The problem of these scalable approaches is well documented (not scientifically as far as I know), but in several forms of media. You said well the consequence

The problem of these scalable approaches is well documented (not scientifically as far as I know), but in several forms of media. You said well the consequences: - False positives - the wrongfully labelled ones ; - """hackers""" - because there's not enough amount of '"' to make it sound different but bare with me, it's just those who learn/are lucky to manipulate/circumvent the system; Even if this are small percentage cases, it will be in the order of hundreds of thousands, if not millions of users. How do you give support to this people? Is it going to be another Youtube fuck-fest where you end up talking to bots? Because this shouldn't be allowed - at all. If you want to make automated control measures, then people should have the right to talk directly to a human who can actually solve their problems. Yet that seems to be impossible, because of the scale.