4 ms·

This is dumb. Predictive policing algorithms predict where the police will be, not where crime takes place: as crime data is simply a measurement of policing ac

This is dumb. Predictive policing algorithms predict where the police will be, not where crime takes place: as crime data is simply a measurement of policing activity. If there is bias in policing, it would therefore be amplified. Ban police from using this technology.

mc32 6y ago
As departments get fewer funds and simultaneously try to keep a lid on crime, they will resort to more of these “more bang for the buck” strategies.
- ideashower 6y ago
  Indeed. Especially as companies like Palantir serve the tools on a silver platter.
- Ensorceled 6y ago
  Fewer funds? Metro Toronto policing budget went up 8% last year.
creddit 6y ago
What if the model data is citizen reported crimes?
ghostpepper 6y ago
This is my first thought as well. If police go to a place, there is more likely to be a recorded instance of a crime investigated in that place, and thus the algorithm will be more likely to send the police there in the future. The crime map for Vancouver is interesting - I'm glad to see that the tool reported in the article is at least exposed to the public. https://geodash.vpd.ca/ https://geodash.vpd.ca/
- at_a_remove 6y ago
  In some sense, if you did not send police anywhere at all, you would have no crime to report. The solution to end crime, then, is to never send the police anywhere.
  - pbhjpbhj 6y ago
    Ah, the Trump method!
  - ramshorns 6y ago
    I'm not sure if you're kidding. Of course police presence leads to escalation and more crime, and many places would be much safer without police.
    - google234123 6y ago
      This is not a good position. What you would get is vigilantism instead as people take matters into their own hands.
    - lawnchair_larry 6y ago
      Demonstrably false.
    - logarhythmic 6y ago
      ... What? This is not at all obvious to me. How exactly are you arriving at this conclusion?
- vkou 6y ago
  In depends on the type of crime. A break-and-enter, or a theft from/of a car, or vandalism against property, for example, are going to be recorded regardless of whether there's a police officer within 5 miles of where they occur. (Because people file a police report to get an insurance claim.) The distribution of public nuisance arrests, or drug-related crimes, or people jumping the turnstiles at the subway, or parking tickets, or speeding tickets, or DUIs, on the other hand, are almost entirely going to be closely correlated with which areas you deploy police to.
almost_usual 6y ago
> Ban police from using this technology. Ban x from using this technology. This is hard to do when the technology isn’t bound to its physical matter (software or algorithms). The technology will go everywhere eventually because it’s easier to distribute than ban.
- JamisonM 6y ago
  What? You mean police forces will be banned so individual officers will pay for access?
  - almost_usual 6y ago
    Why would you have to pay? It’s an algorithm, run some data through it and you’re now more efficient than your fellow officers. You can do more work in less time. Sounds crazy now but I wouldn’t be surprised if it was normal in the future. You can’t stop software or encryption.
pbhjpbhj 6y ago
>crime data is simply a measurement of policing activity // Are you thinking of a specific jurisdiction's specific statistical set here, because that's not true in general. In the UK I know we have perception surveys of the public that ask 'have you been subject to crime' and those statistics are not measures of policing as many crimes reported in the survey won't have been reported to police. Example for me: someone scratched my car body work ("keyed" it), I didn't report it as I wasn't going to claim on insurance and knew the car park owners had reported similar occurrences in the recent past (I spoke to them: 'yeah, happens a lot here'). Crime survey would show a crime against my property; police statistics wouldn't.

mc32 6y ago

As departments get fewer funds and simultaneously try to keep a lid on crime, they will resort to more of these “more bang for the buck” strategies.

ideashower 6y ago
Indeed. Especially as companies like Palantir serve the tools on a silver platter.
Ensorceled 6y ago
Fewer funds? Metro Toronto policing budget went up 8% last year.

creddit 6y ago

What if the model data is citizen reported crimes?

ghostpepper 6y ago

This is my first thought as well. If police go to a place, there is more likely to be a recorded instance of a crime investigated in that place, and thus the algorithm will be more likely to send the police there in the future. The crime map for Vancouver is interesting - I'm glad to see that the tool reported in the article is at least exposed to the public. https://geodash.vpd.ca/ https://geodash.vpd.ca/

at_a_remove 6y ago
In some sense, if you did not send police anywhere at all, you would have no crime to report. The solution to end crime, then, is to never send the police anywhere.
- pbhjpbhj 6y ago
  Ah, the Trump method!
- ramshorns 6y ago
  I'm not sure if you're kidding. Of course police presence leads to escalation and more crime, and many places would be much safer without police.
  - google234123 6y ago
    This is not a good position. What you would get is vigilantism instead as people take matters into their own hands.
  - lawnchair_larry 6y ago
    Demonstrably false.
  - logarhythmic 6y ago
    ... What? This is not at all obvious to me. How exactly are you arriving at this conclusion?
vkou 6y ago
In depends on the type of crime. A break-and-enter, or a theft from/of a car, or vandalism against property, for example, are going to be recorded regardless of whether there's a police officer within 5 miles of where they occur. (Because people file a police report to get an insurance claim.) The distribution of public nuisance arrests, or drug-related crimes, or people jumping the turnstiles at the subway, or parking tickets, or speeding tickets, or DUIs, on the other hand, are almost entirely going to be closely correlated with which areas you deploy police to.

almost_usual 6y ago

> Ban police from using this technology. Ban x from using this technology. This is hard to do when the technology isn’t bound to its physical matter (software or algorithms). The technology will go everywhere eventually because it’s easier to distribute than ban.

JamisonM 6y ago
What? You mean police forces will be banned so individual officers will pay for access?
- almost_usual 6y ago
  Why would you have to pay? It’s an algorithm, run some data through it and you’re now more efficient than your fellow officers. You can do more work in less time. Sounds crazy now but I wouldn’t be surprised if it was normal in the future. You can’t stop software or encryption.

pbhjpbhj 6y ago

>crime data is simply a measurement of policing activity // Are you thinking of a specific jurisdiction's specific statistical set here, because that's not true in general. In the UK I know we have perception surveys of the public that ask 'have you been subject to crime' and those statistics are not measures of policing as many crimes reported in the survey won't have been reported to police. Example for me: someone scratched my car body work ("keyed" it), I didn't report it as I wasn't going to claim on insurance and knew the car park owners had reported similar occurrences in the recent past (I spoke to them: 'yeah, happens a lot here'). Crime survey would show a crime against my property; police statistics wouldn't.