5 ms·

> I think we need to be more precise about what exactly is spooking us and why. What's spooking me is the normalization of facial recognition technology for id

by skohan 8y ago

> I think we need to be more precise about what exactly is spooking us and why. What's spooking me is the normalization of facial recognition technology for identification without adequate public agreement over what the limits of its applications should be. A machine recognizing me is a lot different than a human recognizing me. If a machine can recognize me, there can be cameras posted at every street corner, and some organization can reconstruct my movements based on where and when my face is pinged by the system. It's sort of possible for humans with access to surveillance footage to do that, but the amount of human effort required to accomplish it makes it impossible to do at scale. An automated system could do this for everyone, always.

bayesian_horse 8y ago
So it's still not the problem that a machine can recognize you, but rather that somebody is tracking and storing such recognitions. Which is an entirely different thing. The former is a technological reality. It's impossible to legislate away devices who can do that. Tracking and storing personal data on the other hand, is very much within the reach of regulation. The GDPR for example restricts this. Even if somebody has devices who can recognize you is not allowed to store and track data about such recognition incidences without your approval. In this particular case, they don't need your approval because the camera is used in lieu of a boarding pass, and the airline has every right and duty to know you are boarding right now. In case you are worried that the airline connects your travel patterns with other companies interested in other patterns, well, that works with a boarding pass or even a simple check mark on a pen-on-paper boarding list just fine.
- reallymental 8y ago
  The machine recognizing you was the harder problem. That's the "Minority Report" stuff that people thought was in the _far far_ future that we don't have to worry about constantly. Storing and using that model across various devices is trivial. So if we've done the hard part, the easy part is _definitely_ going to be done. There's no doubt about that. States, companies that can convince the state that they somehow need the data (cough insurance companies) will, I repeat, WILL utilize it. And now, suddenly your premiums for this month have gone up cause in that ski trip you took last weekend, a camera noticed the lack of headgear for about 5 mins. That's the kind of risk the insurance company cannot take. Pay up.
  - bayesian_horse 8y ago
    "Minority Report" was about something that Humans couldn't really do, meaning to look into the future, and they even needed psychics to do it... Good luck trying to advance science by avoiding to invent anything that can recognize image!
    - skohan 8y ago
      Taken metaphorically we’re not far away. Imagine if we come to trust ML to make decisions for us in everyday life, and we build a system which can predict with high confidence that someone will commit a mass shooting. Is it so hard to believe that might be acted on?
      - bayesian_horse 8y ago
        To avoid these issues it is imperative to hold natural and legal persons responsible for the software they use, just as it has always been, regardless if it is using anything artificial intelligence-ish. Under that rule there will always be a way to handle problems arising from AI. Without following that rule, pretty much nothing else matters. In the case of the mass-shooting prediction: Good luck proving that it is accurate enough to legitimize punitive consequences for the suspected perpetrator, and without violating a half-decent democratic constitution. Some of the predictors are in themselves crimes, and picking those up would justify correctional measures, no problem. Schools and police also need to get a lot better at picking up and handling mental health problems, be it through algorithms or smarter laws and procedures. Make it easier for the mentally ill to get medical help than to buy an assault rifle, for starters.
- skohan 8y ago
  Why do you insist that facial recognition technology could not be regulated? It's possible for anyone to go to a garden supply store and find the ingredients to make home made explosives. In principal we can't stop people from doing so, but that doesn't mean we can't make it illegal and impose penalties on anyone who does it. It's a technological reality that I can watch videos on my phone while driving, but we can still make it illegal to do so.
  - bayesian_horse 8y ago
    Making bombs in your backyards is much less popular and much more consequential than recognizing images. And image recognition happens completely in silicon. There is no way to know from the outside if a device is recognizing patterns in image data, much less differentiate between recognizing faces or dog breeds.
    - skohan 8y ago
      I'm not following your argument. There's also no way of knowing from the outside if a device is storing identifying information to be shared with advertisers or foreign governments. What's special about classifiers which makes them unregulatable?
      - bayesian_horse 8y ago
        I think you are arguing, to try an analogy, that prohibiting any kind of face recognition would be as worthwhile as prohibiting bomb building, while being about equally difficult. I'm rejecting that premise. And of course, storage and utilization of personal data is a lot easier to regulate in a way that can be enforced. That's the main premise of the GDPR legislation, by the way. Of course, criminal and negligent businesses can totally abuse your personal data, often without consequences. But they can't do so legitimately and they don't have access to the legal and open economy. Even Facebook and Google are feeling the pain already. They do violate some laws some of the time and they have a hard time hiding that, because hiding it requires criminal intent itself, and it's really hard for hundreds of people to keep such secrets, regardless how profitable.
- raxxorrax 8y ago
  > It's impossible to legislate away devices who can do that. It is actually be pretty trivial to legislate restriction of these devices. We do that all the time. edit: Especially on planes where shampoo-technology is pretty much outlawed...
  - thwythwy 8y ago
    It's trivial for the tech not to follow the law. Show me the machine specs and I'll believe you. Otherwise, track record on compliance is abysmal.
CaptainZapp 8y ago
Add to that that the convenience argument is pretty much bullshit. At least when you apply it to travels in the Schengen countries. On such flights you essentially (can) check in remotely, receive the QR code via mail and essentially badge yourself into the plane all the way without ever showing ID. I can't fathom that scanning my face is any faster than that.
- vinay427 8y ago
  This isn't quite accurate for some parts of the Schengen area. Everyone on my flight was required to go through ID controlled when flying out of "Switzerland" [1] to an EU Schengen area country, though I'm very aware that CH is usually an oddball in these matters. [1] The Basel airport is technically in France, though the Swiss part of the airport which I entered is treated like it's in Switzerland and not in the EU, as far as I understand.
  - CaptainZapp 8y ago
    I took dozens of flghts out of and into ZRH and was checked exactly twice. Both times upon arrival at the gate: Once upon arrival in Prague and once in Zurich upon arrival from Athens. In all other cases the whole process, at least in ZRH, was completely automated. It does depend on country and airport, though. Customs is a whole different kettle of fish. But as for id, I'm very rarely asked.
    - vinay427 8y ago
      Hmm, this may be unique to the Basel airport, and perhaps the airline I was flying on. The gate was actually past a physical barrier with booths for checking IDs, similar to the Schengen borders in airports.

by skohan 8y ago

bayesian_horse 8y ago

So it's still not the problem that a machine can recognize you, but rather that somebody is tracking and storing such recognitions. Which is an entirely different thing. The former is a technological reality. It's impossible to legislate away devices who can do that. Tracking and storing personal data on the other hand, is very much within the reach of regulation. The GDPR for example restricts this. Even if somebody has devices who can recognize you is not allowed to store and track data about such recognition incidences without your approval. In this particular case, they don't need your approval because the camera is used in lieu of a boarding pass, and the airline has every right and duty to know you are boarding right now. In case you are worried that the airline connects your travel patterns with other companies interested in other patterns, well, that works with a boarding pass or even a simple check mark on a pen-on-paper boarding list just fine.

reallymental 8y ago
The machine recognizing you was the harder problem. That's the "Minority Report" stuff that people thought was in the _far far_ future that we don't have to worry about constantly. Storing and using that model across various devices is trivial. So if we've done the hard part, the easy part is _definitely_ going to be done. There's no doubt about that. States, companies that can convince the state that they somehow need the data (cough insurance companies) will, I repeat, WILL utilize it. And now, suddenly your premiums for this month have gone up cause in that ski trip you took last weekend, a camera noticed the lack of headgear for about 5 mins. That's the kind of risk the insurance company cannot take. Pay up.
- bayesian_horse 8y ago
  "Minority Report" was about something that Humans couldn't really do, meaning to look into the future, and they even needed psychics to do it... Good luck trying to advance science by avoiding to invent anything that can recognize image!
  - skohan 8y ago
    Taken metaphorically we’re not far away. Imagine if we come to trust ML to make decisions for us in everyday life, and we build a system which can predict with high confidence that someone will commit a mass shooting. Is it so hard to believe that might be acted on?
    - bayesian_horse 8y ago
      To avoid these issues it is imperative to hold natural and legal persons responsible for the software they use, just as it has always been, regardless if it is using anything artificial intelligence-ish. Under that rule there will always be a way to handle problems arising from AI. Without following that rule, pretty much nothing else matters. In the case of the mass-shooting prediction: Good luck proving that it is accurate enough to legitimize punitive consequences for the suspected perpetrator, and without violating a half-decent democratic constitution. Some of the predictors are in themselves crimes, and picking those up would justify correctional measures, no problem. Schools and police also need to get a lot better at picking up and handling mental health problems, be it through algorithms or smarter laws and procedures. Make it easier for the mentally ill to get medical help than to buy an assault rifle, for starters.
skohan 8y ago
Why do you insist that facial recognition technology could not be regulated? It's possible for anyone to go to a garden supply store and find the ingredients to make home made explosives. In principal we can't stop people from doing so, but that doesn't mean we can't make it illegal and impose penalties on anyone who does it. It's a technological reality that I can watch videos on my phone while driving, but we can still make it illegal to do so.
- bayesian_horse 8y ago
  Making bombs in your backyards is much less popular and much more consequential than recognizing images. And image recognition happens completely in silicon. There is no way to know from the outside if a device is recognizing patterns in image data, much less differentiate between recognizing faces or dog breeds.
  - skohan 8y ago
    I'm not following your argument. There's also no way of knowing from the outside if a device is storing identifying information to be shared with advertisers or foreign governments. What's special about classifiers which makes them unregulatable?
    - bayesian_horse 8y ago
      I think you are arguing, to try an analogy, that prohibiting any kind of face recognition would be as worthwhile as prohibiting bomb building, while being about equally difficult. I'm rejecting that premise. And of course, storage and utilization of personal data is a lot easier to regulate in a way that can be enforced. That's the main premise of the GDPR legislation, by the way. Of course, criminal and negligent businesses can totally abuse your personal data, often without consequences. But they can't do so legitimately and they don't have access to the legal and open economy. Even Facebook and Google are feeling the pain already. They do violate some laws some of the time and they have a hard time hiding that, because hiding it requires criminal intent itself, and it's really hard for hundreds of people to keep such secrets, regardless how profitable.
raxxorrax 8y ago
> It's impossible to legislate away devices who can do that. It is actually be pretty trivial to legislate restriction of these devices. We do that all the time. edit: Especially on planes where shampoo-technology is pretty much outlawed...
- thwythwy 8y ago
  It's trivial for the tech not to follow the law. Show me the machine specs and I'll believe you. Otherwise, track record on compliance is abysmal.

CaptainZapp 8y ago

Add to that that the convenience argument is pretty much bullshit. At least when you apply it to travels in the Schengen countries. On such flights you essentially (can) check in remotely, receive the QR code via mail and essentially badge yourself into the plane all the way without ever showing ID. I can't fathom that scanning my face is any faster than that.

vinay427 8y ago
This isn't quite accurate for some parts of the Schengen area. Everyone on my flight was required to go through ID controlled when flying out of "Switzerland" [1] to an EU Schengen area country, though I'm very aware that CH is usually an oddball in these matters. [1] The Basel airport is technically in France, though the Swiss part of the airport which I entered is treated like it's in Switzerland and not in the EU, as far as I understand.
- CaptainZapp 8y ago
  I took dozens of flghts out of and into ZRH and was checked exactly twice. Both times upon arrival at the gate: Once upon arrival in Prague and once in Zurich upon arrival from Athens. In all other cases the whole process, at least in ZRH, was completely automated. It does depend on country and airport, though. Customs is a whole different kettle of fish. But as for id, I'm very rarely asked.
  - vinay427 8y ago
    Hmm, this may be unique to the Basel airport, and perhaps the airline I was flying on. The gate was actually past a physical barrier with booths for checking IDs, similar to the Schengen borders in airports.