ABSTRACT: Despite tremendous positive potential, social platforms currently disrupt important social processes and exacerbate offline problems, causing harm to users. For example, social platforms introduced opaque algorithmic actors in the form of feeds that obscure who sees what content online. By obscuring this crucial audience information, these algorithmic feeds disrupt how we make decisions about how to present ourselves online. The result: context collapse, embarrassment, and damage to personal and professional relationships. Cases like these require users to quickly adapt around computationally-imposed problems, and put platforms in the position of needing to quickly mitigate these issues. However, neither party has sufficient understanding of the other to do so: users have insufficient knowledge of the platform, and platforms have incomplete understandings of the nature of the disruptions and how users are actively adapting in real time. The challenge is to close the knowledge gap between people and platforms, to help both users and platforms better understand each other and the problem itself.

In this talk, I will show how users attempt to adapt to platforms in the context of online self-presentation behavior by forming folk theories, or informal and quasi-causal theories of what platforms are and how they operate which guide user decision making. I will show how approaching such problems through a folk theories lens helps us update key social theories to account for modern, computationally-influenced sociality, while also providing us with immediate design guidance for platforms, and potential intervention points for future efforts to promote algorithmic literacy. I will conclude by discussing how the folk theorization lens and related methods can improve our understanding of the person/platform relationship, identify user/system disconnects, and explain how to better support learning, adaptation, and literacy in a constantly-changing platform environment.

BIO: Mike DeVito (she/they) studies how users and communities adapt to the challenges that ever-evolving, algorithmically driven technology introduces to social and informational processes. Their work contributes HCI-relevant theory, platform guidance, and empirical/community-based critique, along with methodological innovations. Mike is currently a doctoral candidate in Media, Technology, and Society at Northwestern University, advised by Jeremy Birnholtz, and frequently publishes in and organizes for ACM venues, such as CHI and CSCW. Their current research includes explorations of how social media users employ folk theories of algorithmic feeds to guide their behavior and determine self-presentation strategy, and how the queer community adapts to local sociotechnical infrastructure which can complicate identity disclosure and exacerbate intracommunity conflict. Additionally, Mike serves as chair of the CHI Queer Special Interest Group and a member of the CHI Transparent Standards Group.