3 ms·

Is this personally identifiable? I don't see any issue with collecting and selling anonymized observations. It would be like police setting up cameras and usin

by umvi 7y ago

Is this personally identifiable? I don't see any issue with collecting and selling anonymized observations. It would be like police setting up cameras and using them to train a machine learning model on drunk driver detection. It's not collecting who is driving, just observing how normal cars subtly drift in and out of a lane and brake vs. intoxicated drifting/braking and using that to train a DUI detection model.

noxToken 7y ago
The true test is if the anonymized data can be filtered and cross-referenced in a way to still link back to people[0]. [0]:https://www.reuters.com/article/us-health-privacy/anonymous-patient-data-may-not-be-as-private-as-previously-thought-idUSKCN1OK24O https://www.reuters.com/article/us-health-privacy/anonymous-...
- umvi 7y ago
  You can always de-anonymize data given enough resources so that argument seems bogus. Anonymous data: Red car observed on Dawn Road at 23:34 on May 31 2019. DUI predictor - likely sober. Speed - 53 mph. Can this information be de-anonymized? Sure can! Police: "The suspect fled 7-11 Around 23:30 and either went on Dawn Road or Lost Haven Road. Let's filter all red car observation events that happened that night between 23:30 and 23:40 on Dawn and Lost Haven. Yup, found him" So you are saying we have to scrub it further? You end up removing all information until there is nothing left. "car observed on May 31"
  - noxToken 7y ago
    It would depend on what is included in the data and what other data is available for reference. You can have medical data that tracks all of the symptoms reported by a patient without prompt ("What brings you in?"), symptoms reported by a patient with prompt ("Can you describe the pain? Where is it located?"), and the diagnosis by the doctor. It tells you nothing about the patient nor the doctor, but it still provides useful information.
JohnFen 7y ago
> I don't see any issue with collecting and selling anonymized observations. In this day and age when correlating and analyzing data from a wide variety of sources is commonplace, the only effectively "anonymized" data is data that has been discarded.
- umvi 7y ago
  Right so you are basically saying companies/governments should not be able to make generic observations about anything involving people in any way without their consent because it is impossible to have truly anonymous data. I don't know if I agree with that.
  - JohnFen 7y ago
    No, I'm not saying that. It is possible, for instance, to collect individual data, tally up certain characteristics in the aggregate, then discard the individual data points and only keep the aggregate statistics. With certain narrow exceptions, though, no data about me personally should be collected without my explicit permission regardless.
    - umvi 7y ago
      > With certain narrow exceptions, though, no data about me personally should be collected without my explicit permission regardless. Well, I don't agree with that. I don't see any reason people should have a right to "own" information observed about them. I make observations all the time about other people. You can't force me to forget what I've observed. Computers make it easier to "remember" and process observations, true, but at the root there is no difference between me observing stuff and writing it down and a computer observing stuff and writing it down.
      - JohnFen 7y ago
        I suspect your objection may mostly fall into what I consider exceptions, such as things done in public spaces (although I don't think it's sustainable to assert that you give up all privacy rights in public spaces). > at the root there is no difference between me observing stuff and writing it down and a computer observing stuff and writing it down. If that computer isn't talking with other computers, I agree. I actually don't have much of a problem with individuals making individual observations of public behavior and writing them down or storing them in a computer. My concern is more about the parameters around sharing that data (mostly because of the existence of databases and data mining). Further, I'm far more concerned about data collected about me on the internet than in the physical public square. That said, I do and will continue to go out of my way to avoid as much surveillance as possible even in physical public spaces. For instance, any store using those surveillance devices intended to analyze my shopping behavior, moods, etc., in order to target ads at me is a store I won't be stepping into.
xkcd-sucks 7y ago
The example, run anonymously, can be no more than unsupervised clustering. In order to train "a DUI detection model" there must be actual DUIs and BACs recorded, which isn't anonymous.

by umvi 7y ago

noxToken 7y ago

The true test is if the anonymized data can be filtered and cross-referenced in a way to still link back to people[0]. [0]:https://www.reuters.com/article/us-health-privacy/anonymous-patient-data-may-not-be-as-private-as-previously-thought-idUSKCN1OK24O https://www.reuters.com/article/us-health-privacy/anonymous-...

umvi 7y ago
You can always de-anonymize data given enough resources so that argument seems bogus. Anonymous data: Red car observed on Dawn Road at 23:34 on May 31 2019. DUI predictor - likely sober. Speed - 53 mph. Can this information be de-anonymized? Sure can! Police: "The suspect fled 7-11 Around 23:30 and either went on Dawn Road or Lost Haven Road. Let's filter all red car observation events that happened that night between 23:30 and 23:40 on Dawn and Lost Haven. Yup, found him" So you are saying we have to scrub it further? You end up removing all information until there is nothing left. "car observed on May 31"
- noxToken 7y ago
  It would depend on what is included in the data and what other data is available for reference. You can have medical data that tracks all of the symptoms reported by a patient without prompt ("What brings you in?"), symptoms reported by a patient with prompt ("Can you describe the pain? Where is it located?"), and the diagnosis by the doctor. It tells you nothing about the patient nor the doctor, but it still provides useful information.

JohnFen 7y ago

> I don't see any issue with collecting and selling anonymized observations. In this day and age when correlating and analyzing data from a wide variety of sources is commonplace, the only effectively "anonymized" data is data that has been discarded.

umvi 7y ago
Right so you are basically saying companies/governments should not be able to make generic observations about anything involving people in any way without their consent because it is impossible to have truly anonymous data. I don't know if I agree with that.
- JohnFen 7y ago
  No, I'm not saying that. It is possible, for instance, to collect individual data, tally up certain characteristics in the aggregate, then discard the individual data points and only keep the aggregate statistics. With certain narrow exceptions, though, no data about me personally should be collected without my explicit permission regardless.
  - umvi 7y ago
    > With certain narrow exceptions, though, no data about me personally should be collected without my explicit permission regardless. Well, I don't agree with that. I don't see any reason people should have a right to "own" information observed about them. I make observations all the time about other people. You can't force me to forget what I've observed. Computers make it easier to "remember" and process observations, true, but at the root there is no difference between me observing stuff and writing it down and a computer observing stuff and writing it down.
    - JohnFen 7y ago
      I suspect your objection may mostly fall into what I consider exceptions, such as things done in public spaces (although I don't think it's sustainable to assert that you give up all privacy rights in public spaces). > at the root there is no difference between me observing stuff and writing it down and a computer observing stuff and writing it down. If that computer isn't talking with other computers, I agree. I actually don't have much of a problem with individuals making individual observations of public behavior and writing them down or storing them in a computer. My concern is more about the parameters around sharing that data (mostly because of the existence of databases and data mining). Further, I'm far more concerned about data collected about me on the internet than in the physical public square. That said, I do and will continue to go out of my way to avoid as much surveillance as possible even in physical public spaces. For instance, any store using those surveillance devices intended to analyze my shopping behavior, moods, etc., in order to target ads at me is a store I won't be stepping into.

xkcd-sucks 7y ago

The example, run anonymously, can be no more than unsupervised clustering. In order to train "a DUI detection model" there must be actual DUIs and BACs recorded, which isn't anonymous.