Designing for the Echo Chamber

Ok, this is the piece I wanted to write when I started my last post. The idea of the echo chamber isn’t new, but I wanted to work through a few things to figure out how to start tackling it.

So the biggest challenge in designing against the echo chamber seems to be that this phenomena is singular and personal. I could read an article that is duplicative and reinforces my own views, but it might be completely expansive for my very best of friends.

Countering this type of phenomena probably isn’t done completely programatically. We could filter to get you closer to things you might like, but filtering for content you probably won’t like but might respect enough to consider another point of view is close to impossible. To counter the echo chamber, we each need to find a personal balance in the content we consume. We can’t force people to take a balanced view on things, but you could show them that they haven’t seen enough to consider themselves balanced. So, I’ve been wondering what sort of monitoring tools could be use to help us find a better balance.

Some Inspiration
MediaRDI – At The Center for Civic Media (in the MIT Media Lab), Matt Stampeck and Ethan Zuckerman recently constructed a provocation around a balanced media diet. The idea is that there is a meta-story that puts the content we read into a larger context. Only when we have some sort of content can we understand for ourselves how balanced we are.

Percolate – Percolate is a really great recommendation service created by Noah Brier. The service uses your Twitter followers to suggest which content it thinks you will enjoy. It’s a very smart service, instead of just looking for the most linked content and serving it back to you, it has some analysis to understand what your network means to you. They’ve moved the marker on recommendation, (because the I’ve probably already seen the most linked content anyway). So by striving to put the network in context, the content created by your network gains more context. (If you haven’t, check out the service…the daily recommendation email is really amazing….and I never say that about computer generated email, trust me.)

Organizing Framework
Ok, so let’s put this echo chamber in some sort of context. It seems to generally occur when you’re reading too much of the same content from the same sources/network. Changing either of those aspects could help you along, finding content through new sources or possibly seeing content from a new network. (Pardon the 2X2, but I think it works.)


To understand where your content is coming from, we could start with your social network. Through your Facebook/Twitter stream we could know if content has been passed around your network. We could also even start to see a similarity among sources of content and the networks that consumer them. Media has become so fractionated, this may lend to this type of analysis; the fractions will probably cluster.

To understand the type of content people consume, you would probably operate through a browser plug-in. It would probably feel a little like the StumbleUpon plug-in, which (I think) keeps track of where you’ve been so it won’t send you to a similar destination. A little big brother-ish, but you could design it to be pretty transparent and not so creepy.

So, if we know the dynamics of your network and we know the content you consume, we can probably start to push into new content areas. We wouldn’t be able to know exactly the content, but we could know if it’s new for you (out of network) and if it’s resonating (high view count in other networks.)

If we wanted to ouch things further, through some pretty serious network analysis, you could start to understand content areas that are highly contested (ex. Arab Spring) because of the diversity of network sources that tackle 

So, if you could pull this off, you might interact with the user in a couple of ways…

Personal balance indicator
Some sort of indicator that there is some sort of polarizing content outside of their network. They’re reading one thing, but people outside of their network are reading very different sources. (This gets toward the Kony problem from the first post.) I played with a lot of clever icons, but for this purpose I just stuck with the smiley face; seems universal enough. If you’re viewing something where you’re balanced, either because you’ve read enough, or there is isn’t content outside your network, you get a smiley. If you’re reading a more polarized topic, you’ll get a frown and some suggestions of other content that (together with what you’ve read) will help you begin to balance your “media diet”.



Topic map
Imagine a user seeing that they haven’t seen a balanced set of content. They could also click through the “frown” to see a more holistic organization of content. 

Article map

There’s probably a more elegant way to get to here, this was just a thought experiment. It doesn’t seem like there’s much money in balancing content on the web, but it seems important for who we will become based on the content we consume.

It’s interesting that you almost have to build a representation of the universe to draw directions to get elsewhere. This is the part that feels a little wrong to me, understanding the ever-growing whole seems impossible & un-weblike. 

If you have other ideas, please add them in comments. 

Thanks for reading.

Leave a Reply