Comment by delichon
1 day ago
> Regardless, the true takeaway of Isaacs’ experience is that ALPR technology, whether it’s from Flock or any of the other competitors in this space providing the same functionality (Axon, etc), is dangerous not only because it is so easily abused by police (or even by hackers) to invade the privacy and liberty of citizens, but also because it enables a lazy police department to outsource basic critical thinking to machines and then take action based on a single data point to ruin a person’s life.
The accusation that some machine made any decision here goes beyond the reporting in the article. It seems concocted to raise AI alarm. All we have is that it supplied accurate data that police then abused. That isn't in the vicinity of "outsource[d] basic critical thinking to machines".
I want such general government surveillance banned, but the machines in this story had no more volition than a breathalyzer.
The machines matter, because we know how human behavior works. We know that if you give humans an automated tool that works most of the time, they get lazy, to the point where they make more mistakes than if they had no automation at all.
Flock isn’t an automated tool though; you don’t give it a crime, and it gives back an automated package saying “this is who the criminal is”.
As far as I know, it lets law enforcement query for things against the data. Can be specific (like a license plate) or more general (like a dark colored mid-sized SUV around <location> at <time>).
To get where this story is at, cops here must have done the latter, and taken the first name off the top of the pile and patted themselves on the back, and then waited 7 months, and went and arrested her. Unless there’s an explanation for a 7 month delay, it seems designed to allow defensive evidence to spoil.
There isn’t an uncanny-valley of automation here. There is just lazy, grossly-incompetent, bad-faith, and malicious cops/investigators.
> Flock isn’t an automated tool though; you don’t give it a crime, and it gives back an automated package saying “this is who the criminal is”.
Please tell the police that, then, under penalty of being blacklisted if they ever violate it. They're the ones treating it as such.
I think what we're seeing is that flock is presented to police as a reliable technology that can be trusted for this sort of thing. If a random person had snapped a picture they would have questioned it.
But they didn't question the quality of the evidence. And there is a reason for that.
This should be emphasized more. Regardless of how technology could be used, we (should be) are responsible for how it is likely to be used.
Guns can be used for all sorts of things. A handgun in a house is likely to be used for murder/suicide.
> A handgun in a house is likely to be used for murder/suicide.
On an article about people who did no research to actually prove out their assumptions, this is perhaps the most ironic comment to be chastising them.
1 reply →
"A handgun in a house" is statistically extremely unlikely to be "used for murder/suicide."
1 reply →
Volition has nothing to do with it. A breathalyser doesn't point to a person and say this is the likely drunk, so that comparison is fatuous. These identification systems do point the finger.
Police abusing the law? Never.
Piles of evidence that humans tend to treat AI as infallible, just like the police did here. They didn't use this as a point for further investigation, they saw the match and said "let's go get her" without so much as looking at the damaged vehicle for paint transfer, which should be remedial-level police instruction.
Modern AI was and is a mistake. It's only going to get worst before it gets better.
Unfortunately, our current government in the US is hard AI-pilled (oh, sorry, "super-intelligence-pilled"). They use it every day to present propaganda to the public via social media.
Ban AI use in all branches of government, and in policing. At least until there are ethical standards put in place that are very heavily punished when violated.
Also, state and federal level politicians should be prohibited from using social media in anything other than a purely personal basis without mention of political office or without speaking on the topic of politics. Want to post cool pictures of your family? Go for it. Speaking to your constituency? Must be strongly vetted for ethical violations, and punishable by immediate removal from office and a 10-year blacklist.