61–70 of 211 posts
https://innocenceproject.org/what-is-bite-mark-evidence-fore...
https://bostonreview.net/books-ideas/nathan-robinson-forensi...
https://archive.nytimes.com/www.nytimes.com/books/first/k/ke...
Earlier quoted context omitted.
If manual reclassification is routine then what does that say about the AI?
That it doesn't give the answers the customers (police and prosecutors) want. Which is mostly unrelated to whether or not it's accurate.
Earlier quoted context omitted.
That's the point? These systems invariably turn out to not be any sort of AI at all, but sales people ready to make a call any way their customer (police) wants it made. Later it's presented as AI to a jury. It's basically the same scam as lots of forensics, junk science to get a desired result.
I generally agree, but wouldn't necessarily say a lot of forensics is junk science. The science is generally sound (procedures and theory). Usually it's the prosecutors that present the forensics in a way that misrepresents what the results actually mean. Or lab techs make mistakes or perform misconduct (How to Fix a Drug Scandal on Netflix is interesting).
Earlier quoted context omitted.
> Crucially, Williams' lawyers – public defenders Lisa Boughton and Brendan Max – said records showed that ShotSpotter actually initially picked up what sounded like a firework a mile away, and this was later reclassified by ShotSpotter staff to be a gunshot at the intersection where and when Williams was seen on camera. ShotSpotter strongly insisted it had not improperly altered any data to favor the police's case,…
Wow, we need a civil rights lawyer to represent Williams in a wrongful imprisonment suit against the city, if not also a civil suit against ShotSpotter (or maybe even a criminal complaint for knowingly providing false testimony?). People shouldn’t be able to just dust their hands of this.
Yet the system will let them. I've had a prosecutor and trooper hold a charge that they know was incorrect and subject us to pre trial restrictions found only under that charge. The prosecutor also used their position to block our ability to have a witness remotely testify by telling court scheduling not to even speak to us (inappropriately using their position to influence the court). The state police said they see nothing wrong with knowingly holding an incorrect charge and subjecting people to pretrial under it. The Bar said that even though the conduct we described would constitute prosecutorial misconduct, they won't pursue complaints against prosecutors unless the court makes a determination supporting that. That would cost thousands of dollars to hire a lawyer, make an appeal, etc. The ACLU said they have bigger issues to deal with. I submitted a complaint to DOJ, but I'm not holding my breath.
Literally everyone we dealt with as a part of the system made a bunch of mistakes and didn't give damn about correcting them or protecting people's rights. It's us vs them and all about the money.
Earlier quoted context omitted.
I wouldn't call polygraphs forensics. Most states do not allow them as evidence. Do you have a link for the fingerprinting claim?
(Not who you responded to) https://www.crammlawfirm.com/fingerprint-match-jails-innocen... This isn't the case I originally recalled, and doesn't say how many points were matched, but I remember a 5 point match being used to convict a person who was later proven innocent.
Earlier quoted context omitted.
I generally agree, but wouldn't necessarily say a lot of forensics is junk science. The science is generally sound (procedures and theory). Usually it's the prosecutors that present the forensics in a way that misrepresents what the results actually mean. Or lab techs make mistakes or perform misconduct (How to Fix a Drug Scandal on Netflix is interesting).
No, the procedures and theory behind, e.g., bite mark analysis or fire analysis, are total junk.
Earlier quoted context omitted.
It may not be right to blame AI, but it's certainly the case that they were presenting it as "the system says there was a shot fired here." The bigger point is that the "system" can't be trusted. And the fact that they need to manually classify things is proof of that fact. (It's way more complicated than just an audio recording. It's recordings from numerous locations, attempting to triangulate the original location…
> It may not be right to blame AI, but it's certainly the case that they were presenting it as "the system says there was a shot fired here." I think a crucial distinction is to clarify who “they” is: Who was making the claim? From the article, it seems that the prosecutors may have been the ones exaggerating the evidence: > Prosecutors said ShotSpotter picked up a gunshot sound where Williams was seen on surveillanc…
The prosecutors are the "they". The further removed you get from people who know the 'AI' the more the validity of the evidence provided by the AI relies merely on what others have said.
The people working on the system know that it can be faulty. Other software developers have an idea that it could be faulty. Prosecutors are told that it's alright. They turn around and argue it in a court of law. And the further you get the less people have any idea that the system is faulty. Down the line people will go "the machine says so, therefore it must be so".
Who would have thought?
Shocking, shocking I tell you.
https://venturebeat.com/2021/04/05/government-audit-of-ai-wi...
Earlier quoted context omitted.
No, the procedures and theory behind, e.g., bite mark analysis or fire analysis, are total junk.
Do you have some links for this? My understanding is the interpretation part can be BS, but stuff like finding and identifying potential ignition or accelerant locations and types should be more settled.
The article says that employees of the AI company (ShotSpotter) manually reviewed and classified the sounds as gunshots: > records showed that ShotSpotter actually initially picked up what sounded like a firework a mile away, and this was later reclassified by ShotSpotter staff to be a gunshot at the intersection where and when Williams was seen on camera. So the AI didn’t even make the call. The staff did, manually.…
That's the point? These systems invariably turn out to not be any sort of AI at all, but sales people ready to make a call any way their customer (police) wants it made. Later it's presented as AI to a jury. It's basically the same scam as lots of forensics, junk science to get a desired result.
That said, human in the loop ML can be extremely useful. Surfacing possible positive examples for human review and/or cutting down 'obvious' negatives can multiply the productivity of human reviewers in many contexts.