If you don't feed the AI information on race at all, then it comes up with results correlating race and the crime data you fed it. Then the result cannot be predicated on race, only crime. It is impossible for the result of this calculation to be racist.

It amazes me that "overpolicing" poor areas or responding to rich areas more are both racist according to these people. It has to be anything other than reality.

Wasn't there a sci-fi that predicted people doing crimes and in it they 100% took it as legitimate evidence that a crime was being committed - and then how hard would it be to corrupt it like voting machines?

It used Chicago, Atlanta, Austin, Detroit, Los Angeles, Philadelphia, Portland, and San Francisco...come on, that's fucking easy mode.

Truth. It perpetuates truth.

Dat AI be raaaiiisssiiisst n' sheeeiiiitt!

Lol. Racial bias?

Is that how they say "accurately"?

SO you feed the criminal records of criminals into this thing and release them and it "predicts" they will do more crime? This is not prediction anymore than releasing a snake into some ones bed and predicting they will get bitten.