Can Computers Predict Crimes That Haven’t Happened Yet?
The ethics of big data in crime forecasting
The Inquiry · BBC World Service
Audio is streamed directly from the publisher (open.live.bbc.co.uk) as published in their RSS feed. Play Podcasts does not host this file. Rights-holders can request removal through the copyright & takedown page.
Show Notes
Chicago resident Robert McDaniel was surprised when a police commander showed up at his home to warn him that they were watching him. With only a misdemeanour conviction and arrests for a number of suspected minor offenses, he had somehow made it onto the Chicago Police Department’s so called ‘heat list’ - a list of names created by algorithm of those deemed to be most at risk of either being a victim or perpetrator of violent crime.
In this Inquiry we look at whether computers can predict future of when, where and by whom crimes will be committed. Can analysing ‘big data’ help target scarce resources in more intelligent ways? Or are the algorithms exacerbating the already heightened tensions between police and the public? How effective are some of the ‘predictive policing’ systems already in use? The inner workings of many of these programmes are protected by private copyright laws too so how can you challenge the decision made by a secret algorithm?
(Photo:Chicago Police officers standing next to a police car and a taped off crime scene. Credit: Joshua Lott/AFP/Getty Images)