Latest

Facebook Data Knows What You Like Even Better Than You Do

From The Australian News Service (source link):  “SCROLLING through Facebook, it’s become second nature to casually “like” interesting news articles, photos of friends’ babies and funny status updates.  We don’t have to interact with the post, or absorb its content; we may forget we liked it seconds later.

But Facebook will remember — and those “likes” are building a startlingly accurate digital image of you.

Your click-happy behaviour can determine your race with 95 per cent accuracy, your sexuality with 88 per cent accuracy and whether you use drugs with 65 per cent accuracy, Quartz reported.

Not only do our likes reveal the obvious — our politics, interests and friendship connections — they can reveal more subtle information, including our IQ.  You may never have put it in a status update, but Facebook can even tell with 60 per cent certitude if your parents divorced before you were 21, says Christian Rudder, OkCupid founder and author of Dataclysm: Who We Are When We Think No One’s Looking.

The data gathered by “likes” is in many ways superior to information collected using traditional research methods.  First, it is gathered on an unprecedented scale. “Big data” is likely to be much more accurate than small-scale experiments and tests — and there had been as many as 1.13 trillion likes on Facebook by September 2013.

Secondly, we barely recognise that we are participating in it. We are simply living our lives.  The very absence of lab conditions or a questionnaire makes the data more truthful and accurate, according to Rudder.

“This data is real life,” he writes. “Online you have friends, lovers, enemies, and intense moments of truth without a thought for who’s watching, because ostensibly no one is — except of course the computers recording it all.”

The problem with many research projects is that people will lie — about their salaries, their attractiveness or how racist they are.
But look at their likes, and the mask slips.  “Digital data circumvents that old research obstacle: people’s inability to be honest when the truth makes them look bad,” says Rudder.

“You could never ask people these days if they like racist jokes and get a real answer. Yet lo and behold the country’s most notorious slur for black people is very popular as a Google Search term; it still appears in a half-million searches a month in the United States … Digital data’s ability to get at the private mind like this is unprecedented and very powerful.

“Your search history tells us what kind of jokes you like. Your Facebook network reveals not just your friendships, but in some cases the state of your marriage. Your preferences on OkCupid tell us what you find sexy, and your reaction to the strangers the site offers up tells us how you judge people.

“You fold in data points like these for millions and millions of people, and you start to get a whole new picture of humankind.”  And this is from just a few years’ worth of likes. Soon, data will be able to accurately show the development of our tastes and preferences as we age, plotting changes in social and economic views over history and over the course of a human life.

In a recent experiment on Medium, a writer stopped liking anything on Facebook for two weeks. She said her feed became more nuanced and real, as she was forced to interact with posts rather than mindlessly clicking her approval.  “The algorithm does not understand the psychological nuances of why you might like one thing and not another even though they have comparatively similar keywords and reach similar audiences,” she wrote.

“It doesn’t understand the many political, philosophical, and emotional shades of a given topic … Liking a post about a sweet wedding does not equal my wanting to see every inspiring human who ever existed in New York. The algorithm can’t know that, though, because it can’t know individuals.”  But is that true? Increasingly, the algorithm will be able to weed out nuances within nuances, just as Netflix’s great power lies in the fact it contains categories as specific as “Emotional Independent Dramas for Hopeless Romantics.”

Terrified? You’re not alone. Millions of Facebook users baulked this summer at news of a controversial large-scale secret experiment on “emotional contagion”, while Rudder provoked outrage too, with an OkCupid blog post entitled “We experiment on human beings.”  Unfortunately, we are complicit. No one is forcing us to be on Facebook, or any particular dating site.

We get the opportunity to share pictures, thoughts and to make connections. In return, we are invited to like things. No one is forcing us to like something, but once we do, we don’t own that like any more.  Yet the action is incredibly hard to resist. Adam Powers, inventor of a Chrome extension called “Neutralike”, says notifications are “the primary source of what I think of as dark dopamine, or reward centre activation resulting from dark design patterns.”

We are unlikely to give up liking things any time soon. The non-intimate interaction appeals to our lazy, emotionally detached online selves — plus we crave the external validation that comes from getting likes ourselves.  And it would be almost impossible for many of us to locate and remove all our likes from the past.  Our digital footprint is permanent. Even if we leave Facebook, we leave part of ourselves behind.

Royce Christyn
About Royce Christyn (3467 Articles)
Documentarian, Writer, Producer, Director, Author.