5 ms·

I would think that the idea here is that courts aren't currently intentionally giving out "and everyone remembers what they did forever" as part of most crimina

by derefr 3y ago

I would think that the idea here is that courts aren't currently intentionally giving out "and everyone remembers what they did forever" as part of most criminal sentences; that this type of indefinite scarlet lettering even after serving out a sentence was never an intent of criminal sentencing, but only became one as a byproduct of the judicial process due to court cases being public for auditability + the Internet durably remembering and indexing everything forever; and that if courts want to make an indefinite scarlet letter a part of (some) criminal sentences, then they should be doing so explicitly, as with "sex offender registries" and the like, so that that power can be applied only where democratically favored via legislation.

AnthonyMouse 3y ago
> this type of indefinite scarlet lettering even after serving out a sentence was never an intent of criminal sentencing Most people used to live in small towns and die within 100 miles of where they were born. If you got convicted of something, everybody knew.
- derefr 3y ago
  Most people, yes. But criminals in particular used to escape this bubble of knowledge about who they were, by simply leaving town and starting over somewhere else. The possibility of doing this is one of the implicit bedrock assumptions of the design of most legal systems — but, due to everyone being able to look up "who you are and what you've done" everywhere you go, it's currently a broken assumption.
  - AnthonyMouse 3y ago
    You could just change your name, which seems like less of a burden than moving away from everything you've ever known. If you want to replicate the way it used to be then instead of trying to censor the world, just have the government give you a new ID and social security number when you change your name without publishing any record associating the new one with the old one.
    - derefr 3y ago
      Doesn't work when private companies keep facial-recognition databases and use them to auto-tag new photos of you with your old name.
      - AnthonyMouse 3y ago
        Facial recognition vendors: Our systems have 99.97% accuracy! Actual meaning of this: In a nation of 300M, there are 90,000 false positives for every true match.
        mlyle 3y ago
        This is wrong and not how the numbers are computed. It's not pairwise. Pairwise error rates are more like 1 in 10 billion. NIST did a review in 2019 (published 2020) with datasets containing 12 million people. For searching a given photo against the database, solutions might, say, fail to pull anyone 3% of the time and pull the wrong person 0.5% of the time. Thus, you get about 1 false positive for every ~190 true matches. It was also under pessimistic conditions: a big fraction of the searches were for individual not enrolled in the dataset. https://github.com/usnistgov/frvt/blob/nist-pages/reports/1N/frvt_1N_report_2020_03_27.pdf https://github.com/usnistgov/frvt/blob/nist-pages/reports/1N... edit: I see there are newer versions of the data since I bookmarked this, but I have not reviewed them. A glance shows they format the data differently. Still, things have likely improved somewhat since then.
        AnthonyMouse 3y ago
        > This is wrong and not how the numbers are computed. It's not pairwise. I thought that actually was the pairwise number for some algorithms in use. They're designed to produce a list of suspects to investigate, and then if you get a dozen hits against your surveillance camera photo from your database of a few thousand local mugshots, that's what you're after. > NIST did a review in 2019 (published 2020) with datasets containing 12 million people. For searching a given photo against the database, solutions might, say, fail to pull anyone 3% of the time and pull the wrong person 0.5% of the time. Doing the numbers this way implies that the error rate would be ~2800% higher if the database contained the entire population, even under these conditions. > It was also under pessimistic conditions: a big fraction of the searches were for individual not enrolled in the dataset. It was also under optimistic conditions: They had a database of profile photos taken under controlled conditions with an attendant present. This is obviously not available for most people. It's using the most sophisticated algorithms in existence under laboratory conditions rather than describing what happens in practice in most cases. But even supposing that this is a problem, wouldn't it still be better to ban facial recognition databases than news reporting?

by derefr 3y ago

AnthonyMouse 3y ago

> this type of indefinite scarlet lettering even after serving out a sentence was never an intent of criminal sentencing Most people used to live in small towns and die within 100 miles of where they were born. If you got convicted of something, everybody knew.

derefr 3y ago
Most people, yes. But criminals in particular used to escape this bubble of knowledge about who they were, by simply leaving town and starting over somewhere else. The possibility of doing this is one of the implicit bedrock assumptions of the design of most legal systems — but, due to everyone being able to look up "who you are and what you've done" everywhere you go, it's currently a broken assumption.
- AnthonyMouse 3y ago
  You could just change your name, which seems like less of a burden than moving away from everything you've ever known. If you want to replicate the way it used to be then instead of trying to censor the world, just have the government give you a new ID and social security number when you change your name without publishing any record associating the new one with the old one.
  - derefr 3y ago
    Doesn't work when private companies keep facial-recognition databases and use them to auto-tag new photos of you with your old name.
    - AnthonyMouse 3y ago
      Facial recognition vendors: Our systems have 99.97% accuracy! Actual meaning of this: In a nation of 300M, there are 90,000 false positives for every true match.
      - mlyle 3y ago
        This is wrong and not how the numbers are computed. It's not pairwise. Pairwise error rates are more like 1 in 10 billion. NIST did a review in 2019 (published 2020) with datasets containing 12 million people. For searching a given photo against the database, solutions might, say, fail to pull anyone 3% of the time and pull the wrong person 0.5% of the time. Thus, you get about 1 false positive for every ~190 true matches. It was also under pessimistic conditions: a big fraction of the searches were for individual not enrolled in the dataset. https://github.com/usnistgov/frvt/blob/nist-pages/reports/1N/frvt_1N_report_2020_03_27.pdf https://github.com/usnistgov/frvt/blob/nist-pages/reports/1N... edit: I see there are newer versions of the data since I bookmarked this, but I have not reviewed them. A glance shows they format the data differently. Still, things have likely improved somewhat since then.
        AnthonyMouse 3y ago
        > This is wrong and not how the numbers are computed. It's not pairwise. I thought that actually was the pairwise number for some algorithms in use. They're designed to produce a list of suspects to investigate, and then if you get a dozen hits against your surveillance camera photo from your database of a few thousand local mugshots, that's what you're after. > NIST did a review in 2019 (published 2020) with datasets containing 12 million people. For searching a given photo against the database, solutions might, say, fail to pull anyone 3% of the time and pull the wrong person 0.5% of the time. Doing the numbers this way implies that the error rate would be ~2800% higher if the database contained the entire population, even under these conditions. > It was also under pessimistic conditions: a big fraction of the searches were for individual not enrolled in the dataset. It was also under optimistic conditions: They had a database of profile photos taken under controlled conditions with an attendant present. This is obviously not available for most people. It's using the most sophisticated algorithms in existence under laboratory conditions rather than describing what happens in practice in most cases. But even supposing that this is a problem, wouldn't it still be better to ban facial recognition databases than news reporting?