance, as if it was the probability that an account that doesn't distribute CSAM gets flagged by their system. But that probability depends strongly on adversarial questions, and cannot be calculated using optimistic assumptions about all images being "normal".