We have these AR-15s and all dressed up in SWAT gear, why not start something where we get to use them. /s
Sometimes when you get all the equipment, it can encourage usage, just like military gear to police, or large databases gained from surveillance that make detective work way too easy.
I feel like law enforcement has relied on data from surveillance for so long that everyone has forgotten real detective work and there is just lots of data noise and toys they are using to up budgets, in the end it might be the end of us.
This reminds me of when I was a little kid, maybe 2nd grade, and our class took a field trip to the police station. We all had our fingerprints taken. At the time, it just seemed like a fun activity to show us a little bit about how police work is done.
Fingerprinting kids is a regular thing in the US. I had it done when I was young, maybe 30 years ago. I believe it's used to help identify kidnapped (murdered?) children.
It was so long ago I don't know if they still do it or what's done with the fingerprints now.
Yup. 30 years ago or so, I was in 1st or 2nd grade and I remember they fingerprinted all of us in the school cafeteria. i think they told us it was in case we get kidnapped.
Fingerprints were collected with their permission. There was a lot of stranger-danger anxiety back then.
I guess it’d be good if I had been kidnapped and they could find my fingerprints at the kidnappers house, or prove that I am “me” if I was kidnapped and escaped years later.
I know that this might be an unpopular opinion, but if privacy really is dead, why don't we just make this data public? I know there's significant risks, but I feel like it's a better step forward than having this technology developed and deployed in secret.
I'd be ok with that if we also a) make public every time the data is accessed, including who is doing it, and b) require that anybody accessing the data have at least as much data public as they are accessing.
I think the worst problems here are around asymmetries. If I secretly have all your data but you don't know that and have none of mine, there's a power imbalance. But if we have equal amounts of data and know when each is looking at the other, that seems far more reasonable to me.
Depends on how it's done, but if penalties are sufficiently large for that, hopefully not. Especially if we give big awards to bounty hunters and whistleblowers who find malfeasance.
Yes, transparency for all, and I mean all, is the only other way to balance the scales of power.
It's either privacy for everyone, or transparency for all, but this in between is the worst where some entities get leverage over others because they have access to information that others don't have about them.
Right, so then we'll just have balanced power between everyone who can afford the AWS or other compute service bill do to the facial recognition matches against all those millions (and later, hundreds of millions, then billions) of records. That'll be...
Oh. Hm.
Maybe "privacy for everyone" is a better point to aim for.
The alternative seems to be no privacy as a baseline and then still results in an everybody-loses arms race to top it off.
Because these are all pictures of juveniles. Not even children who were convicted of a crime, those who were arrested. Children should be able to make mistakes and to learn and recover, without a bad decision following them for the rest of their lives. That's not only good for the child, it's good for society. It's why juvenile records are usually sealed and/or expunged and why data like this should never ever be public, even if "privacy is dead."
Dissent is useless if you can't give voice to it, and that's exactly what these systems seek to do -- make it dangerous to do anything the state doesn't want you to do. From speaking in public to peaceful protest these databases will eventually be used to identify anyone who has ever peacefully participated in a democratic society.
This was what the FBI wanted in the 50s, 60s and 70s, and now, 70 years later, they finally have it.
We can't stop eyes (biological and electronic) from seeing faces. We can't stop memory banks (biological and electronic) from storing the likenesses of faces.
Better than trying to craft legislation that we know the state will secretly violate is to make all of this data public and make the world's human facial lexicon part of our shared understanding.
This is the most ideal solution for an optimistic society.
A real answer: because I personally don't want to make job easier for crazies. Your comments stick around on the web instead of evaporating in thin air. Among hundreds who may read them, there may occur a couple angry people with skewed priorities. I don't want such people to easily know how I look and where I live.
Which way are things trending where you are today? The cost to surveil is going down at a tremendous rate so it is just a matter of time before someone decides to pay to surveil you and then sell their insights or take advantage of them. I don't think it's possible for politicians to keep up with the rate of tech advancement. Did you ever watch a congressional hearing in the US on any technology realated subject? It's a comedic bit.
>Did you ever watch a congressional hearing in the US on any technology realated subject? It's a comedic bit.
This comment could be straight out of 1870 when things were industrializing rapidly. We'll figure it out eventually. Hopefully without any serious oppression and/or bloodshed in the process.
My point is that if corporate surveillance with creepy camera technology and facial recognition is going to be rolled out everywhere (for the real safety benefits) then isn't it better for us all to spread the power out as much as possible rather than it being concentrated in the hands of a few giant entities?
The ultimate "f&#$ it" move. It's true ... your thoughts, words and actions are all going to be documented and preserved to be later audited and evaluated into perpetuity. The train is too far down the tracks to do anything about it.
On a related note, I was recently disgusted by Bunk1 [0], a SaaS for summer camps that allows parents to stay connected with their children, as it highlights "Facial Recognition" as one of its features on its homepage.
A friend's child is in a camp that unfortunately uses them, and "invited me" so that I could send a letter to the child (which is a useful feature). The URL of the child's photo is:
When parents upload a photo of their camper, our facial recognition software scans each photo posted by camp and notifies parents when photos of their camper are available. Family members can easily see their camper in action without hunting through the camp’s entire gallery."
The whole problem is how do you selectively use facial recognition? It's not like there's a human judging whether the camera took a photo of someone underage, and somehow classify it by hands. The system is probably deployed by 90% software engineer and NONE of them probably even saw 0.01% of the actual database outside of the few example they used to debug. Those images will never be visited again unless there's a specific need to see them.
> The system is probably deployed by 90% software engineer
And this is the problem: The ruthlessness of engineers who then will wash their hands in innocence.
Physicists creating nuclear weapons. Chemists creating chemical weapons. Software Engineers creating an Orwellian surveillance wet dream. These people need to get the Perillos of Athens treatment, so they and those who come after them can learn that ethics is a thing in their profession.
Honestly, I have no problem with a database of public images showing/tracking people's public actions. It used to be in a small town you knew who was good and bad by their actions. With the population explosion, the only way we have a practical means to return to that is with facial recognition databases.
First of all, there is a strict process in proving that someone is a homosexual in Saudi Arabia. It’s not as simple as “here’s some pictures of this guy leaving a gay club, execute him”. If it were that easy the system would be rife for abuse obviously, especially in an age where any image can be faked or taken out of context. Second, you can simply deny the images were even real, or that it was not you, and further increase the burden of proof required.
Assuming everything you say is true ... your point is what? This hypothetical person would probably not be executed? Going through a trial for your own sexual orientation but being acquitted doesn't sound like a great time to me.
If you're gay, you should know the risks about going to Saudi Arabia or any place with similar laws and avoid going there. One does not usually "end up" in a different country unless they're kidnapped and at that point the homosexuality laws are really the least of your problems.
Your argument implies that 'it's the gay person's fault for getting caught,' so that tracking organisations have less responsibility for their actions.
Instead of Saudi Arabia, how about the United States 100 years ago? It is not as though there are no more harmless activities for people to participate in that they will suffer for.
A comparatively trivial but important example: how about tracking your employees movements to punish them for talking to other companies? Is that a 'you should know your risks' situation too?
It implies no such thing. There is no blame because this is a contrived hypothetical that makes a number of ridiculous assumptions.
1. That someone going to a gay club in NYC later decides to go to Saudi Arabia, knowing full well that the government there is actively hostile to them.
2. That the government of Saudi Arabia somehow has access to facial recognition data from the NYPD.
3. This data for some reason includes not only mugshots directly collected by the NYPD (which was what TFA was actually about) but also private survellience footage from gay clubs, and this footage is clear enough to make a match (remember people tend to go to clubs at night)
4. The government of Saudi Arabia decides that they need to expend a tremendous amount of computing resources to run facial recognition of all of their survellience footage of every face in their country against this fictional database that they somehow obtained.
5. Further they conclude that a single facial recognition match, which is known the flawed, of a person in the vicinity of a gay club is sufficient to convict someone.
There are real privacy concerns yes, but this is absurdist fearmongering and deserves to be ridiculed.
The point is that the risks are much lower if there's not a database storing this information to be leaked in the first place.
The whole "if you don't want to go to jail, don't commit crimes" line of thinking really falls apart when you realize that the legal system is well beyond the ability for a layman to navigate on a day to day basis and gets even worse when you have a perfect surveillance state capable of seeing and storing massive amounts of data.
When the government can trivially make the case that a citizen is a criminal just by looking deep enough in the archive it's time to start considering that maybe your government isn't protecting/serving you anymore.
If you're gay and were born in Saudi Arabia, and managed to get to a point where you're photographed leaving a gay club in NYC, then don't go back to Saudi Arabia.
Governments change, as do treatments of different groups of people. Perhaps the government had no issue with your practicing of a particular religion or being an immigrant before some disaster, war, or nationalist wave changes things.
And anyone can come up with some contrived chain of events about how any particular technology might be misused. That's not a reason to dismiss it wholesale, otherwise we would never have any technology at all.
The next Hitler will be horiffyingly more efficient when they can track everyone by their phone location and know a history of everyone they interacted with as well as facial recognition on every street.
Really? Saudi Arabia has historically, again and again, used facial recognition data collected by the NYPD to hunt down people who went to gay clubs in NYC, who for some reason spontaneously "end up" in Saudi Arabia via some unspecified process they can't control?
Yeah, remember when that happened and then we banned governments from using electronic records or computer data processing? I guess that's why the government keeps all their records on paper now and does all of the processing by hand.
Because computers were used for something evil once so it must be the inanimate objects that are themselves inherently evil and not, you know, the people using them. What a weird, reactionary, alarmist stance to be so prevalent on a supposedly technology based forum.
Christ, how naive. Do you think people have changed any? Different groups are taking advantage of these records every day. It happened in the past and will in the future. The only thing limiting the most horrible examples is opportunity. That's why we must limit our exposure to those who would do us harm.
I also recommend a reading of the history behind the "bill of rights."
Are you counting the NYPD among "those who would do us harm"? Because that is what the actual exposure is here. This talk about Saudi Arabia is hypothetical nonsense. TFA is talking about the NYPD using facial recognition to suggest a possible lead to human detectives, who would then use actual judgement to decide if a match to a five year old photo of an eleven year old is something that is likely to be worth pursuing.
If your position is that this is the equivalent of SS soldiers breaking down your door and hauling you away to a concentration camp, then I honestly don't know why you aren't in the street rioting right now.
We could have a useful discussion about realistic pros and cons of things that are actually happening if people would leave off of the nonsensical hyperbole for like, a fucking second.
Definitely. Police are doing real harm (and good) every day. And are dying to get their hands on technology to make their jobs easier with no thought to negative consequences to others.
The most effective way to fix abuses is to prevent them in the first place. We restrict govts and police by design, a need recognized for centuries. If a capability can be abused, it will be abused. Not if, when. That's the last reply such a shortsighted argument will get.
(Also, am not restricted to other's problematic examples upthread.)
There's some horrendous stories of this happening in PNG. The police are present and try to stop the mob but they are outnumbered 100-to-1 by a group of villagers intent on murdering some poor innocent woman accused of being a witch and causing illness in the community.
They stand by and watch as she is lynched.
have you considered the obvious fallout from that? one example: bad actors use database of people's public actions to privately harass them (e.g. following them to work => ID employer => extort/blackmail)
what is your solution to this outside of panopticon?
They have, and I went to court for it, and won. It's not easy, but stating that any system that can make mistakes is not worthwhile is frankly bad logic. Can police make mistakes today?
Well, victims of domestic violence sure have a problem with it! Though the stats are a bit murky and hard to parse, up to 1 in 4 women are victims of domestic violence [0], so it's not an edge case.
>The New York Police Department has been loading thousands of arrest photos of children and teenagers into a facial recognition database despite evidence the technology has a higher risk of false matches in younger faces.
They already have the photos from arrest records. They already have footage from security cameras, etc. I don't see a problem in using computer algorithms to try to find possible matches between these two data sets. We do this for fingerprints and DNA all the time, matching data that was collected during an arrest to that from a crime scene.
Frankly, I do not have a problem with this, and actually think the police would be derelict in their duty if they were not doing this.
Except DNA and Fingerprints are far less accurate than you’re led to believe [1]. Having worked with DNA, contamination is likely insanely probable and I doubt the validity of all cases it is used as the only piece of hard evidence (it’s alright circumstantial evidence).
The greater question has to do with what our rights are. The reality is, our system has been setup such that everyone commits crimes. By applying highly effective methods, you’re essentially enslaving the population. That’s why we had laws protecting freedom of speech and unreasonable search and seizure. The U.S. was setup to be particularly sensitive to those issues. The reason they are in the constitution is it was believed government would eventually try to chip away at the freedom and it did.
That’s why people are concerned. Their duty is to keep the peace. That does not necessarily mean or need to mean spying 24/7.
They are young and learning and deserve a chance to better themselves without being hassled by computers. Also, what chance has facial recognition to match a 11 year old kid three years later? It quickly becomes useless data, don't you agree?
I wonder how this affects their system. If you have 10 years of a changing facial profile on someone, is their algorithm smart enough to drop older items? Or do you, in your 20s, look like your ten year old brother?
I'd imagine the problems are far more fundamental -- do they work at all given the typically poorly constructed training sets. And also what is their FP/FN rates for subsets of the population -- do they work as well/poorly for everyone?
The problem I see with many of these systems are -- they are trained on one population and used for inference on a very different distribution (minorities.) OR their accuracy is only high on the overall population, not universally accurate across all subsets.
When those in charge, those with political power, etc (usually non-minorities) "see" the performance of the system, they experience it in their own context where it works. They don't feel the pain from the experience of the sub-populations where false positives/negatives are larger.
My guess is that these facial rec systems aren't very sophisticated. I suspect that they just compare a sample image with their database of images, so as people age, they just look less like their image in the database, so the similarity scores will simply decrease as the data grows more stale.
Better idea: You have a surveillance network capable of facial recognition. My child goes missing. The only way for you (the police and your network) to know about this is if I inform you. At that point, you can request images of the child and feed them into your machine to see if you can get a hit.
Probably right here, discussing the USA this time because that's the topic. It would probably not contribute to the discussion to drop references to China.
Young people can be criminals too. And if the police catch them while still juvenile, they might have more of a chance to turn their life around.
Young people can also go missing, and facial recognition of security cams could help find them.
I honestly can't see any reason why children shouldn't be included like everyone else.
It goes without saying that a conviction should never depend on computer matching, and that safeguards need to be in place to prevent abuse and limit to legitimate law-enforcement purposes.
But trying to locate suspects or victims in the first place? Children are no less important here than adults.
You can reuse that comment when we get mandatory under skin gps tracker from birth. We could even sacrifice all our liberties and get rid of every single crimes if we went further.
There is a big difference between your skin gps tracker and facial recognition. The skin gps tracker is a new data source. Facial recognition is making better use of already existing data from mug shots and security cameras.
If you are concerned about privacy, then let's regulate what data you can collect and from whom you can collect it. However, once you have the data, police should be able to use computers to better understand, filter, and generally make the data more useful to solving crimes.
Everything is new data source at some point, everyone gets outraged for a bit and then it's business as usual. There is no stopping ""progress"" and technology.
There was a story here locally where a young boy wandered out of his house while he was thought to be napping. The parents noticed and a search was started, but after a few hours they found him inside a neighbors car, dead from the Texas heat. A GPS tracker would have saved his life. I understand the privacy and personal liberty concerns, but I have a hard time coming to a decision on this one myself.
That's a ridiculous argument. There are a myriad of other things that could have saved his life. Safety measures around the house to prevent young kids wandering outside at night, home surveillance to keep an eye on the property and alerts anytime movement is spotted around the house after a certain time, or even teaching kids not to go wandering off in the dead of night. How did he even get into the neighbors car? Was it malicious? Then it has nothing to do with tracking, they should worry about that neighbors record. If not, how would tracking have helped a young kid somehow getting themselves trapped in someone else's car?
Notice how eroding privacy even further is not even in the top 10 things that could have improved this situation
You can find an isolated incident for anything. When you get your GPS trackers, you'll be able to tell an anecdote about a child who was hiding from a criminal, but was found by a GPS tracker.
> Young people can be criminals too. And if the police catch them while still juvenile, they might have more of a chance to turn their life around.
The US justice system isn't designed to turn anyone's life around -- it's designed to punish people. There's no benefit whatsoever to a kid spending time in jail or prison (or to putting their parents in debt over fines or bail payments).
>It goes without saying that a conviction should never depend on computer matching, and that safeguards need to be in place to prevent abuse and limit to legitimate law-enforcement purposes.
Lol. Has the past half century or so of law enforcement hammered in a belief in the moderation of power by law enforcement? That such power is never used asymmetrically against the lower strata of society as opposed to the elites? Will it prevent people from being the victims of selective enforcement of laws, even non-violent ones involving the mere possession of drugs?
ah, yeah, outstanding argument - looking forward to seeing it applied to all other electronic data ever produced, no matter how old they are