A future that never comes

I can gauge my mood by assigning it the name of the science fiction author whose dystopic worldview I think we currently inhabit. The gritty and desolate days when bitter humor is the only reliable source of solace are Vonnegut days. Often in the US, it's clearly Orwell. And increasingly, the days in which the ludicrous becomes real, it's Phillip K Dick's world. 

Score another one for PKD. A firm called PredPol is selling predictive policing models to local law enforcement, suggesting where crimes are most likely to occur so they can move people and resources into the right spots. The title "predictive" sounds like a misnomer; what we're really looking at here is correlation. As I understand it, the predictions are based on computer analysis of crime patterns, which seems like a pretty lo-fi approach. I can think of fascinating ways we could incorporate corporate data to extract more sophisticated correlations. The only cost would be our right to privacy, which is practically a quaint anachronism by now (see: Orwellian days).

How does crime, for example, correlate to television programming? To the location-specific use of social media by identified criminals? To the approach of standardized test dates in the high school, and the flow of the academic calendar? To the usage of netflix and other streaming media? Where does crime take place relative to where stolen credit cards are actually used? How does it correlate to the release dates of specific status-conferring consumer goods? Are crimes related to location-specific patterns of cell phone usage? 

This Forbes article [Server and Protect] talks a bit about the company and its product. Will 2015 be the year that "predictive policing" takes off? There seems to be some difference of opinion on whether this is "proven" technology or not. 

For a while, apparently, one police district in LA was tweeting their predicted locations so citizens could be on the lookout. Of all aspects of this program, I find this application most interesting. On one hand, you might say it's pointless to tip off the criminals. On the other, disrupting patterns of criminal activity may be an effective model for prevention. Like many little disasters, they may be functions of several elements aligning in one location, and removing one element destroys the entire chain. 

Someone call me when the Bevatron comes online.