Social media algorithms on X promote content that conflicts with your values
A new study reveals that X's recommendation algorithm actively promotes content that conflicts with users' personal values. This mismatch is driven by users arguing with posts they dislike, creating a frustrating feedback loop that disproportionately affects left-leaning users.
An MIT and Stanford University research team found that X's social media algorithm actively promotes content that conflicts with users' stated values. The researchers suggest this may disproportionately affect left-leaning users and may be driven by users replying to content they dislike. The study, published in the Proceedings of the National Academy of Sciences, examined the relationship between users' explicit values and the values promoted by the platform's algorithm.
Researchers used a custom browser extension and tracked the content users saw on X's "Following" and "For You" feeds. The algorithm consistently promoted content that was misaligned with users' stated values, especially in cases of self-transcendent values like caring for others and dependability. The algorithm showed a preference for content expressing individualist values over collectivist ones.
It also promoted posts expressing a need for conservation and order, while suppressing posts expressing self-transcendent values. The researchers found that users' replies to content were less aligned with their personal values than liking or reposting. This confrontational behavior, especially among Democrats, may create a feedback loop where the algorithm promotes more divisive content in response to user engagement.
Written by urgent.news from PsyPost's reporting — not their text. Machine-written — may contain errors; check the original before relying on it.
