Recent discussions and movies like Social Dilemma tend to emphasize the second aspect. However, two papers written before these events by Aimdyn researchers in collaboration with Criminal Justice professor Jason Gravel from Temple University showed both aspects are the basic ingredients for more frequent and larger insurgencies and political violence based on mathematical models and algorithms used in Artificial Intelligence (AI).
“Algorithms on social media platforms are able to “optimize” homophily—a powerful natural human tendency to associate with similar other—by suggesting content that your friends like even before those friends share that information with you. By sharing content an algorithm has already determined your friends are likely to be interested in, you give legitimacy to that content and to your friends’ interest in that content,” said Jason Gravel, Assistant Professor of Criminal Justice at Temple University.
“Information and misinformation alike can rapidly permeate pockets of individuals with similar views, unencumbered by geographic, or time constraints. Soon things become “obvious” in some social circles, no matter how farfetched they may seem from the outside looking in. Under these conditions, little to no coordination among actors in a network is necessary for large scale mobilization around an issue,” Gravel said.
The models also suggest a possible remedy: a balancing of suggested content and suggested network associates, similar to the fairness doctrine of the United States Federal Communications Commission (FCC), introduced in 1949 and abandoned in 1987.
The results could thus be used to inform future decisions made by the tech giants in order to move in a direction less susceptible to huge political outbursts.
Aimdyn Inc is an AI and ML Research company based in Santa Barbara, CA. It has a long history of ML and AI development and analysis with predictive capabilities, applied to large societal issues, from damage control for the Gulf Oil Spill to mapping out the spread of the current COVID-19 pandemic.
To learn more about Aimdyn Inc. and these projects, visit www.aimdyn.com
M. Fonoberova, I. Mezic, J. Mezic, J. Hogg, J. Gravel. “Small-world networks and synchronisation in an agent-based model of civil violence”, Global Crime, 20:3-4, 161-195, 2019. https://doi.org/10.1080/17440572.2019.1662304
M. Fonoberova, I. Mezic, J. Mezic, R. Mohr. “An agent-based model of urban insurgence: Effect of gathering sites and Koopman mode analysis”, PLOS ONE 13(10): e0205259, 2018. https://doi.org/10.1371/journal.pone.0205259