The Internet is designed to confine users to a bias matrix that prevents them from experiencing the realities of the world. This unsettling phenomenon, known as the filter bubble, was exposed in a Ted Talk presented by Eli Pariser titled ‘Beware of Online Filter Bubbles, (2014).’ According to the Ted Talk, the Internet utilizes an immense number of signals (i.e., type of web browser, previously clicked on links, geographical location) to create a ‘safe space’ for users, only displaying to them what they want to see and what they feel comfortable with. Upon the surface, the filter bubble seems to promote efficiency: why present users with a clutter of information they have no interest in, when you can quickly display exactly what they want to find? However, hiding from users what they don’t want to see creates a single-minded view of how the world operates. It does not paint the whole picture; it does not show the other side of the coin. This fabricated version of reality not only modulates personal viewpoints, but also can have an immense impact on decisions that affect the world (e.g., presidential elections, the use of fossil fuels).
To test the filter bubble phenomenon, I compared my girlfriend’s social media, Netflix, and news sources to my own. First stop, Facebook:
Facebook has admitted, and has been sued for, filtering ads based on race and ethnicity. Furthermore, Facebook filters individual newsfeeds based geographic locations, relationship status, what links people have clicked on first, etc. Additionally and most likely having the most impact, newsfeeds are strongly filtered by the type of friends people have and what they post.
Surprisingly, my girlfriend and I had similar Facebook newsfeeds. Although I had more information and posts regarding politics, both newsfeeds had a pro-conservative, anti-democrat taste. This can be explained by both of us growing up in the South, thus having friends that are eager to get that wall built ASAP. In reality, I am more Liberal on a lot of issues; however the links I press on first are usually from my family or close friends-in which despise everything about Obama.
Other then the few political posts, the majority of my girlfriends newsfeed consisted of comical memes usually involving cats or shiba inu’s, relationship quotes, DIY projects, and sponsored ads by fast food chains (e.g., Wendy’s) and retail (e.g., Spencers). On the flip side, my newsfeed consisted of a lot of workout routines (my friends are workoutaholics), wrestling and waterskiing videos, and inspirational quotes. Although some of the memes were the same, overall her newsfeed more aligned with Reddit while mine aligned more with a news website (I think hers was more interesting).
In relation to every other difference between my girlfriend and I’s filter bubbles, Netflix was the most dramatic. My recommended movies consisted of one-star kung-fu action films (I purposefully watch a lot of low rated movies, looking for that hidden treasure), and acclaimed classics such as Pulp Fiction and Blow. Parks and Rec, Scrubs, and other comedies were the first to appear on the list of TV shoes. Conversely, her ‘recommended to watch’ consisted of a handful of documentaries, dramas (i.e., Gray’s anatomy), and animas. Because we often watch Netflix together, we did have some similarities in our recommended to watch, as American Horror Story and Shameless were on the top (if you have never seen Shameless, please watch it). Not to stereotype, but I felt like the differences in movie and TV-show recommendations were based on that I enjoy watching things most men would prefer, while she enjoys watching what most women would prefer.
Poker, Kid Cudi, poker, Red Hot Chili Peppers, Ted Talk, Poker, NCAA wrestling finals, how to run statistical software, different workouts, poker.. That’s primarily what my YouTube filter Bubble consisted of. My overabundance of poker recommendations is most likely due to me utilizing YouTube as a coping mechanism when I come home with my pockets empty after a ‘bad beat’ at a house-game; watching other people lose a lot more money then me gives me a sense of comfort (even though they have a lot more money to lose). Kind of comical, after doing a YouTube Search about how to run a statistical analysis of variance on a program we use on my research team, all of my recommended videos below were still about poker. I guess this is proof that I lose more money then I win most of the time J.
Exotic animals, Keith Urban, DYI projects, and ‘WhatsTrending..’ that’s primarily what my girlfriends YouTube filter Bubble consists of. She is into country music while I like alternative and rap. She would rather watch grass grow then play poker, and she is much more interested in wildlife as compared to my interest statistical psychology. A lot of her YouTube recommendations consisted of horse back riding, how to do your makeup, and how to complete basic algebraic equations (she is in algebra).
There were notable differences when searching news sources such as CCN.com, The Huffington post, and Foxnews. Although we both had news about upcoming weather in our area, the ‘top stories’ differed significantly. The majority of my top stories dealt with how the repeal of Obama Care care would cause millions of citizens to go broke (which is interesting considering most of my Facebook News Feed is conservative), and other articles related to Trumps political Agendas. My girlfriend’s top stories via news sources were mainly about relationship struggles among celebrities (EXTREMELY IMPORTANT STUFF). The search also displayed how she was more pro LGBT while I didn’t have much care on the topic; as one of her top stories was “Aarow Star Colton Haynes Is Engage to Jeff Leatham,” while mine was “Todd Starnes: Hey, Sean, I was Attacked by a Raging Feminist, too.”
In conclusion, I was surprised by what my girlfriend saw on the Internet as compared to what I saw on the Internet. This is a reflection of our difference in interests, difference in age, where we were raised and whom we are associated with. Because she is still new to my life, if we decide to stay together I would hypothesize that our filter bubbles will become more similar in the future (as our values and interests will become more similar).
Both of our filter bubbles primarily consisted of what was relevant near us. There was much about Trump, recent storm destruction, and police arrests; and so little about what was happening globally. We may not know that a school in Africa was just got shot up and thirty children died, if the people who committed the attacks were allied with the American government. We won’t know when a white man kills ten individuals in Australia, but we will sure know when a Muslim shoots someone in America.
At last, I wanted to do an incognito search to see what the Internet would show me if I was anonymous. It should be noted that I am not using an IP changer; thus the Internet can still present information relevant to my geographical location.
Recommendations on Incognito YouTube are primarily based on what others have been recently viewing. From a ‘Thai Noodle Challenge,” to “What Happens When You put 20,000 Volts into a Watermelon,” the variety is limitless. Unfortunately, there are a lot of anti-Muslim videos, exemplifying how the majority of the United States scapegoats the religion. Thus, although I feel incognito is less bias, it is still bias in regards what the common internet user in America is thinking/interested in.
Although the new stories were practically the same, they were arranged in a different order. While my top stories dealt with politics (mentioned above), top stories going incognito were about the dangers of marijuana and fake weed. Furthermore, there were a lot of posts related to childcare and rearing; things I have zero interest in.
Conclusion/In Text References
According to Wikipedia and other internet sources, the filter bubble is an algorithmic approach; which is a procedure that guarantees to achieve a goal (Hillis, 1998). However, I am confused about why the filter bubble is considered an algorithm and not a heuristic, as it is not right 100% of the time. Although it combines a number of factors to determine what should and what should not be present on a selected browser, it is not 100% accurate; rather it gives the ‘almost right answer (Hillis, 1998). For instance, I have more interest in what natural disasters are currently occurring in the world as opposed to the tornado warning in Manhattan last week; thus, if the filter browser was 100% accurate shouldn’t it have presented world news first?
I found difficulty in referencing Andrew Blum’s novel ‘Tubes,’ as the first four chapters mainly covered the development of the internet. However, Blum’s (2013) novel discusses how infinite networks and servers peer together so the internet becomes less congested. If filter bubbles did not exist, and websites would supply everyone with every single bit of information, the exchange of traffic would be immense and (I think) the internet would become slower. Thus, filter bubbles are in place to monitor and limit the amount of traffic that is being sent to a server, and to ensure the traffic is relevant to the viewer. Furthermore, the novel said the most important people that contribute to the internet today are not hardware engineers anymore, but analysts that can predict what people want to view and see. These are the people that assess all of the factors that go into a filter bubble.
Blum, A. (2013). Tube: A journey to the center of the internet. New York, NY: HarperCollins.
Hillis, D. W. (2015). The pattern on the stone: The simple ideas that make computers work. New York, NY: Basic Books.
Pariser, E. (2014). Beware of the filter bubble . Retrieved from http://blog.ted.com/how-do-i-cite-a-tedtalk/