Michal Kosinski – an Assistant Professor at the Stanford Graduate School of Business – discussed this subject during a session at the 2017 Cannes Lions International Festival of Creativity.
“Going forward in the long term – and when I say ‘long term’, I mean five years, maybe ten years – basically there's going to be no privacy left,” he said. (For more details, read WARC’s exclusive report: Politics and data: A double-edged digital sword.)
“For a motivated individual, for a motivated foreign government, or an institution, there will be very little to stop them from learning about you whatever they want to learn – even if you have never mentioned it to anyone.”
Kosinski based these assertions, in part, on a series of projects he has worked on that mixed intelligent digital tools with basic Facebook information – say, an individual’s age, gender, location, “likes”, sharing history and a few self-perceptions.
His experiments have, among other things, proved that sophisticated systems crunching meaningful amounts of data can predict a person’s sexual orientation, skin colour, political affiliation, level of extroversion with extremely high accuracy.
Using the personality self-ratings of 86,220 people, one automated system effectively knew more about an individual than a friend or roommate did with just 70 “likes” to work with – a total hitting 150 for family members and 300 for their spouse.
Coupled with the rise of everything from geo-location services and fitness trackers to smart homes, the analytic tools on offer will soon be able generate vast knowledge about individuals.
“The message is that now you can have companies, politicians, marketers, PR agencies [and] governments using seemingly innocent pieces of data – like the books you've read or the movies you have watched or what you like on Facebook – to reveal information about yourself that you may prefer to keep private,” said Kosinski.
Data sourced from WARC