the AI decided ¯_(ツ)_/¯

Outside of law enforcement, this is certainly how shitty customer service policies get enforced. In other words, "Computer says no".

Now where does this thought come from?

Do you not know what a computer is? It's literally a digital logical accountant! Yeah yeah, we should probably blame the programmers and engineers instead when shit goes sideways, but now I think we need to also hold CEOs accountable when they decide to inject faulty AI into mission critical systems...

https://lemmy.dbzer0.com/post/55990956

IT disagrees. Misbehaving hardware can be taken out back and shot.

That's why cops love using dogs too. Courts have ruled that dogs can't lie. That means if a dog indicates you have contraband, then a search is warranted, even if nothing is found. This of course ignores that it is entirely possible the dog indicated contraband because the cop trained it to do so on command.

You could hold developers of algorithms, logic and even symbolic AI accountable.

However, it's a completely different story for AI based on deep neutral networks. After training they're just a bunch of weights and parameters without individual meaning and it's not a few, it's billions or trillions of them. And almost none of them were individually set, they're often randomly initialized and then automatically tuned by deep learning algorithms during training until the behavior / predictions of the neural net are "good enough".

It's practically impossible to review the network and when you test it you just get the result for the concrete test cases, you can't interpolate or assume even slightly different cases will behave similarly. You also can't fix an individual bug. You can just train again or more and this effort might fix the problem but it could also destroy something that worked before (catastrophic forgetting).

We don't jail the gun for murder.

this is straight out of the book 'Do Android Dream Of Electric Sheep' later turned into the movie 'iRobot'

All technology has the potential to be both liberatory and oppressive, all that ever matters is who wields it and to what end.

Lewis Herber (Murray Bookchin) - Towards a Liberatory Technology

Law enforcement will seize and use computers and the data they hold as evidence to convict criminals, just like any other tool that they might be warranted to seize.

Courts will examine the evidence of what it did to determine what role it played in the offence and whether it supports the allegation.

Likewise police complaints authorities could do the same in principle against the police; if someone were to give them a warrant and the power to execute it.

If a thing happens in public that was unwarranted and can be traced back to a police force or how they deployed any equipment, they can be judicially reviewed* for any decision to deploy that bit of kit. It's more a matter of will they actually be JR'd and will that be review be just and timely. * - in my country.

I don't think it's much different from how they deploy other tech like clubs and pepper spray, tear gas, tazers or firearms. If they have no fear of acting outwith their authority , that's a problem.

In some ways it might be easier to have an 'our word' vs 'their word' defense when they shoot someone, compared to a computer program that might literally document the abuse of power in its code or log files.

"Oops i dropped my notebook", is maybe easier than, "oops i accidentally deleted my local file and then sent a request to IT - that was approved by my manager - asking them to delete instead of restore any onsite or offsite backups".

This is unironically one of the main drivers of AI. As soon as all crucial social systems are inundated with AI, the built-in bias will be excused as "minor glitches" of the system, but the real reason was always a total lack of accountability.

Yeah, not like people in power, who are held accountable all the time!

Same argument for cars