The Rise of Filter Bubbles
eldavojohn writes "Eli Pariser gave a talk at TED which posits that tailoring algorithms are creating 'filter bubbles' around each user, restricting the information that reaches you to be — unsurprisingly — only what you want to see. While you might be happy that your preferred liberal or conservative news hits you, you'll never get to see the converse. This is because Google, Facebook, newspaper sites and even Netflix filter what hits you before you get to see it. And since they give you what you want, you never see the opposing viewpoints or step outside your comfort zone. It amounts to a claim of censorship through personalization, and now that every site does it, it's becoming a problem. Pariser calls for all sites implementing these algorithms to embed in the algorithms 'some sense of public life' and also have transparency so you can understand why your Google search might look different than someone with opposing tastes."
Hit the link below to watch a video of Pariser's talk.
True, but if Bing will produce customized searches equivalent to holding a mirror up to someone's face, people might opt for Bing instead of Google's "high road." I agree with you that it's better for society to have an opt-in system, I just imagine it might be too risky for a company to implement such a system.
These two systems revolve around how badly people want their mirrors.
Would it exacerbate the problem, or merely hide it? Discarding information that contradicts currently held beliefs is natural enough that most people aren't aware of it, even without personalized search algorithms. I think the bigger issue is the ready availability of like-minded communities that will reinforce your beliefes, no matter how outrageous and outlandish they are.
In his presentation he gave an interesting example. He says he leans liberal, but has conservative friends in facebook, because he's interested in their viewpoint. Then he started noticing that he stopped seeing news links from his conservative friends because the facebook algorithm noticed he didn't click on them. Basically, despite saying that he's interested in the opposing viewpoint, he actually isn't, and was filtering the information himself. The algorithm merely made it transparent and more convenient. Nothing actually changed about the information he was consuming.
It is a problem that people tend to ignore information when it goes against their preconceived notions, but it's not a problem that technology does what we want it to do. If a website kept bombarding me with stories that I didn't want to see, I'd stop visiting it, I wouldn't suddenly start reading those stories.
On second thought, I'm reminded of every April 1st on slashdot, and how every story is bombarded by comments from idiots saying how much they hate slashdot on April Fools' day, and yet they don't seem to leave even for that one day. They keep reading every story and then talking about how much they hate it. Maybe you can make people read what they don't want to read after all...
Warning: Opinions known to be heavily biased.
One of the nice things about slashdot is actually the fact that the readers are not segregated politically.
True, but the more important thing, I think, is that over the years I have often (but not always) discovered that opposing ideas I find on Slashdot have some merit behind them. Hence when someone says something I think it wrong I will often trust it enough to check into it a little and see whether I need to re-evaluate my position. This is why I like Slashdot.
However when reading some random website and encountering something contradictory I am far more likely to assume that the author was some random idiot that doesn't understand what they are talking about than I am to re-evaluate my position simply because experience has shown that this is the most probable case. Hence I would argue that the biggest problem is not so much a "filter bubble" but more that when you hear a dissenting voice you are unlikely to believe it because you do not trust it to be right...although I suppose you could call that a self-filter bubble.