Philip Guo (Phil Guo, Philip J. Guo, Philip Jia Guo, pgbovine)

Enforcing diversity in online media consumption

Summary
The over-personalized nature of online media consumption causes us to be myopic, closed-minded, and less able to empathize with people who are different from us. To combat these ill effects, I propose an idea for a web browser home page that aggregates and summarizes articles from a diverse set of sources, thus naturally exposing people to different facets of society.

The problem with online media consumption

Think about how you choose which articles to read online. Chances are,

  • You probably have a news site set as your web browser's home page, so that's the first page you see when you open your browser. This news site reflects your educational level and political attitudes. For example, a vegan hipster from San Francisco is not going to have foxnews.com as her home page.

  • You probably bookmark a dozen or so web sites, blogs, and news aggregator sites that you read in your free time. Again, this selection is (by definition) customized for your own tastes and world-view.

  • You read articles that your friends email to you or share on social networking web sites like Facebook, Twitter, or Google+. Since your friends are similar to you, they will inevitably recommend articles that reinforce your preconceived notions about the world. For example, my friends only send me articles from "high-brow" liberal-leaning publications like the New York Times, NPR, and the New Yorker. I've never received a single recommendation to read a Fox News article (except maybe as a joke).

  • Now here's the kicker: All of the news articles, blog posts, and other sites that you read link to other like-minded web sites. For example, The Huffington Post will link to rabid-liberal bloggers, religious fundamentalist websites will link to creationist bloggers, and Latino empowerment blogs will link to articles that sympathize with their social causes.

  • More broadly, services like Netflix and Amazon are using collaborative filtering to make personalized recommendations based on what thousands of like-minded strangers also enjoy. For example, If you buy a certain book X, Amazon will recommend five more books that other like-minded people also bought.

The effect of this "over-personalization" of the web is that it is way too easy and addictive to only read what we are already comfortable with and ignore (or worse, ostracize) the cultural values of people in different demographics.

Inspiration from the doctor's office waiting room

Now here's an alternative scenario for media consumption: Imagine you're bored waiting for your appointment at the doctor's or dentist's office. There are over a dozen different magazines available in the waiting room, each catering to a specific target demographic. Medical offices bulk-subscribe to a diverse selection of magazines because their patients come from all sorts of backgrounds and occupations. For instance, there are men's interest magazines (filled with bikini pics), women's magazines, celebrity gossip magazines, Asian-American focused magazines, Hispanic magazines, African-American magazines, science and technology magazines, etc.

During a 20-minute time span, you can quickly skim through the headlines and pictures in a dozen magazines to get a vague "feel" for what each magazine's target demographic values and respects. You don't need to read each article in detail, but just the act of flipping through a magazine can give you a good indication of what each demographic is concerned with at the present moment. Although this experience might seem superficial, it is still far more exposure to diversity than we would usually get from the myopic way in which we consume online media.

My proposal: An online version of the waiting-room experience

I propose to emulate that waiting-room experience in the digital world by creating a home page containing a summary of headlines and photos from a diverse variety of online publications. Thus, when people open their web browsers, they immediately get exposed to a variety of different viewpoints, most of which are on topics that they either have little interest in, or even better, actively disagree with!

In its simplest form, this page would be like a Google Reader or a news aggregator site that comes pre-populated with a fixed selection of sources. This is analogous to the medical office staff picking a selection of magazine subscriptions for their waiting room. The key idea behind this proposal is not to let people customize anything, because over-customization will inevitably cause us to more deeply entrench ourselves within our silos of comfort.

A more sophisticated implementation of such a system would be to flip traditional collaborative filtering algorithms on their heads: Instead of making recommendations based on what like-minded people are consuming, have the system make recommendations based on what radically-different people are consuming. For example, web sites like Google and Facebook are already able to build up "taste profiles" to bucket you into a demographic for the purposes of targeting advertisements. They could easily leverage this data to suggest online media that people in different demographics enjoy rather than what people in your own demographic enjoy.

The form of "diversity" that I'm advocating isn't limited to the cliched call for people to learn about both liberal and conservative political ideologies. Rather, being exposed to diverse forms of media means reading about things that people in your demographic would never think to read in the first place. For example, you might skim a recommended article that lots of long-distance truck drivers enjoyed. Even if you had no interest in the truck driving business, you might be pleasantly surprised to learn about how that business is actually similar to your own, albeit with different technical jargon. Or you might be shocked to learn that working-class black women and Asian men with Ph.D.s face similar struggles in their jobs and complain about the same woes (that's a hypothetical example, but who knows, it might be true!). My point is that if you just stay within your comfort zone of only reading articles that your friends or recommendation algorithms send to you, then you lose many such opportunities for serendipity.

In closing, I strongly believe that forcing people to be exposed to the values and viewpoints of different demographics can ultimately lead to improved empathy and understanding. Since all of the forces of online over-personalization are driving us towards greater entrenchment and polarization, we must actively fight these effects by enforcing some diversity in online media consumption.

Created: 2011-09-25
Last modified: 2011-09-25
Related pages tagged as social observations:
Related pages tagged as human-computer interaction: