No. Cameras will only help potentially identify someone after the fact. At some point we may find ourselves in a Black Mirror dystopia where AI is used to precog individuals to determine someone is going to commit a crime where video surveillance helps locate this person before the crime is committed. However, since no crime has been committed, there's no reason to detain them. Unless, you just really want to commit to that dystopian nightmare of allowing AI to make those decisions.

I think it's unquestionably correct; if someone follows you around and sees you in a bar drinking alcohol and stops you from driving for 8 hours in a way you can't reason with, can't object to, can't override, and that happens to everyone all the time, then drink-driving accidents would drop to zero overnight. If everyone is surveilled constantly then every transgression can be blocked or punished immediately - instant fines deducted from bank accounts (I read that China does that with jaywalkers, facial recognition identifies them and they get a smartphone alert that they have been fined), but beyond that things like everyone having an ID tag and all doors and gates working on it would stop a lot of trespassing. Then swarming the perpetrator with electronic mosquitos with taser zappers and beeping noises would stop a lot of cases of casual harassment and casual theft. Then calling the police to a location and locking up the perpetrator in minutes instead of months, would pretty quickly communicate that crime doesn't pay.

I don't think it's a good idea for a free society, it would be hell to live in. The chilling effect, the number of laws is too high to keep track of, it probably isn't be possible to always be on best behaviour 24/7 for a lifetime.

When robot camera insects with wireless mesh networking and power scavenging hit a few dollars each, there won't be a private space on the planet ever again. Any spider, beetle, fly, in any room, outside any window, on any surface, will be a potential camera drone.