The Filter Bubble Problem That Tech Companies Have To Solve

What is the best solution to the internet’s filter bubble problem?

For the first part of this post click here.

The first part of the post ended with a research fellow at the Oxford Internet Institute, Grant Blank, talking about the ways in which researchers defined the term echo chamber and that according to results his team obtained, there was no echo chamber.

Grant Blank also said that people living in the United States of America actually got their information from multiple forms of media.

According to Blank, Americans on average consumed about five different and separate media sources.

More specifically, they consumed around three offline sources and two online sources.

Because of that, it was hard to believe that the American people did not come across a diverse range of opinions.

Blank also said that people living in the United States of America did encounter things that they did not agree with.

Moreover, Americans also changed their mind about specific things based on the things which they regularly encountered in the media.

This is a point that even the likes of Pariser (who coined the term filter bubble several years ago) agreed on that one could not simply blame the internet for all the ills in the country’s polarized political scene.

Some also believe that it might even help to explain why the US liberal elite did not have the foresight to see the coming of Trump.

Afterall, there is no doubt about the fact that a major portion of middle America did not engage in various social-media feeds belonging to liberal people and liberal sources.

And indeed, Blank (more specifically Blank’s work) did conclude that the vast majority of researchers who found such an effect found so because they only studied the cultural liberal elites.

However, as far as most of the Trump supporters went, pre-internet filter bubble such as,

  • Local news
  • Fox
  • Talk radio

managed to play a far important role in terms of news sources than fake news on social media sites like Facebook and tweets on platforms such as Twitter.

It is also true that data coming from Pew, a polling firm, backed up this idea.

In short, the internet did not represent the only source of polarizing news in the country.

It came from a multiple number of news platforms and sources.

After the United States presidential elections in 2016, Pew actually found out that around 60 percent of all Americans got their news from sources such as social media sites.

However, most of the articles covering the study ignored a parenthetical in the study which said that only around 18 percent of Americans said that they did so regularly.

Moving forward to a newer Pew study, researchers found that around 5 percent of all Americans made the statement that they actually had a lot of trust in all the information that they received.

The director of the Center for Civic Media Massachusetts Institute of Technology, Ethan Zuckerman recently said that he did not consider the internet as the absolute causal factor of political polarization.

However, he also said that he believed the society, as a whole, was experiencing a phenomenon which began with news sources such as Fox News and now had sort of extended itself into the space that we call social media.

So if that makes sense, then is there something that we, the society, can do about this problem?

Well, some people have attempted to fix the problem of the filter bubble with varying degree of success.

The three attempts to fix the internet bubble

After the United States presidential elections in 2016, Zuckerman along with a number of other collaborators actually managed to design a tool that they called Gobo.

What did Gobo do?

It allowed people to adjust and modify their own filter bubbles with the help of sliders which allowed them to control all the content filters.

To take an example, there may be a political slider and it could have a range from “many perspectives” to “only one perspective”.

If the user selects the “lots of perspectives” option, then the tool would expose the user to those media outlets which they are unlikely to see on a regular basis.


However, the social media giant Facebook showed from little to no interest in using and/or adopting the tool called Gobo.

Zuckerman recently commented that what Facebook was worried about was that the company believed (and according to Zuckerman they were probably right) very few Facebook users would actually want to move ahead with diversifying their news feed.

Members from Deb Roy’s lab also developed a tool which they called Social Mirror.

In the early part of 2018, members of the lab reported their results of an experiment that they conducted with the help of the new tool.

The tool, called Social Mirror, used techniques such as data visualization in order to give users on the Twitter platform a wide bird’s-eye view of the way in which their whole network of friends and followers fitted into the larger and more general universe of the platform, Twitter.

Deb Roy’s lab members tried to recruit only those Twitter users who were politically active on the platform.

It turns out, the recruits themselves found it very surprising to learn the extent of how much they had cocooned themselves inside far-left and/or far-right bubbles on the platform.

However, the experiment was unable to make any type of long-term impact.

Researchers found that even though some participants in the experiment did follow a more diverse range of political news via Twitter accounts than before even after a week of the experiment ending, after about two or three weeks the majority of them had given up and had gone back to their previous set of homogenous news sources.

Moreover, researchers came across another twist.

Participants who had agreed to start following a larger number of contrarian accounts as researchers suggested to them in order to assist them in diversifying their Twitter news feeds, eventually reported that they had actually started to have even less inclination to actually talk to people who held opposing political views.

It should not be hard for anyone to see that Zuckerman’s team were only able to achieve lousy results.

Such results, among other things, actually pushed Zuckerman towards a slightly more radical idea in order to counter the internet’s filter bubbles.

He wanted to create a social media platform with the taxpayer’s money but with a civic mission in mind to help and provide users with a global and diverse view of the whole world.

Recently Zuckerman wrote an essay for the Atlantic and noted that the early United States of America did not have a problem with featuring a highly partisan press.


The people working behind news sources tailored the press to some very specific audiences.

However, editors and publishers, for the majority of the cases, did not shy away from abiding by the country’s strong cultural norms.

These publishers and editors published a huge array of stores from all the different sections of the young nation.

They also reflected on various different political proclivities.

In many democracies, public broadcasters have successfully tried to focus on provider citizens with a wide array of different perspectives.

However, Zuckerman makes the argument that following in their footsteps in simply not realistic on part of outlets such as Facebook.


Well, according to Zuckerman, the reason why the society should not expect Facebook to do the same as many public broadcasters is the company’s business model.

The business model which sustains Facebook actually drives the company to pander to the people’s natural human desire.

That desire is to take the easy way and congregate with other people who are just like themselves.

According to Zuckerman, a social media platform which was public and had a civic mission could actually find success in pushing unfamiliar perspectives directly into the user’s feeds.

Of course, such a platform would also push people out of their comfort zones.

Computer scientists and other scholars could actually review various algorithms used in providing news to users in order to make sure that people would see an unbiased and balanced representation of different views.

Zuckerman is not under any illusions though.

He has admitted in the past that it is highly likely that people would not shy away from complaining about such a social media platform receiving public funds.

People are also highly likely to, according to Zuckerman, question the platform’s even-handedness even if it came into existence.

However, Zuckerman believes that it definitely deserves a shot since there were no other viable solutions to problems such as internet filter bubble.

The people are the problem

Or are they?

A social psychologist working at New York University, Jay Van Bavel, has actually studied a good number of social posts and has also analyzed which one of those posts was more likely than other to gain sufficient traction.

Jay Van Bavel found that posts which fell under the category of group identification actually activated the most primitive and non-intellectual sections of the human brain.

To take an example, if a politician from the Republican party told people that all the immigrants that were moving into their neighborhoods were actually changing their country’s culture and/or taking away local opportunities or jobs, or if a politician from the Democratic party informs a group of female students that activists of the Christian faith wanted to ban all women’s rights, the words of that specific politician would carry a lot weight and power.

Jan Van Bavel’s research also suggested that if someone really had a desire to overcome biased and/or partisan divisions then he/she had to focus on the emotions rather than the intellect.

Members from the Deb Roy’s lab did not lose hope after their Social Mirror experiment did not achieve the desired results.

They actually went back to work and a few months later debuted a new project which went by the name of FlipFeed.

This exposed people using the Twitter platform to other Twitter users who had different views on subjects such as politics.

The lead author of the study, Martin Saveski, recently said that the whole point of using such tools was to simply try and change how people on these social media platforms felt about people who were on the other side.

In one of the newer experiments, researchers prompted participants to actually try and change their perspective on things.

In simpler terms, researchers wanted participants to actually not get aggressive whenever they encountered a view that opposed their own and imagine if they simply heard their friend having a similar opposing view.

Researchers found that people who received this prompt showed more willingness to say that they would have little problems in speaking with a person who held opposing views more appropriately in the future.

The same participants also said that they now had a better understanding of why the other person had a view that opposed their own on a given subject.


Researchers found that their results actually had congruence with the observations that Pariser made a while back.

Pariser recently noted the fact that some of the most civil political discussions actually took place in online sports forums.


Because the people participating in such forums have great unity.

And why do they have great unity?

They have great unity because they love a single sports team.

That commonality is the reason why they are so united and hence so balanced in their political views.

Of course, Pariser is assuming here that all the participants in the political discussions consider themselves fans of a given sports team first and liberal and/or conservative second.

These participants in online sports forums have an emotional connection with each other even before they enter the field of discussing politics.

If one take a slightly deeper look at the type of projects that the like of Deb Roy and Zuckerman along with other research teams have carried out, one can easily see that what these people are actually trying to do is to simply find a suitable way to employ technology in order to get people to engage with online content that comes from outside their own political bubble.

However, who is to say that is a workable solution to strive towards?

Deb Roy has himself said in the past that he does not believe that the problem of political polarization had any pure and/or simple technological fix.

Some believe that, in the end, it is actually up to the people themselves to decide if they truly want to expose themselves to content which a larger array of news sources.

Of course, such people then also have to make the decision of whether or not they want to engage with people who hold the views expressed in that new content.

To many of our readers, it may sound rather unappealing to go out on your own to find views opposing your own.

Such readers should also give through to the alternative of that approach.

Readers who have no interest in what the other side has to say will continue to post outraged political posts on social media platforms.

And in the process of doing so, they will not accomplish much.

New research shows that anyone who comes across such a post is already certain in his/her heart and mind that the post and its content are true.

Newer technologies such as artificial intelligence and machine learning may help social media sites such as Facebook, Twitter and others to finally deliver content to users which is as balanced as possible.

Of course, one simply cannot ignore the bias problem with computer algorithms and those who write them as well.

Plenty of research shows that machine learning algorithms are only as good as the data that they train on.

Will social media platforms be able to help people out in staying away from extremes and have a balanced view of the world that is surrounding them?

Will they help people empathy for the other side?

These are all questions which are, at the moment, hard to answer.

In their attempt to keep the growth philosophy going year on year, technologies companies especially those involved in social media websites and other news sources will simply refrain from doing anything that might lead to a decrease stock price or shareholder satisfaction.

All the while these social media platforms also have to make sure that they provide enough “relevant” content to users so that they keep coming back for more.

How can a single social media platform balance the need for a balanced set of news sources with the need for the user to come to the site more often (for which the platform has to provide the user with the content that he/she likes rather than the content from the other side which the user may or may not like) is also a hard question to answer.

All of these might also lead some to think about the early days of the internet and its purpose.

Currently, there are a total of four ways in which ex-internet idealists like to explain where the internet went wrong.

To read more on that click here.


Zohair A. Zohair is currently a content crafter at Security Gladiators and has been involved in the technology industry for more than a decade. He is an engineer by training and, naturally, likes to help people solve their tech related problems. When he is not writing, he can usually be found practicing his free-kicks in the ground beside his house.
Leave a Comment

This site uses Akismet to reduce spam. Learn how your comment data is processed.