There are some things where I think it's fair to never trust that person fully again. Ever. But we need a way to trust them enough to let them live and participate in society if we believe they are rehabilitated while still protecting everyone around them.
I'm sure that's not easy, but it has to be easier than lifetime incarceration.
Look at the prison models of almost any other industrialized Western country - make even the slightest genuine effort to reform people instead of considering them subhuman to be inhumanely tortured by the circumstances of their confinement followed by blocking them from participating in the economy upon release and results will improve.
Improve public education and remove inequalities and you remove crime as the best option for catching up to everyone else.
AI won't be used to help convicts, because nobody in the US wants to help them. It'll be used to better manage their shackles for increased profits.
1) Typically the systems monitoring, if not the systems themselves, is dumped on the police along with the funding. I agree in principle that police data systems should be handled by an arms-length agency without ties to any particular police service. I also believe this should include their body cams, interview room video, and even their fleet and weapons/ammo tracking. They should not have any oversight over their own data because that leads to the potential for abuse.
2) At least where I am... officers can query, but queries of federal databases are audited and monitored. You've never seen someone walked out of a building faster than when they are caught with their hand in that particular cookie jar. And yes, charges happen for the serious incidents. However, that still leaves a lot of room for abuse of non-federal data.
And that title is backed by the fact that a decade ago or so I was implementing proper auditing to track cops because they were... abusing video systems and it made it into the news.
Cops are just people, the badge doesn't confer ethics or strength of character. It often does confer a sense of superiority to the general public and a belief that they're above some of the rules the rest of us abide by.
Even the best, most upright cop should never be taken at their word - there should always be some form of oversight. Because they're humans.
The Genie is out of the bottle. The toothpaste is out of the tube.
ha, i remember those metal toothpaste tubes! Lead? Aluminium? But with the plastic ones, you can suck it back in
Don't we already have the "mass surveillance" you describe?
Payments, phones, number plates
So I asked the ai, and realised what was probably obvious to US readers - the class and race divide in Missouri.
Here its everyone using buses, and scanning their registered cards to get on/off. In Missouri it is poor minorities paying cash?
Do we accept the premise that people can be banned from using the buses for some reason?
Then we must accept you can't get on a bus with your face hidden.
So what are the consequences of being flagged by the face recognition? Is it footage being sent to police?
Is it being asked for ID?
I assume most people using public transport these days are using some sort of electronic payment, so that can be your ID.
What is the scenario where things go horribly wrong?
Yes, bad things have happened where morons treat a face-ID as gospel, but nobody is under any obligation to repeat the worst case stupidity.
while in the west it is private enterprise driven by the profits.
Thats what the cold-war era propaganda said. In reality, public-private partnerships werte essential to the post-war economic boom.
DARPA, public infrastructure, aerospace ,
So why don't you try saying what it actually is, instead of "no it isn't"? Why hide under AC if not actually saying anything?
There is no point saying "its not real", if you can't define what real is, especially in terms of a testable theory. they are empty, meaningless words. Ironically, an AI could show much deeper understanding than the AC - based on observable outputs. (ie empirical evidence)
A. Coward, you keep saying that, but it does not mean anything. You show no understanding, just repeating words you think fit.
So how about defining some terms then? What is understanding, if we are going to try to apply it to non-human intelligence?
For decades we have been applying these questions to animals. Mirror self-recognition, theory of mind etc.
The problem is that most people have really actually thought about it, so do not have a meaningful definition of "understanding". Yet many of them make confident pronouncements anyway. I call this "Natural Stupidity". But they can't help it, they are just a bunch of neurons firing according to simple rules of electrochemistry. There is no such thing as emergence.
current AI has no ability to reason or understand, they are just pattern matching algorithms.
I'm not sure about AI, but human stupidity is very real. And it becomes evident in the slashdot comments for any topic that raises emotions. The rich guy, the orange guy, and AI being prime examples.
Given the rather obvious reasoning of reasoning models, you have to wonder what "no ability to reason" intends to mean. i'm sure bloodhawk isn't a moron, or deliberately trying to sound like a moron, but rather expressing some sort of emotion. I guess "AI reasoning is not the same as human reasoning"? Then why not try to make an observation on the differences?
As for "they are just pattern matching algorithms" - that tells us that not only is the poster ignorant of how AI works, but he has no desire to learn. His hard-wired prejudices are adequate. And that leads to over-confident blanket assertions. Is this what people inaccurately call 'Dunning Kruger"?
Somebody's terminal is dropping bits. I found a pile of them over in the corner.
