4 ms·

Can someone explain the legal theory that diffentiates using something like Dataminr to monitor tweets vs simply having an army of agents manually searching for

by tqi 3y ago

Can someone explain the legal theory that diffentiates using something like Dataminr to monitor tweets vs simply having an army of agents manually searching for them? I agree that it feels like the scale of the former makes it different, but is there a principled reason why that is the case?

theGnuMe 3y ago
I don't know, but Congress passed legislation to prevent the digitization of gun records. So they have massive paper files and a small set of agents to match guns to crimes. You would think with AI we could track all guns and bullets to crimes eventually.
- hiatus 3y ago
  > You would think with AI we could track all guns and bullets to crimes eventually. Not sure how that would be achieved since you can make your own guns and bullets in the majority of states.
  - one_level_deep 3y ago
    [dead]
- giantg2 3y ago
  "Congress passed legislation to prevent the digitization of gun records." Funny thing about that which could also apply to data/AI/etc, is that I understand the ATF is ignoring that law anyways.
snarf21 3y ago
I think the contrast is supposed to be the difference between always monitoring people who post about protesting FOR X and/or protesting AGAINST X versus searching for posts of people who were near some incident related to some protesting or some crime. It is similar to the geo-fence search warrants that were recently discussed here. tl;dr: I see it as the difference between using search tools to monitor for potential suspects vs search tools to implicate already identified suspects.
advisedwang 3y ago
The legal choices and mechanisms we have are based on built in assumptions about how easy and expensive something is, and this is especially true for monitoring of public spaces. The decision to say that stuff you say/do in public can be monitored without oversight was made when that would be fairly rare accidentally or require a lot of resources to deliberately use. If we suddenly find ourselves in a world where every single public moment can be recorded and surveilled, the risk to rights is much large for the same state interest. So with the balance shifted, the legal decision could shift. The law (thankfully) is not some hard coded machine that follows cold logic ignoring all context. (Good) judges are able to look at changes in context and recognize that the underlying factors have shifted and so change rulings.
- bashinator 3y ago
  Sufficient quantitative change in capability causes a qualitative difference to outcomes.
rtkwe 3y ago
My reasoning is the latter requires significant resources thus agencies would exercise discretion in what they monitor. With automated systems they can monitor larger swaths of people without much effort at all.
hwillis 3y ago
https://en.wikipedia.org/wiki/Expectation_of_privacy_(United_States) https://en.wikipedia.org/wiki/Expectation_of_privacy_(United... You are entitled to privacy even for certain public situations. Even if you can see through a window, the government can't go training telescopes inside your home without a warrant. It's an active battle for things like location data.

by tqi 3y ago

theGnuMe 3y ago

I don't know, but Congress passed legislation to prevent the digitization of gun records. So they have massive paper files and a small set of agents to match guns to crimes. You would think with AI we could track all guns and bullets to crimes eventually.

hiatus 3y ago
> You would think with AI we could track all guns and bullets to crimes eventually. Not sure how that would be achieved since you can make your own guns and bullets in the majority of states.
- one_level_deep 3y ago
  [dead]
giantg2 3y ago
"Congress passed legislation to prevent the digitization of gun records." Funny thing about that which could also apply to data/AI/etc, is that I understand the ATF is ignoring that law anyways.

snarf21 3y ago

I think the contrast is supposed to be the difference between always monitoring people who post about protesting FOR X and/or protesting AGAINST X versus searching for posts of people who were near some incident related to some protesting or some crime. It is similar to the geo-fence search warrants that were recently discussed here. tl;dr: I see it as the difference between using search tools to monitor for potential suspects vs search tools to implicate already identified suspects.

advisedwang 3y ago

The legal choices and mechanisms we have are based on built in assumptions about how easy and expensive something is, and this is especially true for monitoring of public spaces. The decision to say that stuff you say/do in public can be monitored without oversight was made when that would be fairly rare accidentally or require a lot of resources to deliberately use. If we suddenly find ourselves in a world where every single public moment can be recorded and surveilled, the risk to rights is much large for the same state interest. So with the balance shifted, the legal decision could shift. The law (thankfully) is not some hard coded machine that follows cold logic ignoring all context. (Good) judges are able to look at changes in context and recognize that the underlying factors have shifted and so change rulings.

bashinator 3y ago
Sufficient quantitative change in capability causes a qualitative difference to outcomes.

rtkwe 3y ago

My reasoning is the latter requires significant resources thus agencies would exercise discretion in what they monitor. With automated systems they can monitor larger swaths of people without much effort at all.

hwillis 3y ago

https://en.wikipedia.org/wiki/Expectation_of_privacy_(United_States) https://en.wikipedia.org/wiki/Expectation_of_privacy_(United... You are entitled to privacy even for certain public situations. Even if you can see through a window, the government can't go training telescopes inside your home without a warrant. It's an active battle for things like location data.