New research by computer scientists from Cornell University and the University of California, Berkeley (UC Berkeley) uncovers major changes in Twitter’s algorithms, exposing users to a greater number of tweets that increase rage and hate. The algorithm change coincided with Elon Musk’s takeover of the social network.
The study examined the content appearing on Twitter’s customized timelines, known as “For You,” of 806 users in February. It then found a significant emphasis on “emotional content” in tweets shown to users, regardless of the original purpose. As a result, as compared to earlier encounters, users were more likely to respond strongly to the information. Anger, pleasure, sorrow, and worry were the most frequently expressed emotions in these tweets, with emphasis on anger.
In addition, the study discovered a link between the display of political tweets via Twitter’s algorithms and a shift in user behavior. This shift frequently resulted in an increase in othering behavior, creating unfavorable sentiments about groups of people in society who have opposite viewpoints. The study also discovered that the majority of Twitter users rely on the algorithmically produced newsfeed. As a result, individuals are more likely to follow suggestions for particular sorts of users, which influences the information they see on the chronological timeline.
The implications of these algorithmic changes extend beyond the Twitter platform, as they can influence user behavior and attitudes in the broader society.
The sources for this piece include an article in EveningStandard.