Social media algorithms, designed to boost user engagement for advertising revenue, amplify the biases inherent in human social learning processes, leading to misinformation and polarization. As humans naturally learn more from their ingroup and prestigious individuals, algorithms capitalize on this, pushing information that feeds these biases—regardless of its accuracy.
In antiquity, knowledge dissemination predominantly occurred within our social circles or from esteemed individuals, guaranteeing credibility and collective progress.
However, in our intricate modern society, particularly within the realm of social media, these inherent biases have eroded. The credibility of online connections is often questionable, and the façade of prestige is prevalent.
A recent study elucidates how social media algorithms have diverged from our intrinsic cooperative instincts, contributing to widespread polarization and the propagation of misinformation. Published in the journal “Trends in Cognitive Science” on August 3rd, this research underscores the need for a realignment between algorithmic mechanisms and human sociopsychological tendencies.
This exploration into social learning provides a framework for comprehending the interplay between human psychology and algorithmic operations. Historically, our cognitive inclinations favored knowledge acquisition that bolstered cooperative problem-solving and societal harmony. This led individuals to gravitate towards learning from those within their social spheres or those deemed prestigious.
However, contemporary algorithms have shifted their focus towards content that enhances user engagement, thereby inadvertently exacerbating existing human biases. These algorithms can inundate social media feeds with a specific type of information termed Prestigious, Ingroup, Moral, and Emotional (PRIME) content, regardless of its accuracy or representation of diverse viewpoints.
Consequently, polarizing and extreme political narratives are amplified, often isolating users from dissenting perspectives and perpetuating misconceptions regarding prevailing group opinions.
William Brady, lead author of the study and a distinguished social psychologist at Northwestern University’s Kellogg School of Management, explains that the discordance between algorithmic objectives and cooperative human nature is not intentional. Rather, the misalignment arises from differing goals, which ultimately contribute to unintended adverse effects.
To address these challenges, the research group proposes a twofold approach. First, there is an emphasis on augmenting users’ awareness of algorithmic functionality and content selection. Despite social media platforms typically maintaining a degree of opacity in revealing the intricacies of their algorithms, potential solutions include providing users with insights into why specific posts appear on their feed. This could involve explaining whether the post gained visibility due to user engagement or overall popularity.
Additionally, the research team is actively developing interventions aimed at fostering a more discerning consumption of social media content. By empowering users with tools to navigate the digital landscape more astutely, this initiative seeks to mitigate the unintentional consequences of algorithmic biases.
Furthermore, the researchers propose recalibrating social media algorithms to prioritize community-building and diverse content presentation. Rather than disproportionately amplifying PRIME information, algorithms could be designed to strike a balance that encourages user engagement while curbing the proliferation of polarizing or extremist content.
In conclusion, the study underscores the imperative for a harmonized relationship between human psychology and algorithmic operations within the contemporary digital panorama. This recalibration holds the promise of preserving the essence of communal engagement while safeguarding against the amplification of undue polarization and misinformation.