What's your threshold for when it becomes a problem? Should we wait until it becomes a problem, or should we try to stop this level of facial recognition?
You should also assume this is a proof of concept. It'll get improved and scaled down to run on every police vehicle, and on every camera the police already control.
It has already been scaled down to android phones (you'll find phones are an excellent platform for this), where you can find apps that are meant to let venue-owners guard entrances against specific individuals. That's illegal, but obviously common enough to make such apps.
Then it will be sold to the public as being successful (they are already claiming that in the article itself that it is successful). Then that will be used to justify them in other places.
