We take in so much information that it’s impossible to sort through everything and process the information that’s most important. This fact is complicated by content creators and reporters trying to get all their information out and making us process all that information.
Regular people aren’t exempt in wanting to share what we have to say–we just don’t usually do it with a corporate bent. We spout out what we’re doing, thoughts on current events, and our dreams and aspirations. We share a small piece of our complex lives with others; whether we do this to keep in touch with friends, get a reaction, or something else entirely is up to us.
There’s a small problem. Not everyone we know is on every site we’re also on. Not all my college friends are on Twitter. I’m not on Tumblr or Pinterest or Instagram, and I hardly ever check Facebook. Most of these sites encourage you to share your content elsewhere, both as a way for you to share with your friends and for that site to get more users. This is all fine and good until users start autoposting all their content to other sites and you can tell–their feed is full of obviously autoposted fb.me and 4sq.com and tvt.ag and tmblr.co links with little context or thought as to whether anyone on the other site is interested. Without that context I’m very unlikely to figure out why I should click, so I don’t. There’s a lot more out there to read.
I saw a lot of this on my Twitter feed, so I took advantage of Tweetdeck and did what the client was built for. I filtered those autoposts out. They were easy to spot: more and more sites own short URLs in an attempt to ease sharing in a character-limited environment.
Over time I saw more and more things that annoyed me for other reasons. Sites like Buzzfeed and Upworthy annoyed me with their listicles, clickbait titles, and low quality articles like “20 Facts That Will Blow Your Mind” or “14 Things Only Other Math Nerds Will Understand”. So I filtered the words Buzzfeed and Upworthy out of my Tweetdeck client.
There are other things that I’d like to see less of as well, as well as other areas of personal content curation that I wanted to explore. So I installed the Rather browser extension. You might have already heard of it–it’s the generalized Unbaby extension. Rather lets you filter by topic and suggests some topics to get you started; some suggestions included politics (yes), babies (for Facebook, very yes), pregnancy (dear Baty yes), and current events. Once you select a topic, you can add or remove keywords and URLs for Rather to filter. Rather also appears to support some regular expressions, though I haven’t tested their full extent yet.
While Rather does filter out some things well, it’s limited by the words we use to describe our updates. One example that still stands out is one Twitter friend saying we should vote that day to be Friday. Since “vote” was a political keyword, Rather showed this as blocked content by its own keywords, even though this update had nothing to do with politics. Since Rather shows who made the update, I deduced that the person who wrote it probably wasn’t talking politics, so I unhid it.
“Vote” isn’t the only term that can have multiple meanings. “Aww” is another one, this one showing up in the kids/babies category. But “aww” can also apply to cute animals, and I’m all about cute animals showing up in my feeds. Without mentioning in the update text whether the photo is of a kid or an animal, I’m still left to my own devices, especially since many folks will just use the kid’s (or animal’s) name instead of saying the photo contains a baby.
Even taken alone, this is a very interesting problem: how can we use existing content to predict future content? That is, if we know someone has posted a lot about kids, can we use that to predict whether that “aww” is about their kid or their dog? It sounds like something for Facebook to work on since they already have so much data.
How far can we go with filtering things out? Most notably, search engines filter your results as an attempt to show you relevant results quickly. Websites track you so they can show you material more relevant to your interests. We tend to like and share things we agree on, and this data is used to show you more things you can like and share (and probably agree with). The result is a filter bubble, not just for your search results, but for how you consume the web–which shows in large part how you see the world around you.
That’s just for the filtering others do for you. What about your self-filtering? There are plenty of reasons for self-filtering that are outside the scope of this post–for example, filtering something traumatic to you. There are also the things we choose to filter just because we don’t like them. I’m guilty of this; I hide posts from Mashable and Buzzfeed and all those autoposts from Facebook and Foursquare and Tumblr.
But what about filtering out terms just because you don’t want to see them? Babies and pregnancy would be among them for me, if such a thing could be accomplished. And it’s not just babies; let’s be honest, sometimes I don’t want to read about politics. While most of the people I follow online (on Twitter, anyway) are similarly minded to me, we still disagree on things. Filtering out those things based on context would mean missing out on potential new points of view that could expand your worldview. This gets us caught in the same filter bubble the web is putting us in.
So should we filter our own content online, or do it more than we already do? I don’t know. While there are some great uses for filtering your own online feeds, doing so is far from perfect. I just don’t want to miss out on something new because I chose not to see something related to it.