Tag Archives: cognitive bias

Internet Filter Bubbles

As anyone reading this knows by now, I like TED talks. One of my recent favorites, and one which I shared with friends and even showed to staff at work, is Eli Pariser’s brilliant talk about Internet filter bubbles. A few days ago at the World Science Fiction conference in Reno, I was on a panel about social media, and I mentioned the talk. Cory Doctorow was also on the panel and he added that Eli has a whole book on the topic. I stopped by Amy Cat’s book booth in the dealer’s room and grabbed a copy. On the long drive home to the Seattle area, the book became a topic of conversation. We read through the introduction out loud in the car and stopped every few paragraphs to talk about it.

Please bear in mind I have NOT finished the book, but I think it’s an important conversation, and so here are my thoughts after hearing the talk and reading the introduction.

I’ll start with a very short summary, but I actually recommend watching the talk, which will be worth the nine minutes of your time that it will take. My summary will not do it justice. But here goes:

The major online players including Google, Facebook, Amazon, and others are filtering content so that we get content directed for us. This includes ads, product recommendations, and even news. For example, if I regularly click on and visit Internet sites associated with liberal views on climate science, then related search results, product offerings, and even social information from like-minded friends will increase in availability and opposing views will decrease. The information I see on the Internet will be tailored to me in a self-referential fashion. Eli’s point is that this is largely bad, and may further cement the polarization of people into silos of belief.

Here are the primary points from our somewhat Worldcon-bleary discussions:

First, we agree that it’s happening to a great extent. Certainly Amazon doesn’t bother to present me with sports books, which I wouldn’t buy if it did. This is largely good. It wastes neither their bits nor my time. When it gets to news, it’s not so good. However neither of us uses only one news source. The New York Times is a favorite read of mine and my partner, Toni, almost never reads it. We had many examples of diversity in news sources between us, and yet we share lives, share many beliefs, and even have similar jobs. We decided that personalization benefits us far more than it hurts us (which is not that same as saying there is no harm).

One of the things Eli says in his introduction is that “In the filter bubble, there’s less room for the chance encounters that bring insight and learning.” I suppose this would be true if there were only one way to learn. If I simply used search engines, and they all filtered with similar algorithms based on my own clicks, whole blocks of information would be effectively “hidden” from me. But that’s not our behavior, and I suspect it is not the behavior of most people. For example, in my social networks I follow political topics, and sure, those look right back at me by mirroring my own beliefs. But I also follow a wide swath of writers, and they have different political leanings as well varied points of view on many other topics. They live all over the world. Toni follows competition dog lovers, and they diverge even wider across the political spectrum. Then, because we have a joined subset of social media contacts that reflects things we’ve shared with each other, I hear about dogs (and the dog lover’s politics) and Toni hears about writers and bloggers (and their politics). The Internet as a whole has done more to widen our beliefs than to narrow them.

We tried to imagine a world where people who followed something more like pop culture than we do might live; say a reality TV addict. I suppose that if there are people who are only interested in the Kardashians and Brittney Spears, than they might be stilted even further by filter bubbles. But we tried to name some. Even though I have an old friend who has Fox news on every day (and is thus closer to my shuddering image of the “average” American), they have wide interests and it would be unfair to say that they aren’t encountering new ideas daily. If nothing else, they have social media conversations with me where we genially argue topics we are on opposite sides of. This enriches both of us. In other words, I couldn’t name a single individual so shallow that they allow the filter bubble to encase them in a wall of information that does nothing but mirror their own beliefs back to them. There is an archetypical “Joe Six Pack” who doesn’t really think about much and lets life come at him as he perches on the couch in front of the TV and drinks beers after work. I’m not sure that person exists. Yes – I‘ve read appalling comments on web articles that tell me there are some narrow thinkers out there. But we couldn’t name even one by name, even though I could certainly identify a few on the internet given a minute or two. Most people are more interesting and capable.

Lastly, the people who create content – like me and everyone else who posts here, like almost all writers and painters and poets and musicians — are generally curious and somewhat intellectual beings. We actively search out deeper information than that found on the front page of Fox, CNN, or MSNBC. I know, because I know what gets shared with me in internet social circles.

Eli ended the introduction with a comment that I believe is completely correct. He says, “…it’s critically important to render [the internet filter bubble] visible.” I agree with that. We should know as much as possible about how choices are being made about the content that we see. We should be smart enough to know what not to trust, just like we know to mistrust TV ads. I’m a strong advocate of all kinds of transparency. I do believe there are inherent dangers in the bubble as well as gifts. But there’s a lot of unformatted and difficult data on the Internet, and content filtering helps me far more than it hurts me. Maybe that’s because I have such varied interests that no algorithm can pin me to sports books or the Kardashians. The point is, almost everyone else does, too.

Unless, of course, I’m buried so deeply in the filter bubble that I truly can’t see outside.

What do you think?


Brenda Cooper’s latest science fiction novel, Wings of Creation, is out now from Tor Books. For more information, see her website!