Federal civil rights watchdog sounds alarm over Feds use of facial recognition
therecord.media
therecord.media
It should be straight-out illegal. Governments do not have "free speech" rights.
As for private citizens doing it: I think there are already sufficient laws about recording people without their consent. You can hire someone to stand on a corner and watch for a specific person to walk by, but a law prohibiting you from recording everyone who walks by is most likely going to withstand a court review. It's a question of scale.
(Yes, I use rewards cards, but I have the option to not enter my phone number and pay cash if I want to exclude a particular purchase from that dataset.)
Anyway, a mall is a public space in the context of recording without consent. You have no reasonable expectation of privacy. The law needs to be updated to reflect the vastly more invasive technology we have now compared to when these statutes were written to reflect actual social mores.
we were talking about the law, as I recall.
> Anyway, a mall is a public space
malls have been held to be public spaces, but not the insides of the stores.
I predict a case about this will reach the Supreme Court in the next few years.
Authoritarian governments, of course, have all the rights they manage to grab. Look what Russian officials say. I suppose that North Korean officials can say anything their bosses tell them to say, any public opinion notwithstanding.
The government is protected for speech related to its duties, in clearly spelled out ways.
..which is a different claim than "governments have no free speech rights." They do, just limited in ways that individual free speech rights are not. But even individual free speech rights aren't absolute.
Language matters.
You can argue that the solution is to stop doubling down on our bad investment, bad that is much less feasible than installing a bunch of traffic cameras.
Meta: I welcome the dog pile that will ensue.
As big of a source of death it is, it's still an infinitesimally small one in he grand scheme of things.
Infinitesimal? It’s one of the leading causes of death among youth on the order of suicide, homicide and drug overdoses. Calling it infinitesimally small is horribly offensive.
It’s a big reason why our life expectancy is lower than peer countries. Again to reiterate, we spend a ton of our GDP just in healthcare to address traffic violence. Modern fire departments mostly address car accidents rather than fires.
You clearly don’t know what you are talking about and you are going to hit a sore spot in many Americans
What was wrong with "traffic accidents"? People had gotten used to it so you had to invent something new?
It is a shame because it would be a fixed problem if we held people accountable or fixed land use or designed infrastructure to be safer
This is an easily solved problem. More on the nuance: humans are imperfect when driving, so design infrastructure around that.
What are you gaining out of this? Like do you also look aside when gun violence is brought up?
Stopping the deliberate debasement and pollution of our language. That's what I'm gaining. Or trying to.
Call it what it is ("traffic accidents"), and we can stop arguing. Other countries have fewer accidents? OK, that's worth talking about. I didn't "miss the nuance" because I'm not responding to that right here.
Or "bad road engineering" if that's what you want to talk about.
Language is living. Get off your high horse. People are allowed to invent and use new words and terms. Language evolves with use, not by people like you holding on to dear life for every little thing.
read https://www.orwellfoundation.com/the-orwell-foundation/orwel...
Inventing inflated words to make something old sound like a fresh crisis is indeed debasement and pollution. I guess you don't know or care what that is.
So don't tell me what to say. You can keep doing it and I can keep calling it out, and there isn't a thing you can do about it.
If you're speeding to pass someone and hit an oncoming car, that's not an accident.
If you're impatient and try to squeeze by a bicycle and hit them, that's not an accident.
If you're texting on your phone and rear-end the car in front of you, that's not an accident, it was a conscious decision.
If your transmission seizes or your wheel fails off so you fly off the road into something (and you haven't been ignoring maintenance on your rust bucket for so long that you should expect this), that's probably an accident. But that's an infinitesimal fraction of vehicle incidents.
I do find it hard to believe rockson doesn't know anyone who has been involved in a car accident if they live in the US.
Calling it propaganda is also crazy. Like you want people to keep dying this way? Great way to show your true colors pal
That is why it is accurate to call "traffic violence" a propaganda device.
"Traffic violence" is not a thing. Call them "traffic accidents" as we always have, and we can move on.
And then, of course, bicyclists often run into other bicyclists, or into pedestrians. So now there's "bicycle violence"?
Soccer players often get injured, or get into fights. So there's "soccer violence"?
This is fair. The problem is it's being done with zero controls, let alone verifiable ones.
I'll take the other side of that bet.
> sufficient laws about recording people without their consent
Such as?
First off, this is not true, you definitely have some rights to privacy in public (an trivial example is using a public restroom). The law also is based on what is reasonable, and I don't think nearly everyone would think it is reasonable to expect some privacy if you walked down an alley to get away from others or went to a more secluded area. These are often explicitly why people do such things.
Second, recognize that many of these laws were written long before the average person had a camera and microphone in their pockets (not to mention high quality). The environment changed and it makes sense that our laws should change to. So even if you reject the first point these arguments about "what the law says" are still not enough to dispel concerns because we can in fact change the laws. More so, we _should_ constantly be updating laws due to our changing environments and how we learn about things as they happen.
And yeah, there's also a significant difference between a private person and a government doing these tasks. The government has a lot more power, and this does mean they require higher scrutiny and accountability even if you have complete faith in them (yes, this also means corporate entities, since the concern is power dynamics). Scale certainly has changed in the last 100, 50, 20, and even 10 years. It happened fast and no matter what side you are on, you can't ignore that things have changed and are about to change even more.
I think it's also worth noting that most people don't want this to happen even when they do know. Of course some people don't care but even they do when they learn how it's being used
Something legal applied a billion times doesn't make it illegal.
That's also one of the legal principles underlying mass warrantless surveillance with the third party doctrine.
This is the world we live in. But I'm sure many of you have money - why not try organizing and changing the law.
Carpenter v. United States I think is one of the landmark cases that has started to signal the impending curtailment of unabridged third-party doctrine. Kyllo v. U.S. laid some groundwork as well, I believe unless I've mixed it up with another case in which the court recognized it must balance the ever expanding capabilities granted to law enforcement by new technology against the right of the populace to maintain their expectations of privacy. In part, those tests have to take into account the impact of the particular means being employed. In the Kyllo case, a thermal imager being used to see through the walls of a structure without a warrant constituted an infringement of the 4th Amendment because accepting that law enforcement could do this would be to let law enforcement intrude through artifice where historically they would not have been allowed to tread without warrants. Applied to the entirety of the population, this would represent a wholesale nullification of the 4th Anendment by mere accident of now having a sensor exist that could make that previously private, public. Based on said analysis, your assertion talls short. It is in fact the goal of the Court to specifically not infer that just because something is authorized for use in a singular instance, does not mean it's character does not change when applied a billion times.
Courts are not computers. They are not stupid. They business of weighing nuance is sorta their thing, in spite of how annoying they can be at utterly torturing language when they want to do something a plain reading won't let them do.
Perhaps "I'll do the scan just now, but also, while it has my motivation and attention, I'll press the easy button to donate $N to some org effectively fighting it". Or "I'll opt out, and press the easy button to encourage others with a somewhat-anonymized "I/someone at airline-or-city-or-region-or-occupation-or... just opted-out! N+1! - Yay future!" tweet/counter"?
One way to shift the equation is to shift understanding of the tradeoff:
Some minor inconvenience - I think none in this case - for freedom and other human rights is hardly a tradeoff that needs much consideration. Honestly, I think it's people following the herd to the cultural trend of despair, powerlessness, and worship of strongmen - a incredible reversal from American individualism I'd read has dominated the culture for centuries.
More practically, a little courage and action has a large effect on others. Imagine you are in a meeting where something obviously morally wrong is likely to be approved, and there is pressure to go along:
Imagine nobody raises their hand to object. Imagine the despair and powerlessness everyone would feel, that it just passes without objection, that everyone is cowed.
Imagine one person raises their hand. Not only does it greatly reduce the feeling of despair and powerlessness, it heartens people, and even more it makes it far easier for a second person to raise their hand, or for others to object to this issue or to others later.
Imagine two raise their hand ...
Frankly, it's intimidating, and I don't blame people for not opting out even assuming they know that's possible
Boil the frog. If they start to collect more, it’s not as simple as “they already have my picture.” Should there be a breach, I can’t easily change my face.
I’m generally of the opinion that when it comes to personal data collection we should always say “no, convince me you should” rather than “sure, why not?” Other than an ever-so-slightly faster processing time is there evidence of any real security gain? My initial reaction, uninformed on the specifics but fairly versed in this type of security, is that’s unlikely.
But more to the point the reason for doing it is there, for better or worse — want a passport? Have a picture taken. In this case I am (at least for now) comparing opting out or not. You can still take your flight without them taking that additional picture. That of course may not stay true.
https://travel.state.gov/content/travel/en/passports/how-app...
The result? Faster passage through gate security, significantly faster than TSA Precheck in fact. As a frequent flyer, I like it.
So that's an easy 10~20 minutes shaved off. Once I'm ushered through to screening I have the same convenience benefits as TSA Precheck since it's a prerequisite for Digital ID anyway.
The TSA Precheck lines at all three airports are horrifically long. ;_;
Some people care a lot. But take great care not to be noticed.
For some people the concentrated attention of the state is too be feared. So that is why they say "yes sir, no sir".
It is a matter of survival
But I have global entry which is all done through facial recognition now (and matching to passenger manifests). No passport needed. The camera they used was just a normal DSLR, but it had no problem matching me.
The tech is getting really good.
In Singapore you can enter through automated gates without ever interacting with a person or showing a passport. All through facial recognition if they have a photo on file (citizens and permanent residents).
I assume the tech will get good enough to scan entire crowds very quickly and identify everyone.
It was much easier and less gropey than opting out of the mm wave scans.
CBP is another matter. When returning from abroad I find their agents to be more aggressive or snarky when I (politely) opt out, although I wouldn’t say it’s gone as far as retaliation. I often get asked why I’m opting out, and sometimes they follow that with additional questions. Several times I’ve received a standard spiel explaining that the photo is not stored permanently or transmitted; the vibe I get is that they’re trained to respond this way to “kooks.”
_He_ then held up the line as he walked off, consulted with someone who obviously knew better, came back and pouted while he verified my identity "manually."
https://news.ycombinator.com/item?id=41052152
It is possible that people's experience vary. But unless we know opt out is something that is done quickly as a routine, and running into clueless TSA agent is a rare exception, I would not try that.
- A few people report they’ve had issues opting out
- Most people report they have no issues at all
Personally, I’d feel like I’m perpetuating the problem (of privacy violation/overreach) by not opting out. Even if most people don’t care, it seems critical to have at least some portion of the population holding back the tide somewhat.
If nothing else, a consistent stream of people opting out helps us understand how big the problem has become, i.e. if discussions about this flip over time and most people are getting hassled.
Furthermore, the TSA scanners use a hi res stereoscopic camera to take a 3D photo of your facial features. It's not the same as your passport photo or a regular CCTV camera
I assume they must be face-matching on the passport photo database.
The camera was right by the counter, not behind the officer.
Given the reputation of groups involved, I expect many people who might have otherwise opted out felt that trying to do so would have revealed there was no efficient opt-out process, just an empty offer of one, or that opting out would have lasting repercussions for their treatment during travel in the future.
However, non-passengers (such as family members or friends) could accompany travelers to the gate, or meet them there when they disembark, passing through security without a ticket
Well not quite. They keep getting money every year, but officially their mandate expired in the 90’s and has never been renewed. As crazy as that is by itself, they certainly don’t have the authority to do anything.
> anonymous facial recognition
For a non-technical intuition of this you might notice that you can often recognize a friend walking away from you (no visible face) down the street. Even if you don't know why you can recognize them, it is down to subtle things like the way they walk, dress, hair style, etc. It might be harder to identify people this way and more prone to mistakes, but even when less accurate you could significantly narrow down possible candidates (even more with correlating information like location and time).
[0] This was a surprising thing to me because I suspected clothing would be a stronger indicator especially when face isn't always visible but this could also happen for other reasons and can be highly influenced in pretraining. Even saw it with blurred faces in testing, but depended on how these were blurred. Gate tracking also seemed to be picked up but I never confirmed.
The reason I ask is that india has a bit of a littering problem at bus stands, and the fines for littering are hardly enforced due to lack of budget.
I was hoping to fix the issue by applying a small fine, all the time.
To do it, I was hoping I could build a sort of person-hood tracking system that says, see, that fellow who threw a packet of chips is now standing at platform 10, near the concession stand, go catch him, easy 100₹.
By consistently enforcing penalties, we can change the behavior of people. It is worse than useless if it is inconsistently applied, for those caught will be likely to reoffend, considering it plain chance that they got caught.
I think this is one of the difficult things about creating things. It is easy to get caught up in the benefits of a technology and how it can be used for good. But technology isn't good or bad, it is a tool. Like any tool, you need to consider how it will be used, and that means how it will be abused. My undergrad was in physics, and this lesson is often taught when you get to atomics (or maybe because my advisor worked at Los Alamos and with many of the people who built the atomic bombs). Physics tends to not hide the skeletons in the closet because atomic weapons. But I think CS still does and maybe we haven't caught up to how influential our work is in modern day. How software can (and does) kill people. Or maybe it is the abstraction of software, working at a high level it is easy to feel disconnected. Either way, I think we need to tread more carefully and just make sure we're concerned about ethics. It is a personal decision where the lines are drawn, but they can't be ignored. After all, the road to hell is paved with good intentions, right? Unfortunately evil isn't only created by evil men, but often good men with good intentions.
FWIW, I'm all for the intent. But I question if there are other ways to solve these issues. But that may require a more complex and nuanced approach[2]. But if enforcement is already a problem, then I think there are bigger problems at play that this technical solution might make worse. Unfortunately as we advance the world gets more complex haha
[0] You can get pretty good if you do upfront processing to blur people, but that first means you need to detect people lol. So this would need to be done on device (as in, the camera itself. Which is possible btw).
[1] I saw a lot of videos from India (as well as the whole subcontinent and South East Asia), where the customers were trying to prevent theft (good usage), but there were some other things that had me concerned and were being hinted at. I also think one big difference is that I appeared to be the only person on my team that would look at data we were processing. Not to snoop or be invasive, but because this is the best way to solve data processing problems; you need to check your assumptions, learn what is creating errors, find things you didn't consider, and all that. To be good at ML you can't just pipe in data and expect results, and you shouldn't treat the algorithms like black boxes. While they aren't transparent ("white boxes"), that doesn't mean they're completely opaque either (and we're learning more every day).
[2] I think one key piece of information is that we need to ask why things like littering and theft is less common in richer neighborhoods. Or at least by their residents. Personally, my belief is that this is more to do with a personal safety net. It's easy for ethics to be eroded when you're worried about your next meal or being able to afford rent. These are obviously harder issues to resolve, but I think they are worth addressing. Because I think lifting up the quality of life of the lowest class of people benefits everyone. A rising tide lifts all ships, and I do believe it even lifts the biggest ships more.
How so?
Preventing facial recognition misuse (or use even). How is that impinging on the freedoms guaranteed in the First Ammendment?
On a slightly different note, why is the impasse "pathological"? Are you just throwing in a pejorative term?
> why is the impasse "pathological"? Are you just throwing in a pejorative term?
The current jurisprudence is a pathlogical (https://en.wikipedia.org/wiki/Corner_case) interpretation that clamps the solution space to maximal commercial/corporate rights at the expense of individual rights.
The example of the press is actually quite poignant. The current legal regime neuters the free speech rights of most of the press. People can be fired for writing articles saying things their employer does not want to be said. They can even be fired for saying things in their own personal time that their employer does not want to be said! The end result is greatly diminished rights for the individuals making up the press, in order to support greatly increased rights for the controlling owners of the press.
Maybe we shouldn't treat corporate speech the same as human speech under the 1st amendment, but that would allow regulation of corporations giving money to political organizations? Ok, well that's beyond the pale. That's obviously originally the founders intent in writing the bill of rights. I'm sure it's in the Federalist Papers somewhere.
Also, abridging corporate speech creates a precedent for abridging the speech of all human collective entities like religions, political parties, advocacy groups and the press, because there is no particular reason why corporations as an abstract entity created by humans to express collective human will should be unique in this regard.
You can't simply nullify rights when people exercise them in aggregate. That isn't how free societies are supposed to work, and it's too easy to undermine.
...corporations have thrown tons of money at making it so
Sure there is. If corporations (/LLCs) were prohibited from directing their humans to say specific things as part of their employment, those humans can still say whatever they want on their own time.
> Also, abridging corporate speech creates a precedent for abridging the speech of all human collective entities
It's disingenuous to pretend that a corporation (/LLC) is just some group of people. Rather it's an entity that has gone out of its way to obtain a government created liability shield. Conditioning that grant on following extra regulations aimed at mitigating the harm caused by the extreme scale fostered by the limited liability makes perfect sense.
> corporations as an abstract entity created by humans to express collective human will
Wut? The corporate mechanic is that of directing humans in a top-down fashion. The only people thinking corporations represent "collective human will" are the ownership and managerial classes who either are the people directing the corporation top-down, or at least have to repeat enough of the corporate kool-aid to get promoted. The majority of people view them as the least-worst option to receive a paycheck and often see their own wills suppressed - even though respecting that distributed intelligence would often help the corporation (ie "shit rolls downhill").
No they are not.
Legal fictions aside, it should be obvious
Then make it not so, we aren't forced to live with the status quo for eternity.
The law wouldn't be banning people recording faces. It would be much more specific: corporations can't record peoples faces at a mass scale to log them away and sell the data for a profit. If there's any edge cases to this, then let the courts handle it. We let the courts sort out grey areas already, that's their job.
The cynic might retort that such laws are perpetually vulnerable to subversion under the guise of national security. While there's a kernel of truth in that, it's equally undeniable that if 80% of the citizenry truly felt impassioned, any rogue elements within the government would face severe repercussions. Acts of retribution against such actors would not only be tolerated but perhaps even tacitly endorsed by the public.
If we could reliably attain passion, cooperation, consciousness? And ultimately, a belief in agency?
Then we would hardly even need to protect institutions against such subversion.
I think it takes both agreement and desperation for that kind of thing to happen, though. Comfortable people don't tend to buck the people who can freeze their bank accounts in any serious way at simultaneous scale, even when they're in mass disagreement.
I've always personally envisioned some terrible end-state of human history where vast swathes of people are in something more like a zoo than a prison. Without escape, but also without enough pain to push them toward resistance, until it's too late to turn back.
Heck, I want an aid like that - my built-in wetware face recognition is barely functional (I don't even recognize actors in the movies, save for a few most iconic ones), and tying faces only works after repeating it a dozen times. I can pinky swear I won't be sharing this with anyone. I probably won't even keep records, because it's literal digital junk (but I'd like to have a right to record everything that happened around me that I could've possibly remembered naturally). But that's not about it - I just hate how mass surveillance made a lot of people hate all the cameras without making any distinctions (sure, I think I get it).
What this is about is that while individual privacy in public is not expected (save for stalking or other forms of harassment, which aren't exactly about privacy anymore), privacy at scale is an open question. Large entities, like governments or big companies, can collect unprecedented amounts of data, and at that scale it holds a significant potential both for good use and for abuse. So, in a functional democracy, it must prompt a public discussion and search for a consensus. It can be argued that there are benefits, and whenever benefits outweigh the risks, and it also can be argued that old principles must be still upheld to the letter, or if reality had changed too much and those principles are not matching to the ideas and reasons...
It especially matters when it's about the government, because government abuse can get really nasty, even deadly. That's why all the safeguards, checks and balances.
Basically, I think it's not really about cameras, it's about ability to build large databases with information that may be, uh, socially sensitive.