According to Google News, I am one of the most popular writers of open source software today, with a total of four articles that I have written on the front page of the Open Source Software category today.
Go, look for yourself. The articles are right there on my screen. Except they probably won't be on yours. That's the filter bubble at work.
"The filter bubble" is a term popularized by the book of the same name written by Eli Pariser. In his book, and subsequent talks Pariser has given at venues such as TED, Pariser highlights the dangers and extent of the personalized searches on the Internet, and how such searches are being monetized right now.
The idea is simple. As you go along your daily business, Google keeps an eye on where and what kind of content you're looking at. It does this better when you are signed on, but it can still track general user traffic flows even when they aren't logged in. As you surf, a more specific profile is developed for you.
The benefit touted to you, of course, is that when you run a search you are more likely to find the results you really want. And to a certain extent, that's true. If I type in "Zonker," for example, Google's pretty smart about knowing that I may not want to see pages about the fictional Doonebury character Zonker Harris, but rather my not-quite-so-fictional colleague Zonker Brockmeier.
Google's benefit is clear. Know the customer, know what they like, and send them to pages they are likely to click and (most importantly) ads they are more likely to click.
The drawback is what Pariser calls the filter bubble, where we are constantly fed a diet of content based on our preferences, and perhaps not on what's really going on. A stark example Pariser presents in his book is when a staunch liberal and an equally strong conservative each run a search on "Barack Obama." Both individuals will see content that favors their worldview, and not the other's. And, since we all love to have our worldviews validated, they may not even notice the problem.
Or take my example. I might look at today's Google News page and think that this Brian Proffitt fellow is certainly very popular. I might therefore conclude I am deserving of a raise from my publisher, and would start a campaign to lobby for a pay increase. Imagine my disappointment when I see that the raw traffic numbers don't really reflect a big jump today. If I didn't know about the filter bubble, I might accuse my publisher of doctoring the web analytic reports, and then the next thing you know I'm out on the street, scrimping by writing articles about Windows Phone 7.
Which, suddenly, will seem very popular on my Google News page.
We've been living with the filter bubble for some time now; this is nothing new. But it's like buying a red car: as soon as you do, you notice red cars everywhere. What's new is how fast data vendors and personal add companies are harvesting and selling your information to advertisers. What's also new is where your data is living more and more these days.
Cloud services offer a lot of convenience to users, but they may have the unfortunate consequence of making your data even easier to get at in the days to come. Right now my cookies (one of many ways to be tracked on the web) are on my local machine. What happens, then, when someone gets the bright idea to save cookies remotely? Xmarks, the popular Firefox extension, does this already:
Great. Of course, Xmarks is an opt-in service. I don't have to use it. But what if the cloud is used to store cookies instead? Where's my control then? (And this is coming, make no mistake.)
I am not trying to sound like a privacy-above-all fanatic. I do appreciate a little tracking here and there, if only to keep passwords on non-critical sites in storage. But if I rely on searching to do my job (and I do), then I don't want results slowly being managed over time into something Google thinks I want.
There are solutions, of course, though they are still in the do-it-yourself realm. The open source community has a real opportunity here to automate solutions like this. But I don't think it will happen: should Mozilla try to bake something like this into Firefox, I have a feeling Google, a major supporter of the Mozilla Foundation, would object on behalf of itself or its data-collection partners. Strongly. An add-on could be created, but such tools can be hit or miss.
For now, we can urge such development to take place, because honestly I can't see any effort coming from anyone with a commercial stake in keeping filter bubbles going.
But we can hope.