Overview of TED Talk: Eli Pariser's Beware Online "Filter Bubbles"

Summary of TED Talk

According to Eli Pariser, if the internet is based upon the idea that only relevant information should be available, then we all have a problem on our hands.  He uses the following quote to begin his argument:

"A squirrel dying in front of your house may be more relevant to your interests right now than people dying in Africa."

 - Mark Zuckerberg, Facebook

He first uncovered this concept when routinely viewing his Facebook news feed.  Eli describes how even though he personally preferred to view the links posted by his conservative-minded friends, his feed eventually removed them all from his view, simply because the algorithm in which he talks about more in depth later in his speech recorded how Eli actually viewed more links from his liberal-minded friends.  Eli then proceeds with telling us that there is no longer a standard Google.  He effectively conveys this idea by showing us screenshots of two separate Google searches, conducted by friends of his, for Egypt.  We can clearly see the differences between each of the search results.

Screenshot from "Beware Online Filter Bubbles" Video

Eli then goes on to explain how a person's searches become so-called personalized, without having logged in to anything:

There are 57 signals, including which device you're using, the location in which you're using that device, and which browser you're using to  search.

Eli continues with this idea that the internet shows us what it thinks we would like to see, rather than what we need to see.  He uses the following quote to further explain this idea:

"It will be very hard for people to watch or consume something that has not in one sense been tailored for them."

- Eric Schmidt, Google 

Eli then gives us a visual of what he likes to call a "Filter Bubble". Originally, the idea behind the internet was worldwide connectivity.  He explains that today, the internet is "your own personal, unique universe of information", and that it "depends on who you are, and what you do".  The only problem is, "you don't decide what gets in", and "you don't decide what gets edited out".

Screenshot from Video

Eli then presents to us a study done by Netflix, explaining that "Filter Bubbles" may be the reason for certain types of movies being more accessible than others.  He compares this idea to our "future aspirational selves" versus our "impulsive present selves".  And since these search engines, companies, and so on, look at what we tend to click on the most, this balance between what we want to watch and what we should be watching becomes unbalanced.

Screenshot from Video

To conclude, Eli explains to us the idea that gatekeepers, or those who choose what gets published based upon merit and ability to sell, used to be human.  But with the internet, gate-keeping has become algorithmic.  These algorithms decide what you get to see, based upon what you typically view or search.  Eli makes the point that these algorithms have "no embedded ethics".  And since these gatekeepers have become inclined to just show us what is relevant, we seldom have the opportunity to see the uncomfortable, the important, the challenging information; the information that makes us see another point of view.  Eli ends his speech by stressing that the lack of "civic responsibility" on the web is isolating us, when it needs to connect us.

Filter Bubbles Experiment

I set up an experiment with a friend of mine to see if filter bubbles actually exist, due to my own speculation.  We began with a basic search in Google for "apple pie".  Here are the results...

My Google Search for "Apple Pie"

My Friend's Google Search for "Apple Pie"

Unfortunately, there is absolutely no difference between my search and my friends search.  So I decided to try this experiment again, but this time, using "Newtown Connecticut".  I chose this search because I figured if it was a relevant topic, then based on which one of us usually clicks on the more serious links, I might be able to see a difference between the two.  Turns out, I was right. Here are the results...

My Google Search for "Newtown, Connecticut"

My Friend's Google Search for "Newtown, Connecticut"

Notice how my Google search lists "News for Newtown Connecticut" which included the recent tragedy at Sandy Hook Elementary School. Meanwhile, my friend's Google search lists a map of Newtown, Connecticut first, followed by Wikipedia.  Nowhere on her search does it say "Sandy Hook Elementary".

The conclusion that can be drawn from these searches is that I tend to search for more news stories and world events than my friend does. As of now, I am no longer a skeptic of "Filter Bubbles".

How do Filter Bubbles work in my life?

I am an avid Pinterest user, and I have noticed since learning about Filter Bubbles, that whenever I pin or re-pin an item onto a board under a certain category, and then I immediately go to that category, my pins or re-pins never show up.  This might be due to Filter Bubbles because it doesn't want me to mistakenly re-pin an item of my own, although this might be a stretch.

A more realistic view of how Filter Bubbles affect my viewing of the web would be my use of Google+.  With Google+, I can add friends to my circle, and by doing so, I can search for something, and get results from my what my friends post, or what they like.  Although, since I have just recently started to use Google+, I haven't been able to fully take advantage of it's convenience.

Here is a sample screenshot of what people see through Google+ searches...

Screenshot of Google+

Notice at the top of the screen how it says "60 personal results and 187,000,000 other results" with friends faces next to it.  Eli Pariser might be thoroughly upset with the creation of Google+ because it only shows us what is relevant in our lives, instead of showing us more important things.  

Here is the Google+ video explaining how this concept works...

I can honestly say I am not a fan of Filter Bubbles.  Even though Google+ is convenient, I don't like the idea that I am seeing an entirely different search than everyone else.  Plus, I don't like the idea that I might miss very important information, just because some algorithm thinks I should be seeing less important information, based on what I typically click on, when I'm not using the web for research.  And this also makes me think, "What if I'm writing a research paper for school, and through the primary stages of research when I use Google, I miss a crucial piece of information, just because of Filter Bubbles?" 

I would like to take this time to thank Eli Pariser for enlightening us all of the battle we must overcome with these unfortunate Filter Bubbles.