Here is a part that I personally have to wrestle with:
> "They never even asked him any questions before arresting him. They never asked him if he had an alibi. They never asked if he had a red Cardinals hat. They never asked him where he was that day," said lawyer Phil Mayor with the ACLU of Michigan.
When I was fired by an automated system, no one asked if I had done something wrong. They asked me to leave. If they had just checked his alibi, he would have been cleared. But the machine said it was him, so case closed.
Not too long ago, I wrote a comment here about this [1]:
> The trouble is not that the AI can be wrong, it's that we will rely on its answers to make decisions.
> When the facial recognition software combines your facial expression and your name, while you are walking under the bridge late at night, in an unfamiliar neighborhood, and you are black; your terrorist score is at 52%. A police car is dispatched.
Most of us here can be excited about Facial Recognition technology but still know that it's not something to be deployed in the field. It's by no means ready. We might even consider the moral ethics before building it as a toy.
But that's not how it is being sold to law enforcement or other entities. It's _Reduce crime in your cities. Catch criminals in ways never thought possible. Catch terrorists before they blow up anything._ It is sold as an ultimate decision maker.
52% is little better than a coin flip. If you have a million individuals in your city, your confidence should be in the ballpark of 99.9999% (1 individual in 1 million). That has really been my concern with this, the software will report any facial match above 75% confidence. Apart from the fact that it appalling confidence, no cop will pay attention to that percentage; immediately arresting or killing the individual.
Software can kill. This software can kill 50% of black people.
Software can kill if we put blind trust in it and give it full control over the situation. But we shouldn't do that.
Even if it was correct 99% of the time, we need to recognize that software can make mistakes. It is a tool, and people need to be responsible enough to use it correctly. I think I agree with your general idea here, but to put all of the blame on software strikes me as an incomplete assessment. Technically the software isn't killing anyone, irresponsible users of it are.
Then the justice system would implode. Judicial policy is "software" too, and nobody holds the judiciary or police to that absurd level of excellence, even if we're talking about the death penalty.
I would hope that the software/ML engineers who wrote it knows about probability theory, and why the prior probability should be set at 0.0001% or so.
So that if we print 52% on the screen, that means we've already gathered like 30-bits of evidence (30 coin flips all coming up heads), at which point the suspicion would be real.
It would have to be even higher than that level of accuracy, because every person is going to be 'tested' multiple times.... if everyone's face is scanned 100 times a day, the number of false positives is going to be even higher.
It’s not at all obvious to me that the accuracy threshold should scale with city size. Some small town shouldn’t use a system that is 1000x less accurate.
This shit right here. This is why I don't stop for the inventory control alarms at department store doorways if they go off. I know I've paid, and little sirens are just a nuisance at this point.
This is why I've never stopped for receipt checks, because it's my receipt, and I've paid. The security theatre is just bad comedy.
Just because the machine says I've done a no no, doesn't mean I can't come back and win a lawsuit later. It doesn't abdicate cops from their jobs. I have a winning complexion, so I'll never enjoy a false positive, but if I do, I'll make sure it bankrupts whatever startup incubator garbage decided to shill a replacement for real law enforcement.
Worse - an AI decision puts an obligation on the user to follow it. What do I mean? Well - imagine you are a cop, you get an auto flag to arrest someone and use your discretion to overide it. The person goes on to do something completely different; like they are flagged as a murderer but then go and kill someone DUI. You will be flayed, pilloried. So basically safety first, just do the arrest. The secret is that these systems should not be making calls in this kind of context because they just aren't going to be good enough, it's like cancer diagnosis - the oncologist should have the first say, the machine should be a safety net.
You see it everywhere with AI and other tools. We overly trust them. Even when doctors have a high confidence in their diagnosis, they accept wrong AI-recommended conclusion that contradicts it.
That's interesting. I can imagine in cases like this it's not necessarily that the doctor doubts their own diagnosis, but rather the AI is essentially offering to relieve them of responsibility for it either way.
It's like in human hierarchies - it's often not the person who is more likely to make the best decision who gets to decide, it's the one who is going to bear the consequences of being wrong.
Exactly what I thought when I've read about this. It's not like humans are great at matching faces either. In fact machines have been better at facial recognition for over a decade now. I bet there are hundreds of people (of all races) in prison right now who are there simply because they were mis-identified by a human. Human memory, even in the absence of bias and prejudice, is pretty fallible.
There is a notion of "mixture of experts" in machine learning. It's when you have two or more models that are not, by themselves, sufficiently good to make a robust prediction, but that make different kinds of mistakes, and you use the consensus estimate. The resulting estimate will be better than any model in isolation. The same should be done here - AI should be merely a signal, it is not a replacement for detective work, and what's described in the article is just bad policing. AI has very little to do with that.
The problem is not with the technology, but with how it's used. A medical test is also not 100% error-proof which is why a professional needs to interpret the results, sometimes conducting other tests or disregarding it completely.
A cop stopping someone that has a resemblance to a criminal for questioning seems like a good thing to me, as long as the cop knows that there's a reasonable chance it's the wrong guy.
>But that's not how it is being sold to law enforcement or other entities. It's _Reduce crime in your cities. Catch criminals in ways never thought possible. Catch terrorists before they blow up anything._ It is sold as an ultimate decision maker.
None of those selling points logically lead to the conclusion that it is the ultimate decisions maker.
This story is really alarming because as described, the police ran a face recognition tool based on a frame of grainy security footage and got a positive hit. Does this tool give any indication of a confidence value? Does it return a list (sorted by confidence) of possible suspects, or any other kind of feedback that would indicate even to a layperson how much uncertainty there is?
The issue of face recognition algorithms performing worse on dark faces is a major problem. But the other side of it is: would police be more hesitant to act on such fuzzy evidence if the top match appeared to be a middle-class Caucasian (i.e. someone who is more likely to take legal recourse)?
Essentially, an employee of the facial recognition provider forwarded an "investigative lead" for the match they generated (which does have a score associated with it on the provider's side, but it's not clear if the score is clearly communicated to detectives as well), and the detectives then put the photo of this man into a "6 pack" photo line-up, from which a store employee then identified that man as being the suspect.
Everyone involved will probably point fingers at each other, because the provider for example put large heading on their communication that, "this is not probable cause for an arrest, this is only an investigative lead, etc.", while the detectives will say well we got a hit from a line-up, blame the witness, and the witness would probably say well the detectives showed me a line-up and he seemed like the right guy (or maybe as is often the case with line-ups, the detectives can exert a huge amount of bias/influence over witnesses).
EDIT: Just to be clear, none of this is to say that the process worked well or that I condone this. I think the data, the technology, the processes, and the level of understanding on the side of the police are all insufficient, and I do not support how this played out, but I think it is easy enough to provide at least some pseudo-justification at each step along the way.
That's interesting. In many ways, it's similar to the "traditional" process I went through when reporting a robbery to the NYPD 5+ years ago: they had software where they could search for mugshots of all previously convicted felons living in a x-mile radius of the crime scene, filtered by the physical characteristics I described. Whether the actual suspect's face was found by the software, it was ultimately too slow and clunky to paginate through hundreds of results.
Presumably, the facial recognition software would provide an additional filter/sort. But at least in my situation, I could actually see how big the total pool of potential matches and thus have a sense of uncertainty about false positives, even if I were completely ignorant about the impact of false negatives (i.e. what if my suspect didn't live within x-miles of the scene, or wasn't a known/convicted felon?)
So the caution re: face recognition software is how it may non-transparently add confidence to this already very imperfect filtering process.
(in my case, the suspect was eventually found because he had committed a number of robberies, including being clearly caught on camera, and in an area/pattern that was easy to narrow down where he operated)
> and the detectives then put the photo of this man into a "6 pack" photo line-up, from which a store employee then identified that man as being the suspect.
This is absurdly dangerously. The AI will find people who look like the suspect, that’s how the technology works. A lineup as evidence will almost guarantee a bad outcome, because of course the man looks like the suspect!
> Essentially, an employee of the facial recognition provider forwarded an "investigative lead" for the match they generated (which does have a score associated with it on the provider's side, but it's not clear if the score is clearly communicated to detectives as well)
THIS DOCUMENT IS NOT A POSITIVE IDENTIFICATION. IT IS AN INVESTIGATIVE LEAD ONLY AND IS NOT PROBABLE CAUSE TO ARRREST. FURTHER INVESTIGATION IS NEEDED TO DEVELOP PROBABLE CAUSE TO ARREST.
This is why you should be scared of this tech. Computer assisted patsy finder. No need to find the right guy when the ai will happily cough up 20 people nearby who kinda sorta look like the perp enough to stuff them into a lineup in front of a confused and highly fallible witness.
I'm becoming increasingly frustrated with the difficulty in accessing primary source material. Why don't any of these outlets post the surveillance video and let us decide for ourselves how much of a resemblance there is.
> the detectives then put the photo of this man into a "6 pack" photo line-up, from which a store employee then identified that man
This is not correct. The "6-pack" was shown to a security firm's employee, who had viewed the store camera's tape.
"In this case, however, according to the Detroit police report, investigators simply included Mr. Williams’s picture in a “6-pack photo lineup” they created and showed to Ms. Johnston, Shinola’s loss-prevention contractor, and she identified him." [1]
> Does this tool give any indication of a confidence value?
Yes.
> Does it return a list (sorted by confidence) of possible suspects,
Yes.
> ... or any other kind of feedback that would indicate even to a layperson how much uncertainty there is?
Yes it does. It also states in large print heading “THIS DOCUMENT IS NOT A POSITIVE IDENTIFICATION IT IS AN INVESTIGATIVE LEAD AND IS NOT PROBABLE CAUSE TO ARREST”.
You can see a picture of this in the ACLU article.
The police bungled this badly by setting up a fake photo lineup with the loss prevention clerk who submitted the report (who had only ever seen the same footage they had).
However, tools that are rife for misuse do not get a pass because they include a bold disclaimer. If the tool/process can not prevent misuse, the tool/process is broken and possibly dangerous.
That said, we have little data on how often the tool results in catching dangerous criminals versus how often it misidentifies innocent people. We have little data on if those innocent people tend to skew toward a particular demographic.
But I have a fair suspicion that dragnet techniques like this unfortunately can be both effective and also problematic.
I think the software would be potentially less problematic if the victim/witness were given access, and (ostensibly) could see the pool of matches and how much/little the top likely match differed from the less confident matches.
> The police bungled this badly by setting up a fake photo lineup...*
FWIW, this process is similar to traditional police lineups. The witness is shown 4-6 people – one who is the actual suspect, and several that vaguely match a description of the suspect. When I was asked to identify a suspect in my robbery, the lineup included an assistant attorney who would later end up prosecuting the case. The police had to go out and find tall slight-skinned men to round out the lineup.
> ... with the loss prevention clerk who submitted the report (who had only ever seen the same footage they had).
Yeah, I would hope that this is not standard process. The lineup process is already imperfect and flawed as it is even with a witness who at least saw the crime first-hand.
Intresting and related, a team made a neat "face depixelizer" that takes a pixelated image and uses machine learning to generate a face that should match the pixelated image.
What's hilarious is that it makes faces that look nothing like the original high-resolution images.
Interesting... Neat... Hilarious... In light of the submission and the comment you're responding to, these are not the words I would choose.
I think there's genuine cause for concern here, especially if technologies like these are candidates for inclusion in any real law enforcement decision-making.
What's sad is that a tech entrepreneur will definitely add that feature and sell it to law enforcement agencies that believe in CSI magic: https://www.youtube.com/watch?v=Vxq9yj2pVWk
Ironically, if the police had used and followed the face depixelizer then we may not have had the false arrest of a black man - not because of accuracy but because it doesn't produce many black faces
People are not good at understanding uncertainty and its implications, even if you put it front and center. I used to work in renewable energy consulting and I was shocked by how aggressively uncertainty estimates are ignored by those whose goals they threaten.
In this case, it's incumbent on the software vendors to ensure that less-than-certain results aren't even shown to the user. American police can't generally be trusted to understand nuance and/or do the right thing.
I blame TV shows like CSI and all the other crap out there that make pixelated images look like something you could "Zoom" into or something the computer can still understand even if the eye does not. Because of this, non tech people do not really understand that pixelated images have LOST information. Add that to the racial situation in the U.S. and the the inaccuracy of the tool for black people. Wow, this can lead to some really troublesome results
> But the other side of it is: would police be more hesitant to act on such fuzzy evidence if the top match appeared to be a middle-class Caucasian (i.e. someone who is more likely to take legal recourse)?
Honest question: does race predict legal recourse when decoupled from socioeconomic status, or is this an assumption?
Race and socioeconomic status are deeply intertwined. Or to be more blunt - US society has kept black people poorer. To treat them as independent variables is to ignore the whole history of race in the US.
>Honest question: does race predict legal recourse when decoupled from socioeconomic status, or is this an assumption?
I think the issue is that regardless of the answer, it isn't decoupled in real world scenarios.
I think the solution isn't dependent upon race either. It is to ensure everyone have access to legal recourse regardless of socioeconomic status. This would have the side effect of benefiting races correlated with lower socioeconomic status more.
Middle class black people often get harassed by police, and there is a long history of far steeper sentences for convictions for drugs used more by the black population (crack) than that used more by the white population (cocaine).
So unequal treatment based on race has quite literally been a feature of the US justice system, independent of socioeconomic status.
> The issue of face recognition algorithms performing worse on dark faces is a major problem.
This needs to be coupled with the truth that people (police) without diverse racial exposure are terrible at identifying people outside of their ethnicity. In the photo/text article they show the top of the "Investigative Lead Report" as an image. You mean to say that every cop who saw the two images side by side did not stop and say "hey, these are not the same person!" They did not, and that's because their own brains' could not see the difference.
This is a major reason police forces need to be ethnically diverse. Just that enables those members of the force who never grew up or spent time outside their ethnicity can learn to tell a diverse range of similar but different people outside their ethnicity apart.
This is a classic example of the false positive rate fallacy.
Let's say that there are a million people, and the police have photos of 100,000 of them. A crime is committed, and they pull the surveillance of it, and match against their database. They have a funky image matching system that has a false positive rate of 1 in 100,000 people, which is way more accurate than I think facial recognition systems are right now, but let's just roll with it. Of course, on average, this system will produce one positive hit per search. So, the police roll up to that person's home and arrest them.
Then, in court, they get to argue that their system has a 1 in 100,000 false positive rate, so there is a chance of 1 in 100,000 that this person is innocent.
Wrong!
There are ten people in the population of 1 million that the software would comfortably produce a positive hit for. They can't all be the culprit. The chance isn't 1 in 100,000 that the person is innocent - it is in fact at least 9 out of 10 that they are innocent. This person just happens to be the one person out of the ten that would match that had the bad luck to be stored in the police database. Nothing more.
If I'm searching for a murderer in a town of 1000, it takes about 10 independent bits of evidence to get the right one. And when I charge someone, I must already have the vast majority of that evidence. To say "oh well we don't know that it wasn't Mr. or Mrs. Doe, let's bring them in" is itself a breach of the Does' rights. I'm ignoring 9 of the 10 bits of evidence!
Using a low-accuracy facial recognition system and a low-accountability lineup procedure to elevate some random man who did nothing wrong from presumed-innocent to 1-in-6 to prime suspect, without having the necessary amount of evidence, is committing the exact same error and is nearly as egregious as pulling a random civilian out of a hat and charging them.
There's a good book called "The Drunkards Walk", that describes a woman who was jailed after having 2 children die from SIDS. They argued that the odds of this happening is 1 in a million (or something like that), so probably the woman is a baby killer. The prosecution had statisticians argue this. The woman was found guilty.
She later won on appeal in part because the defense showed that the testimony and argument of the original statisticians were wrong.
This stuff is so easy to get wrong. A little knowledge of statistics can be dangerous.
He wasn't arrested until the shop owner had also "identified" him. The cops used a single frame of grainy video to pull his driver's license photo, and then put that photo in a lineup and showed the store clerk.
The store clerk (who hadn't witnessed the crime and was going off the same frame of video fed into the facial recognition software) said the driver's license photo was a match.
There are several problems with the conduct of the police in this story but IMHO the use of facial recognition is not the most egregious.
The story is the same one that all anti-surveillance, anti-police militarization, pro-privacy, and anti-authoritarian people foretell. Good technology will be used enable, amplify, and justify civil rights abuses by authority figures from your local beat cop, to a faceless corporation, a milquetoast public servant, or the president of the United States.
Our institutions and systems (and maybe humans in general) are not robust enough to cleanly handle these powers, and we are making the same mistake over and over and over again.
Correct, and this has been the story with every piece of technology or tool we've ever given to police. We give them body cameras and they're turned off or used to create FPS-style snuff films of gunned down citizens. Give them rubber bullets and they're aimed at protesters eyeballs. Give them tasers and they're used as an excuse to shoot someone when the suspect "resists." Give them flashbangs and they'll throw them into an infant's crib. Give them mace and it's used out of car windows to punish journalists for standing on the sidewalks.
The mistake is to treat any police department as a good-faith participant in the goal of reducing police violence. Any tool you give them will be used to brutalize. The only solution is to give them less.
"The Shinola shoplifting occurred in October 2018. Katherine Johnston, an investigator at Mackinac Partners, a loss prevention firm, reviewed the store’s surveillance video and sent a copy to the Detroit police"
"In this case, however, according to the Detroit police report, investigators simply included Mr. Williams’s picture in a “6-pack photo lineup” they created and showed to Ms. Johnston, Shinola’s loss-prevention contractor, and she identified him. (Ms. Johnston declined to comment.)"
I think you're correct that the person was not an owner or clerk. IMHO the salient point is that the person was not any sort of eyewitness but merely comparing the same grainy photo as the algorithm.
Yes, this is a story of police misconduct. The regulation of facial recognition that is required is regulation against police/authority stupidity. The FR system aids in throwing away misses, leaving investigative leads. But if a criminal is not in the FR database to begin with, any results of the FR are wastes of time.
> "I picked it up and held it to my face and told him, 'I hope you don't think all Black people look alike,' " Williams said.
I'm white. I grew up around a sea of white faces. Often when watching a movie filled with a cast of non-white faces, I will have trouble distinguishing one actor from another, especially if they are dressed similarly. This sometimes happens in movies with faces similar to the kinds I grew up surrounded by, but less so.
So unfortunately, yes, I probably do have more trouble distinguishing one black face from another vs one white face from another.
This is known as the cross-race effect and it's only something I became aware of in the last 5-10 years.
Add to that the fallibility of human memory, and I can't believe we still even use line ups. Are there any studies about how often line ups identify the wrong person?
I lived in South Africa for a while and heard many times, with various degrees of irony, "you white people all look the same" from black South Africans. So yeah it's definitely a cross-racial recognition problem, and it's probably also a problem with distinguishing between members of visible minorities using traits beyond the most noticable othering characteristic.
There is just so much wrong with this story. For starters:
The shoplifting incident occurred in October 2018 but it wasn’t until March 2019 that the police uploaded the security camera images to the state image-recognition system but the police still waited until the following January to arrest Williams. Unless there was something special about that date in October, there is no way for anyone to remember what they might have been doing on a particular day 15 months previously. Though, as it turns out, the NPR report states that the police did not even try to ascertain whether or not he had an alibi.
Also, after 15 months, there is virtually no chance that any eye-witness (such as the security guard who picked Williams out of a line-up) would be able to recall what the suspect looked like with any degree of certainty or accuracy.
This WUSF article [1] includes a photo of the actual “Investigative Lead Report” and the original image is far too dark for a anyone (human or algorithm) to recognise the person. It’s possible that the original is better quality and better detail can be discerned by applying image-processing filters – but it still looks like a very noisy source.
That same “Investigative Lead Report” also clearly states that “This document is not a positive identification … and is not probable cause to arrest. Further investigation is needed to develop probable cause of arrest”.
The New York Times article [2] states that this facial recognition technology that the Michigan tax-payer has paid millions of dollars for is known to be biased and that the vendors do “not formally measure the systems’ accuracy or bias”.
Finally, the original NPR article states that
> "Most of the time, people who are arrested using face recognition are not told face recognition was used to arrest them," said Jameson Spivack
It gets even crazier to think about when you realise that cities like Detroit are overwhelmingly Black. The police there are just not providing good value for the people who live there.
It isn't just facial recognition, license plate readers can have the same indefensibly Kafka-esque outcomes where no one is held accountable for verifying computer-generated "evidence". Systems like in the article make it so cheap for the government to make a mistake, since there are few consequences, that they simply accept mistakes as a cost of doing business.
Someone I know received vehicular fines from San Francisco on an almost weekly basis solely from license plate reader hits. The documentary evidence sent with the fines clearly showed her car had been misidentified but no one ever bothered to check. She was forced to fight each and every fine because they come with a presumption of guilt, but as soon as she cleared one they would send her a new one. The experience became extremely upsetting for her, the entire bureaucracy simply didn't care.
It took threats of legal action against the city for them to set a flag that apparently causes violations attributed to her car to be manually reviewed. The city itself claimed the system was only 80-90% accurate, but they didn't believe that to be a problem.
I agree that's bad, and license plate readers come with their own set of problems.
But being biased by the skin color of the driver is (AFAIK) not one of them. Which is exactly the problem with vision systems applied to humans, at least the ones we've seen deployed so far.
If a system discriminates against a specific population, that's very different from (indiscriminately) being unreliable.
That suddenly reminded me why I feel so privileged to not own a car, a distinct contrast from when I was a teenager and felt it was a rite of passage!
I had forgotten about the routine of fighting traffic tickets multiple times a year as a fact of life. Let alone fender benders. I had only been reveling in the lack of a frustrating commute.
Last decade I did get a car for 3 months, and the insurance company was so thrilled that I was "such a good driver" because of my "spotless record" for many years. Little do they know I just don't drive and perhaps have now less experience than others. Although tangentially, their risk matrix actually might be correct, if I can afford to live in dense desirable areas then maybe it is less likely that I would be going fast and getting into circumstances that pull from their insurance pool at larger amounts.
They probably thought "one of the largest companies in the world probably chauffeurs him down the highway in a bus anyway"
This is a big reason I wish it was easier to not have a car in the US. There's always the potential to get things like parking tickets, and you have to deal with license, insurance, parking permit, etc.
The volume of tickets issued is quite staggering, and each one is a huge annoyance for someone.
> How many people actually stick to the page and listen to that?
I just did. 3 minutes wasn't that bad and I wasn't somewhere where it would be a problem.
> Why do sites do this?
NPR is a radio network. I have seen that often they do transcribe their clips. I am not sure what the process they have for that looks like, but it seems this particular clip didn't get transcribed.
Edit: looks like they do have a transcription mentioned elsewhere in the thread. So seems like some kind of UI fail.
NPR does transcribe (many, most?) its audio stories, but usually there's a delay of a day or so – the published timestamp for this story is 5:06AM (ET) today.
edit: looks like there's a text version of the article. I'm assuming this is a CMS issue: there's an audio story and a "print story", but the former hadn't been linked to the latter: https://news.ycombinator.com/item?id=23628790
Most people are going to hear the story on the radio or in a podcast app / RSS feed. It’s useful to have the story indexed on a shareable web link where it can be played on different platforms without any setup. If I wanted to share a podcast episode with friends in a group chat, a link like this would be a good way to do it. Since this is more of a long-form text discussion forum I’d probably look for a text format before posting here.
Third, Robert’s arrest demonstrates why claims that face recognition isn’t dangerous are far-removed from reality. Law enforcement has claimed that face recognition technology is only used as an investigative lead and not as the sole basis for arrest. But once the technology falsely identified Robert, there was no real investigation.
I fear this is going to be the norm among police investigations.
> Federal studies have shown that facial-recognition systems misidentify Asian and black people up to 100 times more often than white people.
The idea behind inclusion is that this product would have never made it to production if the engineering teams, product team, executive team and board members represented the population. But enough representation so that there is a countering voice is even better.
Would have just been "this edge case is not an edge case at all, axe it."
Accurately addressing a market is the point of the corporation more than an illusion of meritocracy amongst the employees.
This is so incredibly common, it's embarrassing. I was on an expert panel about "AI and Machine Learning in Healthcare and Life Sciences" back in January, and I made it a point throughout my discussions to keep emphasizing the amount of bias inherent in our current systems, which ends up getting amplified and codified in machine learning systems. Worse yet, it ends up justifying the bias based on the false pretense that the systems built are objective and the data doesn't lie.
Afterward, a couple people asked me to put together a list of the examples I cited in my talk. I'll be adding this to my list of examples:
* A hospital AI algorithm discriminating against black people when providing additional healthcare outreach by amplifying racism already in the system. https://www.nature.com/articles/d41586-019-03228-6
* Misdiagnosing people of African decent with genomic variants misclassified as pathogenic due to most of our reference data coming from European/white males. https://www.nejm.org/doi/full/10.1056/NEJMsa1507092
* When AI criminal risk prediction software used by judges in deciding the severity of punishment for those convicted predicts a higher chance of future offence for a young, black first time offender than for an older white repeat felon. https://www.propublica.org/article/machine-bias-risk-assessm...
I agree 100% about how common it is. The industry also pays lip service about doing something about it. My last job was at a research institution and we had a data ethics czar, who's a very smart (Stats phd) guy and someone I consider a friend. A lot of his job was to go around the org and conferences talking about things like this.
While there's a lot of head nodding, nothing is ever actually addressed in day to day operations. Data scientists barely know what's going on when they throw things through TensorFlow. What matters is the outcome and the confusion matrix at the end.
I say this as someone who works in data and implements AI/ML platforms. Mr. Williams needs to find the biggest ambulance chasing lawyer and file civil suits not only the law enforcement agencies involved, but top down everyone at DataWorks from the president to the data scientist to the lowly engineer who put this in production.
These people have the power to ruin lives. They need to be made an example of and held accountable for the quality of their work.
>When AI criminal risk prediction software used by judges in deciding the severity of punishment for those convicted predicts a higher chance of future offence for a young, black first time offender than for an older white repeat felon.
>When Amazon's AI recruiting tool inadvertantly filtered out resumes from women
>When Google's hate speech detecting AI inadvertantly censored anyone who used vernacular referred to in this article as being "African American English
There's simply no indication that these aren't statistically valid priors. And we have mountains of scientific evidence to the contrary, but if dared post anything (cited, published literature) I'd be banned. This is all based on the unfounded conflation between equality of outcome and equality of opportunity, and the erasure of evidence of genes and culture playing a role in behavior and life outcomes.
The discussion about this tech revolves around accuracy and racism, but the real threat is in global unlimited surveillance. China is installing 200 million of facial recognition cameras right now to keep the population under control. It might be the death of human freedom as this technology spreads
Another reason that it's absolutely insane that the state demands to know where you sleep at night in a free society. These clowns were able to just show up at his house and kidnap him.
The practice of disclosing one's residence address to the state (for sale to data brokers[1] and accessible by stalkers and the like) when these kinds of abuses are happening is something that needs to stop. There's absolutely no reason that an ID should be gated on the state knowing your residence. It's none of their business. (It's not on a passport. Why is it on a driver's license?)
Perhaps we, as technologists, are going about this the wrong way. Maybe, instead of trying to reduce the false alarm rate to an arbitrarily low number, we instead develop CFAR (Constant false alarm rate) systems, so that users of the system know that they will get some false alarms, and develop procedures for responding appropriately. In that way, we could get the benefit of the technology, whilst also ensuring that the system as a whole (man and machine together) are designed to be robust and have appropriate checks and balances.
If you follow the link, you'll see that the computer report had this message right at the top in massive letters:
THIS DOCUMENT IS NOT A POSITIVE IDENTIFICATION. IT IS AN INVESTIGATIVE LEAD ONLY AND IS _NOT_ PROBABLE CAUSE TO ARREST. FURTHER INVESTIGATION IS NEEDED TO DEVELOP PROBABLE CAUSE TO ARREST.
I mean, what else could the technologists have done?
Well .. it looks like it isn't enough, and additional human-factors oriented requirements need to be documented - like deliberately engineering the system to produce a certain number of false alarms so the police always have to do additional work to discard them - as opposed to only sometimes.
We, as technologists, probably largely agree with this argument, but those in charge are all laymen: officers, Dept. leaders, politicians, even lawyers, can all be ignorant of that nuance, and will probably remain so unless walked through in detail, and then continuously badgered about it.
I don't think using the facial recognition is necessarily wrong to help identify probable suspects, but arresting someone based on a facial match algorithm is definitely going too far.
Of course really I blame the AI/ML hucksters for part of this mess who have sold us the idea of machines replacing rather than augmenting human decision making.
In a world where some police forces don't use polygraph lie detectors because they are deemed too inaccurate, it baffles me that people would make an arrest based on a facial recognition hit from poor quality data.
This seems similar to self-driving cars where people hold the computer to much higher standards than humans. I don't have solid proof, but I suspect that using facial recognition with a reasonable confidence threshold and reasonable source images is more accurate than eyewitness ID. If for no other reason than the threshold for a positive eyewitness ID is laughably bad.
The current best practice is to have a witness pick out the suspect from 6 photos. It should be immediately obvious that right off the bat there's a 17% chance of the witness randomly picking the "right" person. It's a terrible way to do things and it's no surprise that people are wrongly convicted again and again on eyewitness testimony.
I think it is very wrong. Faces are anything but unique. Having a particular face should not result in you being a suspect. Only once actual policing results in you becoming a suspect then this might be a low quality extra signal.
> Having a particular face should not result in you being a suspect. Only once actual policing results in you becoming a suspect then this might be a low quality extra signal.
Having a picture or just a description of the face is one of the most important pieces of information the police has in order to do actual policing. You can be arrested for just broadly matching the description if you happen to be in the vicinity.
Had the guy been convicted of anything just based on that evidence, this would be a scandal. As it is, a suspect is just a suspect and this kind of thing happens all the time, because humans are just as fallible. It's just not news when there's no AI involved.
Yeah, facial recognition can be useful in law enforcement, as long as it's used responsibly. There was a man who shot people at a newspaper where I lived, and when apprehended, he refused to identify himself, and apparently their fingerprint machine wasn't working, so they used facial recognition to identify him.
From the wiki article and the linked news articles, the police picked him up at the scene of the crime. He also had smoke grenades (used in the attack) when they found him.
> Authorities said he was not carrying identification at the time of his arrest and was not cooperating. … an issue with the fingerprint machine ultimately made it difficult to identify the suspect, … A source said officials used facial recognition technology to confirm his identity.
> Police, who arrived at the scene within a minute of the reported gunfire, apprehended a gunman found hiding under a desk in the newsroom, according to the top official in Anne Arundel County, where the attack occurred.
This doesn't really seem like an awesome use of facial recognition to me. He was already in custody after getting picked up at the crime scene. I doubt he would have been released if facial recognition didn't exist.
I don't think there is such a thing as responsible use of facial recognition technology by law enforcement.
The technology is certainly not robust enough to be trusted to work correctly at that level yet. Even if it was improved I think there is a huge moral issue with the police having the power to use it indiscriminately on the street.
A few things I just don't have the stomach for as an engineer, writing software that:
- impacts someones health
- impacts someones finances
- impacts someones freedoms
Call me weak, but I think about the "what ifs" a bit too much in those cases. What if my bug keeps them from selling their stock and they lose their savings? What if the wrong person is arrested, etc?
Why would anyone call you weak, that's principled and it's the correct attitude. It's the people who don't think about the consequences of the products they help build that are the problem.
I think that your prints, DNA, and so forth must be, in the interests of fairness, utterly erased from all systems in the case of false arrest. With some kind of enormous, ruinous financial penalty in place for the organizations for non-compliance, as well as automatic jail times for involved personnel. These things need teeth to happen.
any defence lawyer with more than 3 brain cells would have an absolute field day deconstructing a case brought solely on the basis of a facial recognition. What happened to the idea that police need to gather a variety of evidence confirming their suspicions before an arrest is issued. Even a state prosecutor wouldn't authorize a warrant based on such flimsy methods.
And then in some states employers are allowed to ask have you eve been arrested (never mind convicted of any crime) on employment application. Sure, keep putting people down. One day it might catch up with China's social scoring policies.
The difference is that is a known problem, but with ML, a large fraction of the population thinks it's infallible. Worse, its reported confidence for an individual face may be grossly overstated, since that is based on all the data it was trained on, rather than the particular subset you may be dealing with.
large fraction of the population and ML marketing both believe that.
I still think it insane. We have falling crime rates and we still arm ourselves as fast as we can. Humanity could live without face recognition and we wouldn't even suffer any penalties. Nope, people need to sell their evidently shitty ML work.
> The difference is that is a known problem, but with ML, a large fraction of the population thinks it's infallible.
I don't think anybody actually believes that.
I'm pretty sure the exact opposite is true: People expect AI to fail, because they see it fail all the time in their daily use of computers, for example in voice recognition.
> Worse, its reported confidence for an individual face may be grossly overstated, since that is based on all the data it was trained on, rather than the particular subset you may be dealing with.
At the end of the day, this is still human error. A human compared the faces and decided they looked alike enough to go ahead. The whole thing could've happened without AI, it's just that without AI, processing large volumes of data is infeasible.
The suspect said the picture looked nothing like him. When he was shown the picture, he picked up the picture, put it in front of his face, and said "I hope you don't think all black people look alike".
I see this all the time when working with execs. I have to continually remind even very smart people with STEM undergrad and even graduate degrees that a computer vision system cannot magically see things that are invisible to the human eye.
"the computer said so" is way stronger than you would think.
Not even close to the same thing. People aren't very reliable witnesses either, But they are pretty good at identifying people they actually know.
It's also poor practice to search a database using a photo or even DNA to go fishing for a suspect. A closest match will generally be found even if the actual perpetrator isn't in the database. I think on some level the authorities know this, which is why they dont seed the databases with their own photos and DNA.
It's like asking "is mass surveillance that different from targeted surveillance"?
Yes, of course it is. Orders of magnitude more people could be negatively and undeservedly affected this for no other reason than the fact that it's now cheap enough and easy enough to use by the authorities.
Just to give one example I came up with right now, in the future the police could stop you, take your picture and automatically have it go through its facial recognition database. Kind of like "stop and scan".
Or if the street cameras get powerful enough (and they will), they could take your picture automatically while driving and then stop you.
Think of it like a "TSA system for the roads". A lot more people will be "randomly picked" by these systems from the roads.
Nope. But It's certainly far more accurate than eyewitnesses. And will reduce the frequency of false positives. Compare this to "suspect is a 6' male approx 200lbs with a white shirt and blue jeans" and then having police frantically pick up everyone in the area that meets this description.
This is the story that gets attention though. Despite it representing an improvement in likely every potential metic you can measure.
The response is what is interesting to me. It triggers a 1984 reflex resulting in people attempting to reject a dramatic enchantment in law enforcement ostensibly because it is not perfect. Or because they believe it a threat to privacy. I think people who are rejecting it should dig deep into their assumptions and reasoning to examine why they are really opposed to technology like this.
> think people who are rejecting it should dig deep into their assumptions and reasoning to examine why they are really opposed to technology like this.
Because a false positive ruins lives? Is that not sufficient? This man’s arrest record is public and won’t disappear. Many employers won’t hire if you have an arrest record (regardless of conviction). His reputation is also permanently smeared. These records are permanently public and in fact some counties publish weekly arrest records on their websites and in newspapers (not that newspapers matter much anymore)
Someday this technology may be better and work more reliably. We’re not there yet. Right now it’s like the early days of voice recognition from the ‘90s.
What is a unique use case for facial recognition that cannot be abused and has no other alternative solution?
Even the "good" use cases like unlocking your phone have security problems because malicious people can use photos or videos of your face and you can't change your face like you would a breached username and password.
I've got to be honest: I'm getting the picture the police here aren't very competent. I know I know, POSIWID and maybe they're very competently aiming at the current outcome. But don't they just look like a bunch of idiots?
In this particular case, computerized facial recognition is not the problem.
Facial recognition produces potential matches. It's still up to humans to look at footage themselves and use their judgment as to whether it's actually the same person or not, as well as to judge whether other elements fit the suspect or not.
The problem here is 100% on the cop(s) who made that call for themselves, or intentionally ignored obvious differences. (Of course, without us seeing the actual images in question, it's hard to judge.)
There are plenty of dangers with facial recognition (like using it at scale, or to track people without accountability), but this one doesn't seem to be it.
> The problem here is 100% on the cop(s) who made that call for themselves
I disagree. There is plenty of blame on the cops who made that call for themselves, true.
But there doesn't have to be a single party who is at fault. The facial recognition software is badly flawed in this dimension. It's well established that the current technologies are racially biased. So there's at least some fault in the developer of that technology, and the purchasing officer at the police department, and a criminal justice system that allows it to be used that way.
Reducing a complex problem to a single at-fault person produces an analysis that will often let other issues continue to fester. Consider if the FAA always stopped the analysis of air-crashes at: "the pilot made an error, so we won't take any other corrective actions other than punishing the pilot". Air travel wouldn't nearly as safe as it is today.
While we should hold these officers responsible for their mistake (abolish QI so that these officers could be sued civilly for the wrongful arrest!), we should also fix the other parts of the system that are obviously broken.
The justification for depriving someone of their liberty lies solely with the arresting officer. They can base that on whatever they want, as long as they can later justify it to a court.
For example, you might have a trusted informant who could tell you who committed a local burglary, just this on its own could be legitimate grounds to make an arrest. The same informant might walk into a police station and tell the same information to someone else, for that officer, it might not be sufficient to justify an arrest.
> Even if this technology does become accurate (at the expense of people like me), I don’t want my daughters’ faces to be part of some government database.
Stop using Amazon Ring and similar doorbell products.
The pandemic has accelerated the use of no-touch surfaces specially at places like an airport that are more inclined to now use a face recognition security kiosk. What's not clear is the vetting process for these (albeit controversial) technologies. What if Google thinks person A is an offender but Amazon thinks otherwise. Can they be used as counter evidence? What is the gold standard for surveillance?
And now the poor guy has an arrest record. Which wouldn't be a problem in reasonable legislations, where it's nobody's business whether you've been arrested or not, as long as you've not been convicted.
But in the US, I've heard that it can make it harder to get a job.
I believe I'm starting to get a feel for how the school to prison pipeline may work.
Unlikely unless they were compelled by law or found something else to replace it, and I think it's the latter. Something about machine learning and such.
In alot of police departments around the world, the photo database used is the drivers license database.
There is clothing available that can confuse facial recognition systems. What would happen if, next time you go for your drivers license photo, you wore a T shirt designed to confuse facial recognition, for example like this one? https://www.redbubble.com/i/t-shirt/Anti-Surveillance-Clothi...
I would love to see police trying to take a crack at this from the other side of things. Instead of matching against a database, Set up a style GAN and come up with a mask of the original photo or video to isolate just the face and have the discriminator try to match the face. Then at the end you can see the generated face with a decent pose and more importantly look through the range of generated faces that result in a reasonable match to give a somewhat decent idea of how confident you should be about any identification.
While this case is bad enough, mistakes like this are not the biggest concern. Mistakenly arrested people are (hopefully) eventually released, even though they have to go through quite a bit of trouble.
The consequence that is much worse would be mass incarceration of certain groups, because the AI is too good at catching people who actually did something.
This second wave of mass incarceration will lead to even more single parent families and poor households, and will reinforce the current situation.
How does computerized facial recognition compare in terms of racial bias and accuracy to human-brain facial recognition? Police are not exactly perfect in either regard.
While also enabling finger-pointing, e.g. the police can say "We aren't racist or aren't at fault. The system is just faulty." while the engineers behind the facial recognition tech can say that they, "Were just doing their job. The police should've heeded their disclaimers, etc."
It's supposed to be a cornerstone of "innocent until proven guilty" legal systems that it is better to have 10 guilty people go free than to deprive a single innocent person of their freedom. It seems like the needle has been moving in the wrong direction on that. I'm not sure it that's just my impression on things, or if it's because there's more awareness with the internet/social networking of issues...
No mention of whether a judge signed a warrant for the arrest. In what world can cops just show up and arrest you on your front lawn based on their hunch?
If it's statistically proven to not work with black people then I think the only options are
1) Make it avoid black people, i.e. they aren't stored in the database and aren't processed when scanned.
2) Put a 5 year hiatus on commercial / public use.
Either of these things are more acceptable than too many false positives. #1 is really interesting to me as a thought experiment because it makes everyone think twice.
The issue is that it's hard to determine who is considered "black" and who is not since race is a made-up fiction constructed by white supremacists that is now so engrained in American society we find it difficult if not impossible to reflect on it. Are we using the Fitzpatrick skin color scale? Are we continuing the hypodescent rule (an algorithm, btw)?
Maybe we just outlaw face recognition in criminal justice entirely.
It definitely shouldn't be allowed as the main evidence - perhaps the answer to skin tone is to use the data and see where the algo starts to get too many false positives..
The largest - over arching idea - is to get everyone to think twice by making the majority think twice - If white people think 'its only for us!?' it'll make them really study the effects.. (I'm white.)
This technology will never be ready to use like this.
Similarly we shouldn’t collect vast databases of fingerprints or DNA and search them for every crime.
Why? Because error rates are unavoidable. There is some uncertainty, and in large enough numbers you will find false matches with perfect DNA matching.
We must keep our senses and use these technologies to help us rather than find the hidden bad guy.
well, I'm going to take something out of my chest. every time I shared a project here using machine learning people always gave me crap. saying my models were simplistic, or I did something wrong or the solution didn't work 100% of the time. well, I studied ML back in college. the basics, the algorithms that started all, linear regression, perceptron, adaline, knn, kmeans... and guess what? ML doesn't work 100% of the time. I always wanted to see how people would react when a car driven by ml hits something or when they based an important decision based on the classification of an nn. ML should be used along side human intelligence not by itself. you don't blindly trusts a black box.
The worst part is they use facial recognition which finds someone that looks like the suspect, and then they put the guy in a lineup and have him identified by the victim. Wtf?
I've been thinking this sort of event has become inevitable. Tech development and business models support extending the environments in which we collect images and analyze them. Confidence values lead to statistical guilt. I wrote about it here if interested: https://unintendedconsequenc.es/inevitable-surveillance/
> In Williams' case, police had asked the store security guard, who had not witnessed the robbery, to pick the suspect out of a photo lineup based on the footage, and the security guard selected Williams.
All that rigamaroll for $3800 worth of crap? They should just switch it up and start entrapping people like the FBI does. Then at least they would have perhaps one leg to stand on.
That and we might get some kind of judicial ruling that current incarnations of facial recognition software are racially based.
It would be a great result if a court declared that the use of racially biased facial recognition software is a violation of the 14th amendment violation, and enjoined PDs from using such software unless it can be demonstrated to be free of racial bias.
Will he? I thought it was pretty hard to win cases against the police. What does the evidence about the practicality of pursuing police for torts say here, and for that matter what kind of evidence should we be looking for?
One can file a civil case for false arrest. However, thanks to the qualified immunity cops enjoy coupled with the newness of the facial recognition technology, it will be very hard to prove that they obviously violated clearly established law. Maybe the cops will even point their fingers at the software.
Maybe he can go after the makers of the facial recognition software, but they can probably point their finger at the cops for using it wrong.
So, in any case, the guy will be left with a big legal bill.
Here is a part that I personally have to wrestle with:
> "They never even asked him any questions before arresting him. They never asked him if he had an alibi. They never asked if he had a red Cardinals hat. They never asked him where he was that day," said lawyer Phil Mayor with the ACLU of Michigan.
When I was fired by an automated system, no one asked if I had done something wrong. They asked me to leave. If they had just checked his alibi, he would have been cleared. But the machine said it was him, so case closed.
Not too long ago, I wrote a comment here about this [1]:
> The trouble is not that the AI can be wrong, it's that we will rely on its answers to make decisions.
> When the facial recognition software combines your facial expression and your name, while you are walking under the bridge late at night, in an unfamiliar neighborhood, and you are black; your terrorist score is at 52%. A police car is dispatched.
Most of us here can be excited about Facial Recognition technology but still know that it's not something to be deployed in the field. It's by no means ready. We might even consider the moral ethics before building it as a toy.
But that's not how it is being sold to law enforcement or other entities. It's _Reduce crime in your cities. Catch criminals in ways never thought possible. Catch terrorists before they blow up anything._ It is sold as an ultimate decision maker.
[1]:https://news.ycombinator.com/item?id=21339530
52% is little better than a coin flip. If you have a million individuals in your city, your confidence should be in the ballpark of 99.9999% (1 individual in 1 million). That has really been my concern with this, the software will report any facial match above 75% confidence. Apart from the fact that it appalling confidence, no cop will pay attention to that percentage; immediately arresting or killing the individual.
Software can kill. This software can kill 50% of black people.
Software can kill if we put blind trust in it and give it full control over the situation. But we shouldn't do that.
Even if it was correct 99% of the time, we need to recognize that software can make mistakes. It is a tool, and people need to be responsible enough to use it correctly. I think I agree with your general idea here, but to put all of the blame on software strikes me as an incomplete assessment. Technically the software isn't killing anyone, irresponsible users of it are.
18 replies →
> 99.9999%
Then the justice system would implode. Judicial policy is "software" too, and nobody holds the judiciary or police to that absurd level of excellence, even if we're talking about the death penalty.
3 replies →
I would hope that the software/ML engineers who wrote it knows about probability theory, and why the prior probability should be set at 0.0001% or so.
So that if we print 52% on the screen, that means we've already gathered like 30-bits of evidence (30 coin flips all coming up heads), at which point the suspicion would be real.
What a great comment. This encapsulates my concerns about the topic eloquently. The technology is not ready for use.
Just to be clear, parent is describing fictional software, not the system in the article. You seem to be conflating the two.
1 reply →
It would have to be even higher than that level of accuracy, because every person is going to be 'tested' multiple times.... if everyone's face is scanned 100 times a day, the number of false positives is going to be even higher.
1 reply →
It’s not at all obvious to me that the accuracy threshold should scale with city size. Some small town shouldn’t use a system that is 1000x less accurate.
2021: 50% of black people are now dead due to facial recognition. Trump suspends summary execution.
This shit right here. This is why I don't stop for the inventory control alarms at department store doorways if they go off. I know I've paid, and little sirens are just a nuisance at this point.
This is why I've never stopped for receipt checks, because it's my receipt, and I've paid. The security theatre is just bad comedy.
Just because the machine says I've done a no no, doesn't mean I can't come back and win a lawsuit later. It doesn't abdicate cops from their jobs. I have a winning complexion, so I'll never enjoy a false positive, but if I do, I'll make sure it bankrupts whatever startup incubator garbage decided to shill a replacement for real law enforcement.
Is this an attitude that is safe for people of all races, though?
Also, can everyone afford to pursue lawsuits?
11 replies →
So out of curiosity, you just roll out of Costco with a cart full of food and gear while the receipt checker tries to stop you?
2 replies →
Worse - an AI decision puts an obligation on the user to follow it. What do I mean? Well - imagine you are a cop, you get an auto flag to arrest someone and use your discretion to overide it. The person goes on to do something completely different; like they are flagged as a murderer but then go and kill someone DUI. You will be flayed, pilloried. So basically safety first, just do the arrest. The secret is that these systems should not be making calls in this kind of context because they just aren't going to be good enough, it's like cancer diagnosis - the oncologist should have the first say, the machine should be a safety net.
You see it everywhere with AI and other tools. We overly trust them. Even when doctors have a high confidence in their diagnosis, they accept wrong AI-recommended conclusion that contradicts it.
https://www.nature.com/articles/s41591-020-0942-0
Bit like with self driving cars - if it's not perfect we don't know how to integrate it with people
That's interesting. I can imagine in cases like this it's not necessarily that the doctor doubts their own diagnosis, but rather the AI is essentially offering to relieve them of responsibility for it either way.
It's like in human hierarchies - it's often not the person who is more likely to make the best decision who gets to decide, it's the one who is going to bear the consequences of being wrong.
> The trouble is not that the AI can be wrong
Exactly what I thought when I've read about this. It's not like humans are great at matching faces either. In fact machines have been better at facial recognition for over a decade now. I bet there are hundreds of people (of all races) in prison right now who are there simply because they were mis-identified by a human. Human memory, even in the absence of bias and prejudice, is pretty fallible.
There is a notion of "mixture of experts" in machine learning. It's when you have two or more models that are not, by themselves, sufficiently good to make a robust prediction, but that make different kinds of mistakes, and you use the consensus estimate. The resulting estimate will be better than any model in isolation. The same should be done here - AI should be merely a signal, it is not a replacement for detective work, and what's described in the article is just bad policing. AI has very little to do with that.
Can you add any details about how an automated system was used to fire you? I'm not familiar with systems like that.
We had a whole discussion about it here a couple years ago: https://news.ycombinator.com/item?id=17350645
1 reply →
Seconded! Fascinating situation.
1 reply →
The problem is not with the technology, but with how it's used. A medical test is also not 100% error-proof which is why a professional needs to interpret the results, sometimes conducting other tests or disregarding it completely.
A cop stopping someone that has a resemblance to a criminal for questioning seems like a good thing to me, as long as the cop knows that there's a reasonable chance it's the wrong guy.
>But that's not how it is being sold to law enforcement or other entities. It's _Reduce crime in your cities. Catch criminals in ways never thought possible. Catch terrorists before they blow up anything._ It is sold as an ultimate decision maker.
None of those selling points logically lead to the conclusion that it is the ultimate decisions maker.
This story is really alarming because as described, the police ran a face recognition tool based on a frame of grainy security footage and got a positive hit. Does this tool give any indication of a confidence value? Does it return a list (sorted by confidence) of possible suspects, or any other kind of feedback that would indicate even to a layperson how much uncertainty there is?
The issue of face recognition algorithms performing worse on dark faces is a major problem. But the other side of it is: would police be more hesitant to act on such fuzzy evidence if the top match appeared to be a middle-class Caucasian (i.e. someone who is more likely to take legal recourse)?
I think the NYT article has a little more detail: https://www.nytimes.com/2020/06/24/technology/facial-recogni...
Essentially, an employee of the facial recognition provider forwarded an "investigative lead" for the match they generated (which does have a score associated with it on the provider's side, but it's not clear if the score is clearly communicated to detectives as well), and the detectives then put the photo of this man into a "6 pack" photo line-up, from which a store employee then identified that man as being the suspect.
Everyone involved will probably point fingers at each other, because the provider for example put large heading on their communication that, "this is not probable cause for an arrest, this is only an investigative lead, etc.", while the detectives will say well we got a hit from a line-up, blame the witness, and the witness would probably say well the detectives showed me a line-up and he seemed like the right guy (or maybe as is often the case with line-ups, the detectives can exert a huge amount of bias/influence over witnesses).
EDIT: Just to be clear, none of this is to say that the process worked well or that I condone this. I think the data, the technology, the processes, and the level of understanding on the side of the police are all insufficient, and I do not support how this played out, but I think it is easy enough to provide at least some pseudo-justification at each step along the way.
That's interesting. In many ways, it's similar to the "traditional" process I went through when reporting a robbery to the NYPD 5+ years ago: they had software where they could search for mugshots of all previously convicted felons living in a x-mile radius of the crime scene, filtered by the physical characteristics I described. Whether the actual suspect's face was found by the software, it was ultimately too slow and clunky to paginate through hundreds of results.
Presumably, the facial recognition software would provide an additional filter/sort. But at least in my situation, I could actually see how big the total pool of potential matches and thus have a sense of uncertainty about false positives, even if I were completely ignorant about the impact of false negatives (i.e. what if my suspect didn't live within x-miles of the scene, or wasn't a known/convicted felon?)
So the caution re: face recognition software is how it may non-transparently add confidence to this already very imperfect filtering process.
(in my case, the suspect was eventually found because he had committed a number of robberies, including being clearly caught on camera, and in an area/pattern that was easy to narrow down where he operated)
> and the detectives then put the photo of this man into a "6 pack" photo line-up, from which a store employee then identified that man as being the suspect.
This is absurdly dangerously. The AI will find people who look like the suspect, that’s how the technology works. A lineup as evidence will almost guarantee a bad outcome, because of course the man looks like the suspect!
2 replies →
> Essentially, an employee of the facial recognition provider forwarded an "investigative lead" for the match they generated (which does have a score associated with it on the provider's side, but it's not clear if the score is clearly communicated to detectives as well)
This is the lead provided:
https://wfdd-live.s3.amazonaws.com/styles/story-full/s3/imag...
Note that it says in red and bold emphasis:
THIS DOCUMENT IS NOT A POSITIVE IDENTIFICATION. IT IS AN INVESTIGATIVE LEAD ONLY AND IS NOT PROBABLE CAUSE TO ARRREST. FURTHER INVESTIGATION IS NEEDED TO DEVELOP PROBABLE CAUSE TO ARREST.
5 replies →
This is why you should be scared of this tech. Computer assisted patsy finder. No need to find the right guy when the ai will happily cough up 20 people nearby who kinda sorta look like the perp enough to stuff them into a lineup in front of a confused and highly fallible witness.
1 reply →
I'm becoming increasingly frustrated with the difficulty in accessing primary source material. Why don't any of these outlets post the surveillance video and let us decide for ourselves how much of a resemblance there is.
14 replies →
I can see why you'd only get 6 guys together for a physical "6 pack" line-up.
But for a photo lineup I can't imagine why you don't have least 25 photos to pick from.
2 replies →
> the detectives then put the photo of this man into a "6 pack" photo line-up, from which a store employee then identified that man
This is not correct. The "6-pack" was shown to a security firm's employee, who had viewed the store camera's tape.
"In this case, however, according to the Detroit police report, investigators simply included Mr. Williams’s picture in a “6-pack photo lineup” they created and showed to Ms. Johnston, Shinola’s loss-prevention contractor, and she identified him." [1]
[1] ibid.
Just a tip in case it happens to anyone - Never, ever agree to be in a lineup.
It wasn't just that the employee picked the man out of 6 pack; the employee they interviewed wasn't even a witness to the crime in the first place.
>into a "6 pack" photo line-up
How did the people in the 6 pack photo line-up match up against the facial recognition? Were they likely matches?
3 replies →
> Does this tool give any indication of a confidence value?
Yes.
> Does it return a list (sorted by confidence) of possible suspects,
Yes.
> ... or any other kind of feedback that would indicate even to a layperson how much uncertainty there is?
Yes it does. It also states in large print heading “THIS DOCUMENT IS NOT A POSITIVE IDENTIFICATION IT IS AN INVESTIGATIVE LEAD AND IS NOT PROBABLE CAUSE TO ARREST”.
You can see a picture of this in the ACLU article.
The police bungled this badly by setting up a fake photo lineup with the loss prevention clerk who submitted the report (who had only ever seen the same footage they had).
However, tools that are rife for misuse do not get a pass because they include a bold disclaimer. If the tool/process can not prevent misuse, the tool/process is broken and possibly dangerous.
That said, we have little data on how often the tool results in catching dangerous criminals versus how often it misidentifies innocent people. We have little data on if those innocent people tend to skew toward a particular demographic.
But I have a fair suspicion that dragnet techniques like this unfortunately can be both effective and also problematic.
I think the software would be potentially less problematic if the victim/witness were given access, and (ostensibly) could see the pool of matches and how much/little the top likely match differed from the less confident matches.
> The police bungled this badly by setting up a fake photo lineup...*
FWIW, this process is similar to traditional police lineups. The witness is shown 4-6 people – one who is the actual suspect, and several that vaguely match a description of the suspect. When I was asked to identify a suspect in my robbery, the lineup included an assistant attorney who would later end up prosecuting the case. The police had to go out and find tall slight-skinned men to round out the lineup.
> ... with the loss prevention clerk who submitted the report (who had only ever seen the same footage they had).
Yeah, I would hope that this is not standard process. The lineup process is already imperfect and flawed as it is even with a witness who at least saw the crime first-hand.
Intresting and related, a team made a neat "face depixelizer" that takes a pixelated image and uses machine learning to generate a face that should match the pixelated image.
What's hilarious is that it makes faces that look nothing like the original high-resolution images.
https://twitter.com/Chicken3gg/status/1274314622447820801
Interesting... Neat... Hilarious... In light of the submission and the comment you're responding to, these are not the words I would choose.
I think there's genuine cause for concern here, especially if technologies like these are candidates for inclusion in any real law enforcement decision-making.
That should be called a face generator, not a depixelizer.
1 reply →
What's sad is that a tech entrepreneur will definitely add that feature and sell it to law enforcement agencies that believe in CSI magic: https://www.youtube.com/watch?v=Vxq9yj2pVWk
3 replies →
Ironically, if the police had used and followed the face depixelizer then we may not have had the false arrest of a black man - not because of accuracy but because it doesn't produce many black faces
I wonder if this is trained on the same, or similar, datasets.
1 reply →
People are not good at understanding uncertainty and its implications, even if you put it front and center. I used to work in renewable energy consulting and I was shocked by how aggressively uncertainty estimates are ignored by those whose goals they threaten.
In this case, it's incumbent on the software vendors to ensure that less-than-certain results aren't even shown to the user. American police can't generally be trusted to understand nuance and/or do the right thing.
I blame TV shows like CSI and all the other crap out there that make pixelated images look like something you could "Zoom" into or something the computer can still understand even if the eye does not. Because of this, non tech people do not really understand that pixelated images have LOST information. Add that to the racial situation in the U.S. and the the inaccuracy of the tool for black people. Wow, this can lead to some really troublesome results
I lose hours every day just yelling "enhance" at my computer screen. Hasn't worked yet, but any day now...
1 reply →
> But the other side of it is: would police be more hesitant to act on such fuzzy evidence if the top match appeared to be a middle-class Caucasian (i.e. someone who is more likely to take legal recourse)?
Honest question: does race predict legal recourse when decoupled from socioeconomic status, or is this an assumption?
Race and socioeconomic status are deeply intertwined. Or to be more blunt - US society has kept black people poorer. To treat them as independent variables is to ignore the whole history of race in the US.
1 reply →
>Honest question: does race predict legal recourse when decoupled from socioeconomic status, or is this an assumption?
I think the issue is that regardless of the answer, it isn't decoupled in real world scenarios.
I think the solution isn't dependent upon race either. It is to ensure everyone have access to legal recourse regardless of socioeconomic status. This would have the side effect of benefiting races correlated with lower socioeconomic status more.
2 replies →
Middle class black people often get harassed by police, and there is a long history of far steeper sentences for convictions for drugs used more by the black population (crack) than that used more by the white population (cocaine).
So unequal treatment based on race has quite literally been a feature of the US justice system, independent of socioeconomic status.
2 replies →
> The issue of face recognition algorithms performing worse on dark faces is a major problem.
This needs to be coupled with the truth that people (police) without diverse racial exposure are terrible at identifying people outside of their ethnicity. In the photo/text article they show the top of the "Investigative Lead Report" as an image. You mean to say that every cop who saw the two images side by side did not stop and say "hey, these are not the same person!" They did not, and that's because their own brains' could not see the difference.
This is a major reason police forces need to be ethnically diverse. Just that enables those members of the force who never grew up or spent time outside their ethnicity can learn to tell a diverse range of similar but different people outside their ethnicity apart.
It wouldn't make it into the newspapers, so it doesn't matter.
This is a classic example of the false positive rate fallacy.
Let's say that there are a million people, and the police have photos of 100,000 of them. A crime is committed, and they pull the surveillance of it, and match against their database. They have a funky image matching system that has a false positive rate of 1 in 100,000 people, which is way more accurate than I think facial recognition systems are right now, but let's just roll with it. Of course, on average, this system will produce one positive hit per search. So, the police roll up to that person's home and arrest them.
Then, in court, they get to argue that their system has a 1 in 100,000 false positive rate, so there is a chance of 1 in 100,000 that this person is innocent.
Wrong!
There are ten people in the population of 1 million that the software would comfortably produce a positive hit for. They can't all be the culprit. The chance isn't 1 in 100,000 that the person is innocent - it is in fact at least 9 out of 10 that they are innocent. This person just happens to be the one person out of the ten that would match that had the bad luck to be stored in the police database. Nothing more.
See also: Privileging the hypothesis.
If I'm searching for a murderer in a town of 1000, it takes about 10 independent bits of evidence to get the right one. And when I charge someone, I must already have the vast majority of that evidence. To say "oh well we don't know that it wasn't Mr. or Mrs. Doe, let's bring them in" is itself a breach of the Does' rights. I'm ignoring 9 of the 10 bits of evidence!
Using a low-accuracy facial recognition system and a low-accountability lineup procedure to elevate some random man who did nothing wrong from presumed-innocent to 1-in-6 to prime suspect, without having the necessary amount of evidence, is committing the exact same error and is nearly as egregious as pulling a random civilian out of a hat and charging them.
There's a good book called "The Drunkards Walk", that describes a woman who was jailed after having 2 children die from SIDS. They argued that the odds of this happening is 1 in a million (or something like that), so probably the woman is a baby killer. The prosecution had statisticians argue this. The woman was found guilty.
She later won on appeal in part because the defense showed that the testimony and argument of the original statisticians were wrong.
This stuff is so easy to get wrong. A little knowledge of statistics can be dangerous.
And even if the original stats were right. A 1 in a million event happens to about 100 people per day in the US.
2 replies →
In the legal world it apparently goes by the moniker "Prosecutor's Fallacy":
https://en.wikipedia.org/wiki/Prosecutor%27s_fallacy
See also: the paper "Why Most Published Research Findings Are False". Previously on HN: https://news.ycombinator.com/item?id=1825007
Definitely they should have everyone's 3d image in the system. DNA too.
He wasn't arrested until the shop owner had also "identified" him. The cops used a single frame of grainy video to pull his driver's license photo, and then put that photo in a lineup and showed the store clerk.
The store clerk (who hadn't witnessed the crime and was going off the same frame of video fed into the facial recognition software) said the driver's license photo was a match.
There are several problems with the conduct of the police in this story but IMHO the use of facial recognition is not the most egregious.
The story is the same one that all anti-surveillance, anti-police militarization, pro-privacy, and anti-authoritarian people foretell. Good technology will be used enable, amplify, and justify civil rights abuses by authority figures from your local beat cop, to a faceless corporation, a milquetoast public servant, or the president of the United States.
Our institutions and systems (and maybe humans in general) are not robust enough to cleanly handle these powers, and we are making the same mistake over and over and over again.
Correct, and this has been the story with every piece of technology or tool we've ever given to police. We give them body cameras and they're turned off or used to create FPS-style snuff films of gunned down citizens. Give them rubber bullets and they're aimed at protesters eyeballs. Give them tasers and they're used as an excuse to shoot someone when the suspect "resists." Give them flashbangs and they'll throw them into an infant's crib. Give them mace and it's used out of car windows to punish journalists for standing on the sidewalks.
The mistake is to treat any police department as a good-faith participant in the goal of reducing police violence. Any tool you give them will be used to brutalize. The only solution is to give them less.
It is not clear to me that the person who identified him was shop owner or clerk. From the nyt article: https://www.nytimes.com/2020/06/24/technology/facial-recogni...
"The Shinola shoplifting occurred in October 2018. Katherine Johnston, an investigator at Mackinac Partners, a loss prevention firm, reviewed the store’s surveillance video and sent a copy to the Detroit police"
"In this case, however, according to the Detroit police report, investigators simply included Mr. Williams’s picture in a “6-pack photo lineup” they created and showed to Ms. Johnston, Shinola’s loss-prevention contractor, and she identified him. (Ms. Johnston declined to comment.)"
I think you're correct that the person was not an owner or clerk. IMHO the salient point is that the person was not any sort of eyewitness but merely comparing the same grainy photo as the algorithm.
1 reply →
Yes, this is a story of police misconduct. The regulation of facial recognition that is required is regulation against police/authority stupidity. The FR system aids in throwing away misses, leaving investigative leads. But if a criminal is not in the FR database to begin with, any results of the FR are wastes of time.
> "I picked it up and held it to my face and told him, 'I hope you don't think all Black people look alike,' " Williams said.
I'm white. I grew up around a sea of white faces. Often when watching a movie filled with a cast of non-white faces, I will have trouble distinguishing one actor from another, especially if they are dressed similarly. This sometimes happens in movies with faces similar to the kinds I grew up surrounded by, but less so.
So unfortunately, yes, I probably do have more trouble distinguishing one black face from another vs one white face from another.
This is known as the cross-race effect and it's only something I became aware of in the last 5-10 years.
Add to that the fallibility of human memory, and I can't believe we still even use line ups. Are there any studies about how often line ups identify the wrong person?
https://en.wikipedia.org/wiki/Cross-race_effect
I lived in South Africa for a while and heard many times, with various degrees of irony, "you white people all look the same" from black South Africans. So yeah it's definitely a cross-racial recognition problem, and it's probably also a problem with distinguishing between members of visible minorities using traits beyond the most noticable othering characteristic.
There is just so much wrong with this story. For starters:
The shoplifting incident occurred in October 2018 but it wasn’t until March 2019 that the police uploaded the security camera images to the state image-recognition system but the police still waited until the following January to arrest Williams. Unless there was something special about that date in October, there is no way for anyone to remember what they might have been doing on a particular day 15 months previously. Though, as it turns out, the NPR report states that the police did not even try to ascertain whether or not he had an alibi.
Also, after 15 months, there is virtually no chance that any eye-witness (such as the security guard who picked Williams out of a line-up) would be able to recall what the suspect looked like with any degree of certainty or accuracy.
This WUSF article [1] includes a photo of the actual “Investigative Lead Report” and the original image is far too dark for a anyone (human or algorithm) to recognise the person. It’s possible that the original is better quality and better detail can be discerned by applying image-processing filters – but it still looks like a very noisy source.
That same “Investigative Lead Report” also clearly states that “This document is not a positive identification … and is not probable cause to arrest. Further investigation is needed to develop probable cause of arrest”.
The New York Times article [2] states that this facial recognition technology that the Michigan tax-payer has paid millions of dollars for is known to be biased and that the vendors do “not formally measure the systems’ accuracy or bias”.
Finally, the original NPR article states that
> "Most of the time, people who are arrested using face recognition are not told face recognition was used to arrest them," said Jameson Spivack
[1] https://www.wusf.org/the-computer-got-it-wrong-how-facial-re...
[2] https://www.nytimes.com/2020/06/24/technology/facial-recogni...
Really seems like most police departments in our country are incompetent, negligent, ineffective and systemically racist.
Many of these cops are earning $200k plus annually! Our law enforcement system is ridiculous and needs an overhaul.
It gets even crazier to think about when you realise that cities like Detroit are overwhelmingly Black. The police there are just not providing good value for the people who live there.
Brookings had a great post about this the other day: https://www.brookings.edu/blog/how-we-rise/2020/06/11/to-add...
It isn't just facial recognition, license plate readers can have the same indefensibly Kafka-esque outcomes where no one is held accountable for verifying computer-generated "evidence". Systems like in the article make it so cheap for the government to make a mistake, since there are few consequences, that they simply accept mistakes as a cost of doing business.
Someone I know received vehicular fines from San Francisco on an almost weekly basis solely from license plate reader hits. The documentary evidence sent with the fines clearly showed her car had been misidentified but no one ever bothered to check. She was forced to fight each and every fine because they come with a presumption of guilt, but as soon as she cleared one they would send her a new one. The experience became extremely upsetting for her, the entire bureaucracy simply didn't care.
It took threats of legal action against the city for them to set a flag that apparently causes violations attributed to her car to be manually reviewed. The city itself claimed the system was only 80-90% accurate, but they didn't believe that to be a problem.
I agree that's bad, and license plate readers come with their own set of problems.
But being biased by the skin color of the driver is (AFAIK) not one of them. Which is exactly the problem with vision systems applied to humans, at least the ones we've seen deployed so far.
If a system discriminates against a specific population, that's very different from (indiscriminately) being unreliable.
That suddenly reminded me why I feel so privileged to not own a car, a distinct contrast from when I was a teenager and felt it was a rite of passage!
I had forgotten about the routine of fighting traffic tickets multiple times a year as a fact of life. Let alone fender benders. I had only been reveling in the lack of a frustrating commute.
Last decade I did get a car for 3 months, and the insurance company was so thrilled that I was "such a good driver" because of my "spotless record" for many years. Little do they know I just don't drive and perhaps have now less experience than others. Although tangentially, their risk matrix actually might be correct, if I can afford to live in dense desirable areas then maybe it is less likely that I would be going fast and getting into circumstances that pull from their insurance pool at larger amounts.
They probably thought "one of the largest companies in the world probably chauffeurs him down the highway in a bus anyway"
This is a big reason I wish it was easier to not have a car in the US. There's always the potential to get things like parking tickets, and you have to deal with license, insurance, parking permit, etc.
The volume of tickets issued is quite staggering, and each one is a huge annoyance for someone.
Since the NPR is a 3 minute listen without a transcript, here's the ACLU's text/image article: https://www.aclu.org/news/privacy-technology/wrongfully-arre...
And here's a 1st-person account from the arrested man: https://www.washingtonpost.com/opinions/2020/06/24/i-was-wro...
The mods can change this link to https://www.npr.org/2020/06/24/882683463/the-computer-got-it...
The linked story is audio only and is associated with the Morning Edition broadcast, but the full story appears under our Special Series section.
(I work for NPR)
Ok, changed from https://www.npr.org/2020/06/24/882678392/man-says-he-was-fal.... Thanks!
As soon as I saw it was audio only, i left the site. Why do sites do this? How many people actually stick to the page and listen to that?
> How many people actually stick to the page and listen to that?
I just did. 3 minutes wasn't that bad and I wasn't somewhere where it would be a problem.
> Why do sites do this?
NPR is a radio network. I have seen that often they do transcribe their clips. I am not sure what the process they have for that looks like, but it seems this particular clip didn't get transcribed.
Edit: looks like they do have a transcription mentioned elsewhere in the thread. So seems like some kind of UI fail.
NPR does transcribe (many, most?) its audio stories, but usually there's a delay of a day or so – the published timestamp for this story is 5:06AM (ET) today.
edit: looks like there's a text version of the article. I'm assuming this is a CMS issue: there's an audio story and a "print story", but the former hadn't been linked to the latter: https://news.ycombinator.com/item?id=23628790
1 reply →
Well, if anyone were going to do it, you'd think no one would be surprised about it being the "National Public Radio"
6 replies →
Most people are going to hear the story on the radio or in a podcast app / RSS feed. It’s useful to have the story indexed on a shareable web link where it can be played on different platforms without any setup. If I wanted to share a podcast episode with friends in a group chat, a link like this would be a good way to do it. Since this is more of a long-form text discussion forum I’d probably look for a text format before posting here.
Why do radio sites post audio?
NPR's text-only article served to me:
https://text.npr.org/s.php?sId=882683463
From ACLU article:
Third, Robert’s arrest demonstrates why claims that face recognition isn’t dangerous are far-removed from reality. Law enforcement has claimed that face recognition technology is only used as an investigative lead and not as the sole basis for arrest. But once the technology falsely identified Robert, there was no real investigation.
I fear this is going to be the norm among police investigations.
> Federal studies have shown that facial-recognition systems misidentify Asian and black people up to 100 times more often than white people.
The idea behind inclusion is that this product would have never made it to production if the engineering teams, product team, executive team and board members represented the population. But enough representation so that there is a countering voice is even better.
Would have just been "this edge case is not an edge case at all, axe it."
Accurately addressing a market is the point of the corporation more than an illusion of meritocracy amongst the employees.
This is so incredibly common, it's embarrassing. I was on an expert panel about "AI and Machine Learning in Healthcare and Life Sciences" back in January, and I made it a point throughout my discussions to keep emphasizing the amount of bias inherent in our current systems, which ends up getting amplified and codified in machine learning systems. Worse yet, it ends up justifying the bias based on the false pretense that the systems built are objective and the data doesn't lie.
Afterward, a couple people asked me to put together a list of the examples I cited in my talk. I'll be adding this to my list of examples:
* A hospital AI algorithm discriminating against black people when providing additional healthcare outreach by amplifying racism already in the system. https://www.nature.com/articles/d41586-019-03228-6
* Misdiagnosing people of African decent with genomic variants misclassified as pathogenic due to most of our reference data coming from European/white males. https://www.nejm.org/doi/full/10.1056/NEJMsa1507092
* The dangers of ML in diagnosing Melanoma exacerbating healthcare disparities for darker skinned people. https://jamanetwork.com/journals/jamadermatology/article-abs...
And some other relevant, but not healthcare examples as well:
* When Google's hate speech detecting AI inadvertantly censored anyone who used vernacular referred to in this article as being "African American English". https://fortune.com/2019/08/16/google-jigsaw-perspective-rac...
* When Amazon's AI recruiting tool inadvertantly filtered out resumes from women. https://www.reuters.com/article/us-amazon-com-jobs-automatio...
* When AI criminal risk prediction software used by judges in deciding the severity of punishment for those convicted predicts a higher chance of future offence for a young, black first time offender than for an older white repeat felon. https://www.propublica.org/article/machine-bias-risk-assessm...
And here's some good news though:
* A hospital used AI to enable care and cut costs (though the reporting seems to over simplify and gloss over enough to make the actual analysis of the results a little suspect). https://www.healthcareitnews.com/news/flagler-hospital-uses-...
I agree 100% about how common it is. The industry also pays lip service about doing something about it. My last job was at a research institution and we had a data ethics czar, who's a very smart (Stats phd) guy and someone I consider a friend. A lot of his job was to go around the org and conferences talking about things like this.
While there's a lot of head nodding, nothing is ever actually addressed in day to day operations. Data scientists barely know what's going on when they throw things through TensorFlow. What matters is the outcome and the confusion matrix at the end.
I say this as someone who works in data and implements AI/ML platforms. Mr. Williams needs to find the biggest ambulance chasing lawyer and file civil suits not only the law enforcement agencies involved, but top down everyone at DataWorks from the president to the data scientist to the lowly engineer who put this in production.
These people have the power to ruin lives. They need to be made an example of and held accountable for the quality of their work.
1 reply →
>When AI criminal risk prediction software used by judges in deciding the severity of punishment for those convicted predicts a higher chance of future offence for a young, black first time offender than for an older white repeat felon.
>When Amazon's AI recruiting tool inadvertantly filtered out resumes from women
>When Google's hate speech detecting AI inadvertantly censored anyone who used vernacular referred to in this article as being "African American English
There's simply no indication that these aren't statistically valid priors. And we have mountains of scientific evidence to the contrary, but if dared post anything (cited, published literature) I'd be banned. This is all based on the unfounded conflation between equality of outcome and equality of opportunity, and the erasure of evidence of genes and culture playing a role in behavior and life outcomes.
This is bad science.
6 replies →
The discussion about this tech revolves around accuracy and racism, but the real threat is in global unlimited surveillance. China is installing 200 million of facial recognition cameras right now to keep the population under control. It might be the death of human freedom as this technology spreads
Edit: one source says it is 400 million new cameras: https://www.cbc.ca/passionateeye/m_features/in-xinjiang-chin...
Reminds me of this-
Facial recognition technology flagged 26 California lawmakers as criminals. (August 2019)
https://www.mercurynews.com/2019/08/14/facial-recognition-te...
Another reason that it's absolutely insane that the state demands to know where you sleep at night in a free society. These clowns were able to just show up at his house and kidnap him.
The practice of disclosing one's residence address to the state (for sale to data brokers[1] and accessible by stalkers and the like) when these kinds of abuses are happening is something that needs to stop. There's absolutely no reason that an ID should be gated on the state knowing your residence. It's none of their business. (It's not on a passport. Why is it on a driver's license?)
[1]: https://www.newsweek.com/dmv-drivers-license-data-database-i...
Perhaps we, as technologists, are going about this the wrong way. Maybe, instead of trying to reduce the false alarm rate to an arbitrarily low number, we instead develop CFAR (Constant false alarm rate) systems, so that users of the system know that they will get some false alarms, and develop procedures for responding appropriately. In that way, we could get the benefit of the technology, whilst also ensuring that the system as a whole (man and machine together) are designed to be robust and have appropriate checks and balances.
If you follow the link, you'll see that the computer report had this message right at the top in massive letters:
THIS DOCUMENT IS NOT A POSITIVE IDENTIFICATION. IT IS AN INVESTIGATIVE LEAD ONLY AND IS _NOT_ PROBABLE CAUSE TO ARREST. FURTHER INVESTIGATION IS NEEDED TO DEVELOP PROBABLE CAUSE TO ARREST.
I mean, what else could the technologists have done?
Well .. it looks like it isn't enough, and additional human-factors oriented requirements need to be documented - like deliberately engineering the system to produce a certain number of false alarms so the police always have to do additional work to discard them - as opposed to only sometimes.
I wonder if the problem here is just the way traditional policing works, not even the technology.
1 reply →
We, as technologists, probably largely agree with this argument, but those in charge are all laymen: officers, Dept. leaders, politicians, even lawyers, can all be ignorant of that nuance, and will probably remain so unless walked through in detail, and then continuously badgered about it.
I don't think using the facial recognition is necessarily wrong to help identify probable suspects, but arresting someone based on a facial match algorithm is definitely going too far.
Of course really I blame the AI/ML hucksters for part of this mess who have sold us the idea of machines replacing rather than augmenting human decision making.
In a world where some police forces don't use polygraph lie detectors because they are deemed too inaccurate, it baffles me that people would make an arrest based on a facial recognition hit from poor quality data.
But no, its AI, its magical and it must be right.
This seems similar to self-driving cars where people hold the computer to much higher standards than humans. I don't have solid proof, but I suspect that using facial recognition with a reasonable confidence threshold and reasonable source images is more accurate than eyewitness ID. If for no other reason than the threshold for a positive eyewitness ID is laughably bad.
The current best practice is to have a witness pick out the suspect from 6 photos. It should be immediately obvious that right off the bat there's a 17% chance of the witness randomly picking the "right" person. It's a terrible way to do things and it's no surprise that people are wrongly convicted again and again on eyewitness testimony.
1 reply →
I think it is very wrong. Faces are anything but unique. Having a particular face should not result in you being a suspect. Only once actual policing results in you becoming a suspect then this might be a low quality extra signal.
> Having a particular face should not result in you being a suspect. Only once actual policing results in you becoming a suspect then this might be a low quality extra signal.
Having a picture or just a description of the face is one of the most important pieces of information the police has in order to do actual policing. You can be arrested for just broadly matching the description if you happen to be in the vicinity.
Had the guy been convicted of anything just based on that evidence, this would be a scandal. As it is, a suspect is just a suspect and this kind of thing happens all the time, because humans are just as fallible. It's just not news when there's no AI involved.
2 replies →
Those hucksters should be worried about the Supreme Court swatting away their business model, because that's where I see this headed.
I don't think they'll worry about that. Even if that did happen there are foreign markets who would still invest in this.
Yeah, facial recognition can be useful in law enforcement, as long as it's used responsibly. There was a man who shot people at a newspaper where I lived, and when apprehended, he refused to identify himself, and apparently their fingerprint machine wasn't working, so they used facial recognition to identify him.
https://en.wikipedia.org/wiki/Capital_Gazette_shooting
From the wiki article and the linked news articles, the police picked him up at the scene of the crime. He also had smoke grenades (used in the attack) when they found him.
> Authorities said he was not carrying identification at the time of his arrest and was not cooperating. … an issue with the fingerprint machine ultimately made it difficult to identify the suspect, … A source said officials used facial recognition technology to confirm his identity.
https://en.wikipedia.org/wiki/Capital_Gazette_shooting#Suspe...
> Police, who arrived at the scene within a minute of the reported gunfire, apprehended a gunman found hiding under a desk in the newsroom, according to the top official in Anne Arundel County, where the attack occurred.
https://www.washingtonpost.com/local/public-safety/heavy-pol...
This doesn't really seem like an awesome use of facial recognition to me. He was already in custody after getting picked up at the crime scene. I doubt he would have been released if facial recognition didn't exist.
> as long as it’s used responsibly
At what point can we decide that people in positions of power are not and will not ever be responsible enough to handle this technology?
Surely as a society we shouldn’t continue to naively assume that police are “responsible” like we’ve assumed in the past?
2 replies →
I don't think there is such a thing as responsible use of facial recognition technology by law enforcement.
The technology is certainly not robust enough to be trusted to work correctly at that level yet. Even if it was improved I think there is a huge moral issue with the police having the power to use it indiscriminately on the street.
A few things I just don't have the stomach for as an engineer, writing software that: - impacts someones health - impacts someones finances - impacts someones freedoms
Call me weak, but I think about the "what ifs" a bit too much in those cases. What if my bug keeps them from selling their stock and they lose their savings? What if the wrong person is arrested, etc?
Why would anyone call you weak, that's principled and it's the correct attitude. It's the people who don't think about the consequences of the products they help build that are the problem.
Thank you for that really helpful reply. I re-framed my thoughts based on it.
I think that your prints, DNA, and so forth must be, in the interests of fairness, utterly erased from all systems in the case of false arrest. With some kind of enormous, ruinous financial penalty in place for the organizations for non-compliance, as well as automatic jail times for involved personnel. These things need teeth to happen.
any defence lawyer with more than 3 brain cells would have an absolute field day deconstructing a case brought solely on the basis of a facial recognition. What happened to the idea that police need to gather a variety of evidence confirming their suspicions before an arrest is issued. Even a state prosecutor wouldn't authorize a warrant based on such flimsy methods.
True but the defendant is still financially, and in many cases professionally, ruined.
Getting a layer that would advise beyond pleading guilty for a reduced sentence is not a default option.
The company that developed this software is Dataworks plus, according to the article. Name and shame.
And then in some states employers are allowed to ask have you eve been arrested (never mind convicted of any crime) on employment application. Sure, keep putting people down. One day it might catch up with China's social scoring policies.
Is that different from somebody getting arrested based on mistaken eyewitness.
The difference is that is a known problem, but with ML, a large fraction of the population thinks it's infallible. Worse, its reported confidence for an individual face may be grossly overstated, since that is based on all the data it was trained on, rather than the particular subset you may be dealing with.
large fraction of the population and ML marketing both believe that.
I still think it insane. We have falling crime rates and we still arm ourselves as fast as we can. Humanity could live without face recognition and we wouldn't even suffer any penalties. Nope, people need to sell their evidently shitty ML work.
2 replies →
> The difference is that is a known problem, but with ML, a large fraction of the population thinks it's infallible.
I don't think anybody actually believes that.
I'm pretty sure the exact opposite is true: People expect AI to fail, because they see it fail all the time in their daily use of computers, for example in voice recognition.
> Worse, its reported confidence for an individual face may be grossly overstated, since that is based on all the data it was trained on, rather than the particular subset you may be dealing with.
At the end of the day, this is still human error. A human compared the faces and decided they looked alike enough to go ahead. The whole thing could've happened without AI, it's just that without AI, processing large volumes of data is infeasible.
1 reply →
The suspect said the picture looked nothing like him. When he was shown the picture, he picked up the picture, put it in front of his face, and said "I hope you don't think all black people look alike".
I see this all the time when working with execs. I have to continually remind even very smart people with STEM undergrad and even graduate degrees that a computer vision system cannot magically see things that are invisible to the human eye.
"the computer said so" is way stronger than you would think.
Not even close to the same thing. People aren't very reliable witnesses either, But they are pretty good at identifying people they actually know.
It's also poor practice to search a database using a photo or even DNA to go fishing for a suspect. A closest match will generally be found even if the actual perpetrator isn't in the database. I think on some level the authorities know this, which is why they dont seed the databases with their own photos and DNA.
Yes.
A computer can make a mistake across literally any person who has a publicly available photo (which is almost everyone).
Also, the facial recognition technologies are provably extremely racially biased.
It's like asking "is mass surveillance that different from targeted surveillance"?
Yes, of course it is. Orders of magnitude more people could be negatively and undeservedly affected this for no other reason than the fact that it's now cheap enough and easy enough to use by the authorities.
Just to give one example I came up with right now, in the future the police could stop you, take your picture and automatically have it go through its facial recognition database. Kind of like "stop and scan".
Or if the street cameras get powerful enough (and they will), they could take your picture automatically while driving and then stop you.
Think of it like a "TSA system for the roads". A lot more people will be "randomly picked" by these systems from the roads.
Nope. But It's certainly far more accurate than eyewitnesses. And will reduce the frequency of false positives. Compare this to "suspect is a 6' male approx 200lbs with a white shirt and blue jeans" and then having police frantically pick up everyone in the area that meets this description.
This is the story that gets attention though. Despite it representing an improvement in likely every potential metic you can measure.
The response is what is interesting to me. It triggers a 1984 reflex resulting in people attempting to reject a dramatic enchantment in law enforcement ostensibly because it is not perfect. Or because they believe it a threat to privacy. I think people who are rejecting it should dig deep into their assumptions and reasoning to examine why they are really opposed to technology like this.
> think people who are rejecting it should dig deep into their assumptions and reasoning to examine why they are really opposed to technology like this.
Because a false positive ruins lives? Is that not sufficient? This man’s arrest record is public and won’t disappear. Many employers won’t hire if you have an arrest record (regardless of conviction). His reputation is also permanently smeared. These records are permanently public and in fact some counties publish weekly arrest records on their websites and in newspapers (not that newspapers matter much anymore)
Someday this technology may be better and work more reliably. We’re not there yet. Right now it’s like the early days of voice recognition from the ‘90s.
2 replies →
What is a unique use case for facial recognition that cannot be abused and has no other alternative solution?
Even the "good" use cases like unlocking your phone have security problems because malicious people can use photos or videos of your face and you can't change your face like you would a breached username and password.
I've got to be honest: I'm getting the picture the police here aren't very competent. I know I know, POSIWID and maybe they're very competently aiming at the current outcome. But don't they just look like a bunch of idiots?
In this particular case, computerized facial recognition is not the problem.
Facial recognition produces potential matches. It's still up to humans to look at footage themselves and use their judgment as to whether it's actually the same person or not, as well as to judge whether other elements fit the suspect or not.
The problem here is 100% on the cop(s) who made that call for themselves, or intentionally ignored obvious differences. (Of course, without us seeing the actual images in question, it's hard to judge.)
There are plenty of dangers with facial recognition (like using it at scale, or to track people without accountability), but this one doesn't seem to be it.
> The problem here is 100% on the cop(s) who made that call for themselves
I disagree. There is plenty of blame on the cops who made that call for themselves, true.
But there doesn't have to be a single party who is at fault. The facial recognition software is badly flawed in this dimension. It's well established that the current technologies are racially biased. So there's at least some fault in the developer of that technology, and the purchasing officer at the police department, and a criminal justice system that allows it to be used that way.
Reducing a complex problem to a single at-fault person produces an analysis that will often let other issues continue to fester. Consider if the FAA always stopped the analysis of air-crashes at: "the pilot made an error, so we won't take any other corrective actions other than punishing the pilot". Air travel wouldn't nearly as safe as it is today.
While we should hold these officers responsible for their mistake (abolish QI so that these officers could be sued civilly for the wrongful arrest!), we should also fix the other parts of the system that are obviously broken.
The facial recognition software is badly flawed in this dimension. It's well established that the current technologies are racially biased.
Who decided to use this software for this purpose, despite these bad flaws and well established bias? The buck stops with the cops.
7 replies →
You are being downvoted but you are 100% right.
The justification for depriving someone of their liberty lies solely with the arresting officer. They can base that on whatever they want, as long as they can later justify it to a court.
For example, you might have a trusted informant who could tell you who committed a local burglary, just this on its own could be legitimate grounds to make an arrest. The same informant might walk into a police station and tell the same information to someone else, for that officer, it might not be sufficient to justify an arrest.
> Even if this technology does become accurate (at the expense of people like me), I don’t want my daughters’ faces to be part of some government database.
Stop using Amazon Ring and similar doorbell products.
The pandemic has accelerated the use of no-touch surfaces specially at places like an airport that are more inclined to now use a face recognition security kiosk. What's not clear is the vetting process for these (albeit controversial) technologies. What if Google thinks person A is an offender but Amazon thinks otherwise. Can they be used as counter evidence? What is the gold standard for surveillance?
NPR article about the same, if you prefer to read instead of listen: https://www.npr.org/2020/06/24/882683463/the-computer-got-it...
I'll be watching this case with great interest
Sadly, there's plenty more where that came from.
And now the poor guy has an arrest record. Which wouldn't be a problem in reasonable legislations, where it's nobody's business whether you've been arrested or not, as long as you've not been convicted.
But in the US, I've heard that it can make it harder to get a job.
I believe I'm starting to get a feel for how the school to prison pipeline may work.
Wait until you hear about how garbage and unscientific fingerprint identification is.
Speaking of pseudoscience, didn't most police forces just start phasing out polygraphs in the last decade?
Unlikely unless they were compelled by law or found something else to replace it, and I think it's the latter. Something about machine learning and such.
In alot of police departments around the world, the photo database used is the drivers license database.
There is clothing available that can confuse facial recognition systems. What would happen if, next time you go for your drivers license photo, you wore a T shirt designed to confuse facial recognition, for example like this one? https://www.redbubble.com/i/t-shirt/Anti-Surveillance-Clothi...
I would love to see police trying to take a crack at this from the other side of things. Instead of matching against a database, Set up a style GAN and come up with a mask of the original photo or video to isolate just the face and have the discriminator try to match the face. Then at the end you can see the generated face with a decent pose and more importantly look through the range of generated faces that result in a reasonable match to give a somewhat decent idea of how confident you should be about any identification.
While this case is bad enough, mistakes like this are not the biggest concern. Mistakenly arrested people are (hopefully) eventually released, even though they have to go through quite a bit of trouble.
The consequence that is much worse would be mass incarceration of certain groups, because the AI is too good at catching people who actually did something.
This second wave of mass incarceration will lead to even more single parent families and poor households, and will reinforce the current situation.
How does computerized facial recognition compare in terms of racial bias and accuracy to human-brain facial recognition? Police are not exactly perfect in either regard.
Face recognition widens the scope of how many people can be harassed.
While also enabling finger-pointing, e.g. the police can say "We aren't racist or aren't at fault. The system is just faulty." while the engineers behind the facial recognition tech can say that they, "Were just doing their job. The police should've heeded their disclaimers, etc."
It's supposed to be a cornerstone of "innocent until proven guilty" legal systems that it is better to have 10 guilty people go free than to deprive a single innocent person of their freedom. It seems like the needle has been moving in the wrong direction on that. I'm not sure it that's just my impression on things, or if it's because there's more awareness with the internet/social networking of issues...
No mention of whether a judge signed a warrant for the arrest. In what world can cops just show up and arrest you on your front lawn based on their hunch?
If it's statistically proven to not work with black people then I think the only options are
1) Make it avoid black people, i.e. they aren't stored in the database and aren't processed when scanned.
2) Put a 5 year hiatus on commercial / public use.
Either of these things are more acceptable than too many false positives. #1 is really interesting to me as a thought experiment because it makes everyone think twice.
The issue is that it's hard to determine who is considered "black" and who is not since race is a made-up fiction constructed by white supremacists that is now so engrained in American society we find it difficult if not impossible to reflect on it. Are we using the Fitzpatrick skin color scale? Are we continuing the hypodescent rule (an algorithm, btw)?
Maybe we just outlaw face recognition in criminal justice entirely.
It definitely shouldn't be allowed as the main evidence - perhaps the answer to skin tone is to use the data and see where the algo starts to get too many false positives..
The largest - over arching idea - is to get everyone to think twice by making the majority think twice - If white people think 'its only for us!?' it'll make them really study the effects.. (I'm white.)
This technology will never be ready to use like this.
Similarly we shouldn’t collect vast databases of fingerprints or DNA and search them for every crime.
Why? Because error rates are unavoidable. There is some uncertainty, and in large enough numbers you will find false matches with perfect DNA matching.
We must keep our senses and use these technologies to help us rather than find the hidden bad guy.
well, I'm going to take something out of my chest. every time I shared a project here using machine learning people always gave me crap. saying my models were simplistic, or I did something wrong or the solution didn't work 100% of the time. well, I studied ML back in college. the basics, the algorithms that started all, linear regression, perceptron, adaline, knn, kmeans... and guess what? ML doesn't work 100% of the time. I always wanted to see how people would react when a car driven by ml hits something or when they based an important decision based on the classification of an nn. ML should be used along side human intelligence not by itself. you don't blindly trusts a black box.
Due process should not be abandoned in favour of automation. This was police negligence as much as it was a software mismatch.
One more thing, the article was being to dramatic about the whole incident.
The worst part is they use facial recognition which finds someone that looks like the suspect, and then they put the guy in a lineup and have him identified by the victim. Wtf?
That’s a good point. The selection from a lineup then gives less information. So the real issue is poorly understood priors.
I’m pretty sure that this can be used fairly with a rigorous Bayesian treatment.
The prosecutor and the police chief should personally apologize to his daughters, assuming that would be age appropriate.
I've been thinking this sort of event has become inevitable. Tech development and business models support extending the environments in which we collect images and analyze them. Confidence values lead to statistical guilt. I wrote about it here if interested: https://unintendedconsequenc.es/inevitable-surveillance/
relevant: https://www.theregister.com/2020/06/24/face_criminal_ai/
> In Williams' case, police had asked the store security guard, who had not witnessed the robbery, to pick the suspect out of a photo lineup based on the footage, and the security guard selected Williams.
Great job police
Boston just banned facial recognition, as have San Francisco, Oakland and a bunch of other cities.
You can join this movement by urging your local government officials to follow suit.
A human still confirmed the match right? That makes this not a facial recognition issue but something else.
A human who had only seen the same grainy security footage that the algorithm saw.
TOTAL FAIL.
All that rigamaroll for $3800 worth of crap? They should just switch it up and start entrapping people like the FBI does. Then at least they would have perhaps one leg to stand on.
Understandable, all black men look the same.
sounds like this guy is about to get a big payday.
That and we might get some kind of judicial ruling that current incarnations of facial recognition software are racially based.
It would be a great result if a court declared that the use of racially biased facial recognition software is a violation of the 14th amendment violation, and enjoined PDs from using such software unless it can be demonstrated to be free of racial bias.
Might have something to do with the...monochromatic...makeup of most software company C-suites.
Fortunately, that is changing, but not that quickly.
Facial recognition is a solution to a problem we don't have. It is the smartwatch of ML.
Will he? I thought it was pretty hard to win cases against the police. What does the evidence about the practicality of pursuing police for torts say here, and for that matter what kind of evidence should we be looking for?
One can file a civil case for false arrest. However, thanks to the qualified immunity cops enjoy coupled with the newness of the facial recognition technology, it will be very hard to prove that they obviously violated clearly established law. Maybe the cops will even point their fingers at the software.
Maybe he can go after the makers of the facial recognition software, but they can probably point their finger at the cops for using it wrong.
So, in any case, the guy will be left with a big legal bill.
most cases are settled by the city/county before it goes to court
at tax payer expense.
It must be frustrating to be a taxpayer, demanding to defund the police while simultaneously having to pay for their mistakes.
If only their elected officials would listen to them...
1 reply →