It's also important to note that every single AI identification schema to date has a disturbingly high false positive rate for it to be used in military or policing actions. Innocent people are being arrested or killed, and yet often the equipment itself isn't even owned or operated by the government.

Reply to this note

Please Login to reply.

Discussion

I completely agree but I still can't help but jokingly say that innocent people are being arrested and killed already so it wouldn't be much of a change 😄

I believe that the existing issue of arrests on false charges/misidentification is due to the abrogation of responsibility on the parts of the decision makers. These AI tools will further contribute to that problem, but it also presents a huge leap in the number of misidentifications leading to death (esp. military use, as in Gaza currently) or incarceration/charges (like the street cams the London "Bladerunners" are famously disabling) already.

Yes absolutely, AI pretty much only intensifies existing trends that humans have come up with on their own in addition to providing a convenient way to outsource agency and accountability. I have no trouble imagining a version of the old Nuremberg defense that has people claiming that they were simply acting on AI-generated recommendations.

I have heard of the Bladerunners and while some claims and numbers sound exaggerated, I obviously hope they're real. I've only been to London once in 2018 and I did feel watched quite a bit, at least in poorer areas, but since then the number of cameras seems to have exploded so what I've seen then must've been nothing compared to now. Chilling indeed.