When you talk about they. You are talking about an inherently imperfect ML model. Even if it was 99.99% accurate there would still be thousands of false positi

When you talk about they. You are talking about an inherently imperfect ML model. Even if it was 99.99% accurate there would still be thousands of false positives positives/negatives like you describe.