Saturday, April 27, 2013

Profiling

This PNAS piece presents evidence based on a 58,000 person sample that based on your trail of "Facebook Likes" that Big Data Nerds can quickly (and correctly) determine your sexual orientation, your race and your political orientation.   This suggests a certain "internal consistency" to the choices we make.  Dick Cheney apparently doesn't click on Prius Ads.    This evidence of "profiling" is good news for advertisers seeking to reach certain markets.   I have been working on this broad topic for a long time in my work on environmentalists being "consistent" in their voting in public markets and their "voting their pocket book" in private markets.   Is this PNAS project a foreshadowing of Big Brother?   This depends on how concerned we are about future scenarios where subgroups of the population are easily identified.  While we celebrate "targeted advertising", we are also scandalized by "targeted oppression".  So, do we want transaction costs of identifying subsets of individuals to be high or low?   When do you value your privacy versus when do you want to let it "all hang out"?

No comments :