Hacker News Reader

Comment by seanmcdirmid | Hacker News Reader

How? A cop is going to spend extra attention on minorities driving through white neighborhoods, a camera doesn't care.

There are initial decisions to be made with the deployment of these automated policing tools. For example, this is from an EFF piece on predictive policiing[0]:

"For instance, if police put majority Black neighborhoods under intense police scrutiny and surveillance, they will be more likely than in other neighborhoods to make stops and arrests according to minor quality of life infractions. Therefore, derivative maps purporting to show where future crimes might be committed will disproportionately weigh those neighborhoods already living under the weight of intense police presence. This can create a self-fulfilling prophecy that uses the supposed objectivity of math to legitimize racially biased police procedures."

There is also a piece on other issues with license plate readers in particular.[1] I'd recommend the whole series if you're interested.[2]

[0]: https://sls.eff.org/technologies/predictive-policing

[1]: https://sls.eff.org/technologies/automated-license-plate-rea...

[2]: https://sls.eff.org/

The solution is easy in this case: just put them everywhere. The technology means that this scales, although where I live it is the richer cities that have deployed it (so Medina Washington has license plate readers all over...not a place you could call poor). But I can see how this is a "damned if we do, damned if we don't" kind of situation, after all poor neighborhoods complain that they don't have enough police coverage at the same time that they complain they have too much.

I think this is just people not liking police in general. Here in Seattle, we complain about not enough police coverage in the north because they mostly deployed in the south, given that this is where much more violent crime is (while we mostly have just property crime). If you station more policing resources where more of the problems are, you are going to get this no matter what (and you get accusations of bias if you and if you don't).

Some quick options come to mind:

  - unjust distribution of the cameras themselves

  - camera data is used for selective enforcement (alert a nearby office when minorities are driving through white neighborhoods)

Tools are only as unbiased as the people deploying and using them.

Those both involve biases from people in the loop, not the devices themselves.

The models are trained on data, if that data is inherently biased, the model will be too.

Reply on news.ycombinator.com