Social Media Naturally Creates Fake News

What is fake news? How do we define what is real? The construction of truth is a critical topic for analysts of all sorts. However, there is now more content being created on the internet every second than any person could reasonably consume in their lifetime.
Social media, monetization, fake news, and censorship
While we may not be worried about the need for critical examination of claims in cat videos, we certainly do care in other about establishing the truthiness of statements related to topics with personal and social implications. Topics with such implications are often political in nature and so such examinations often go hand-in-hand with these topics. As analysts, even if our goal is to remain objective, we often find ourselves put in a position to critically examine these topics via data mining, statistical analysis, or citizen journalism. However, if we are to scale our efforts we need to better understand the means by which such forces can be perpetuated.

Concern over this topic has been ennobled by social media giants such as Google, Facebook, etc. using approaches such as Snopes, social feedback, source monitoring, etc. However, I find it ironic that such companies position themselves as independent curators in such circumstances. While they desire to come across as unbiased the very platforms these companies have created systematically support both the cultivation and propagation of fallacious ideas. Perhaps it is for this reason that creators of digital platforms have begun attempting to right the wrongs their networks have enabled.

How does social media cultivate fallacious ideas?
There are primarily three ways that digital platforms perpetuate fallacies:

  1. Illusory perception of consensus.
  2. Inappropriate application of social constructionist frameworks.
  3. Reinforced perpetuation of incredulity.

This is probably not an exhaustive list- feel free to add to it in the comment. Also, I am only going to briefly provide some examples of how each of these concepts applies and may add more over time.

1. Illusory perception of consensus.
Essentially this takes place for two reasons. First, on average, your friends have more friends than you do. This may seem contradictory but it is true because the mean and median number of friends on a social network are substantially divergent. Such divergence takes place in exponential or power-law distributions. A few individuals on social networks are responsible for a majority of the social connections. This effect is accentuated on networks that do not force reciprocity such as Twitter and Instagram. This effect is also accentuated by local biases in a users' community; as users accrete connections they tend to do so in homogeneous fashion.

  • Feld, Scott L. (1991), "Why your friends have more friends than you do", American Journal of Sociology, 96 (6): 1464–1477
  • Hodas, Nathan; Kooti, Farshad (May 2013). "Friendship Paradox Redux: Your Friends are More Interesting than You"

2. Inaccurate application of social constructionist frameworks.
Ways of thinking and ideas that have practical value typically have applications in applied reality. For example, an effective way of assessing causality will lead to value for a business and to employment for the individual maintaining that way of thinking. In my experience it seems that people with more effective worldviews tend to find themselves increasingly strapped for free time since the opportunity cost for their time has increased. So, the ability for people with successful worldviews to contribute to conversations on social media is diminished relative to their lack of free time. Thus, the content found on social media is confounded with practicality. This manifests itself in three primary ways:

  • There is less support for practical ideas on social media than is present in the real-world.
  • Many practical ideas present in the real-world often do not appear as options on social media. This is accentuated on smaller social networks where the tendency for an echo chamber increases due to the lack of diversity in thought.
  • People with practical worldviews have often honed a level of diplomacy that causes their ideas, when presented, to come across less forcefully. Individuals with high opportunity costs for their time have more to loose and it is more important that they communicate their ideas in less combative fashion.

3 Reinforcement of incredulous ideas
In addition to the biases in audience composition and engagement mentioned above, there are also biases in content creation. It is harder to define quality metrics than it is to define reach metrics. Thus, digital platforms primarily monetize content based upon reach. Engagement, if it effects monetization at all, often only does so as an afterthought (second-order factor). This reinforces content that evokes an emotional response in platform users. Content that causes surprise, anger, fear, arousal, or disgust is much more potent than content that encourages critical thought. This encourages bipolarity in content creation. Content becomes increasingly focused upon in-group norms (ie. virtue-signaling) or upon incredulity (ie. click-baiting).

Where do we go from here? As analysts it is useful to be able to identify fallacious content. However, it is even more helpful if we can identify the mechanisms behind the creation of such content. If we can do the later then we are one step closer to being able to automatically identify this content using eg. algorithms.