Subverting the algorithm : examining anti-algorithmic tactics on social media

Abstract

As social media users have become increasingly aware of algorithms and their potentially negative effects, I argue that some users are challenging algorithmic systems of control. This thesis examines anti-algorithmic tactics on online platforms -- ways in which users actively aim to subvert algorithmic systems. I investigate anti-algorithmic tactics through two case studies of social media users responding to algorithmic content moderation and content curation. The first case study investigates how Black Facebook users have used alternative spellings to avoid detection by content moderation algorithms. The second case study investigates how users of Gobo, a social media browsing tool, have used tactics to minimize the influence of content curation algorithms on their social media feeds. In these case studies, I conduct close readings of public social media posts and interviews with social media users to better understand the perceptions around algorithms and motivations for anti-algorithmic tactics. Based on insights from these case studies, I conclude this thesis by discussing design frameworks that address concerns around transparency and user agency in algorithmic systems.