I was lucky enough to have this witty insight from a college friend pop up in my Facebook feed recently –
“Sometimes I wonder what is going on in the lives of the people Facebook filters out of my news feed. Then I remember the answer: “proportionally too much”.”
I say lucky, because in reality, we only get too see a small proportion of the statuses, photos and links our Facebook friends post – Facebook recently revealed that the average post is only seen by 12% of our friends. There’s only so much news feed space to go around, and Facebook’s not going to waste the time of that lady you met on holidays three years ago with an Instagram of your omelette, delicious as it may be. Facebook’s EdgeRank algorithm knows that you guys haven’t really been that close since that boozy night in Berlin, so it prioritises the posts that she’s more likely to interact with.
Sounds efficient, right? It is. By looking at our past behaviour; the links we’ve clicked, the statuses we’ve liked, the photos we’ve commented on, Facebook can predict with impressive accuracy the content that will appeal to us.
However, there’s a dark side to all this personalisation. When algorithms determine our information diet based on our previous behaviour, we can become isolated from viewpoints and ideas that challenge our biases: Facebook knows my political views, so I won’t see an article posted by a friend who’s at the opposite end of the spectrum: EdgeRank has determined that I’m unlikely to interact with an opinion piece with which I disagree, so it prioritises a story that reinforces my beliefs instead.
Understanding different points of view is at the heart of compromise, and compromise is what makes a modern democratic society possible. When we aren’t exposed to alternative points of view, our opinions harden, and our tolerance for difference decreases. Division and closed-mindedness make for an unhappy, conflicted society.
This isn’t just a Facebook issue. Google does it too: Eli Pariser, who coined the term filter bubble, compared the Google results two friends were served for the keyword “Egypt” while the Arab Spring was at its most febrile. The results speak for themselves.
The articles NewsWhip tracks are categorised, but not personalised. So we see what’s trending among everyone – not the subset represented by those we follow on Twitter (wits, future of news experts and interesting startups). The results side by side sometimes seem schizophrenic when compared with the fairly narrow editorial line of standard news outlets. Check out two of January’s biggest social stories:
More People Killed With Hammers and Clubs Each Year Than Rifles – BreitbartÂ
Five People Shot at Three Different Gun Shows on Gun Appreciation Day – Think ProgressÂ
These are stories that a huge amount of people around the world believe are important, but the people sharing one are unlikely to have seen the other thanks to personalisation algorithms. Gun control is already an extremely polarised issue in the US. Algorithmically separating one camp’s news from the other is one way to make sure compromise never happens.
Whatever your deeply-held beliefs may be, there’s never harm in getting to know the opposing point of view. If you see a huge social story – even if the headline doesn’t appeal to your political leanings – why not click through and see what the people outside your filter bubble are interested in?
