Freedom, critical sense and the Edgerank problem, the filter that determines what you read on Facebook

Freedom, critical sense and the Edgerank problem, the filter that determines what you read on Facebook

The anthropologist Robin Dunbar had already verified, after measuring the size of hunter-gatherer groups in different corners of the planet, that humans can only create deep emotional ties with an average of 150 people.

Naturally, within this set there are several layers: the most intimate layer is made up of up to five people, the next can have up to ten, then there is another layer of 35, and finally one of 100.

On Facebook, however, that whole theory was shattered. Although there were always closer layers (usually made up of people we already know in the real world, such as friends and family), the user must continually choose who to pay more or less attention to . And that decision has to be dynamic: change according to each moment.

The way to facilitate this task has been a filter adapted to your particular interests: Edgerank . The bottleneck of our cognitive and emotional attention.

Echo chamber

Gen X, hippies, mods, hackers… the conceptual appeal of the labels is undeniable. Although its explanatory value is low or even null, people not only liked to be part of a group, to be labeled in some way, but the label even intervenes in their own way of living and acting. Like a self-fulfilling prophecy . Like the fish that bites its tail.

An example of this is the tradition of the Ashanti people in central Ghana of giving each child a spiritual name based on their day of birth. The Kwadwo, for example, were the children born on Monday, and they considered themselves peaceful, calm and withdrawn. The Kwaku, who were characterized by their rebellion, were children born on Wednesdays. In the 1950s, psychologist Gustav Jahoda discovered that this early categorization had a long-term impact on the self-image and lives of Ashanti children. The children simply adjusted to the etiquette, to what was expected of them.

People organize themselves in social networks based on homophilia (love of equals), that is, the conscious or unconscious tendency to associate with subjects who look like them, either because they share their interests, stories or aspirations. As Rutgers University human behavior researcher Helen Fisher wrote in her book The New Psychology of Love :

Most men and women fall in love with individuals of the same ethnic, social, religious, educational, and economic backgrounds, who have similar physical attractiveness, comparable intelligence, similar attitudes and expectations, values ​​and interests, and social skills and abilities. analog communication devices.

This way of organizing ourselves also defines us, both endogenously and exogenously . However, unlike the real world, on Facebook we do not interact with tens or hundreds of individuals, but with thousands or hundreds of thousands. We cannot give our full cognitive and emotional attention to everything that appears in our News Feed. For this reason, Facebook makes that decision for us taking as a reference our behavior (our interactions, our ties with others, the "likes" that we indicate, etc).

In 2011 Facebook stopped using the EdgeRank name as its system evolved into a more complex algorithm based on machine learning . As of 2013 Facebook engineers have reported that more than 100,000 individual signals are used to determine whether a post will be displayed on each user’s wall.

The problem is that most of these filters are based on homophilia: getting together with people who look like us, especially on an ideological level. What ultimately results in an ideological echo chamber, a Facebook Mind, which is a propitious terrain for fake news or extreme political polarization, as well as a diversity trap .