Information Cascades and Echo Chambers
Coming off news of Russian interference in 2016 Presidental Election, there was a Senate Intelligence Hearing on Capitol Hill that had representatives from Google, Facebook, and Twitter. The topic discussed concerned the tech companies’ roles in curbing hate speech while also protecting free speech, as many politicians were aggravated over the Russian government being able to buy ads on these social media sites, and how there seemed to be a “liberal bias” exhibited by the companies. While Colin Strech, representative of Facebook, argued that algorithms will help solve problems of hate speech and biases, there seems to be evidence that these algorithms do more harm than good.
Psychologically speaking, it’s been proven that humans tend to be in groups in which the people have characteristics similar to themselves (“Social Identity Theory” Tajfel and Tuner (1974)), and, in the age of social media, social networks reflect this observation at large scales. How information is passed through these homogenous groups can be understood with information cascades, and the issue with information cascades is that people tend to ignore new information and base their decisions on the decision on others. This could mean that these information cascades can form from faulty or biased information, and, in the process, they can create conformity in a network; making them homogenous.
The issue here is that social media companies expolit these networks in order to keep users on their sites longer and generate more revenue from their advertisers, which is why the algorithms tend to show users news and information they would tend to agree with. In essence, echo chambers are created and new and challenging ideas fail to reach a wide network of people, which only creates more division and hence why algorithms are part of the problem and not the solution.
How this article relates to what we learn in class is that it talks about the role information cascades play in creating echo chambers. This is because in an information cascade, people make decisions one at a time, and they make their decision based on what other poeple before them decided. Since we tend to group around people we are similar towards, if others decide that certain information is truthful, we tend to believe the information is true as well despite the fact that there may be more reliable data that proves otherwise. This is one of the reasons why falsified information can spread so quickly and be deemed truthful, and why other information can be seen as “fake news” despite their legitimately.