How to train your algorithm: The struggle for public control over private audience commodities on Tiktok

Abstract

Social media users are increasingly aware of the politics of their viewing habits, and they attempt to express these politics through interactions with proprietary algorithms. Combining theories about audience commodities with scholarship about “algorithmic imaginaries,” I define “algorithmically imagined audiences” as a kind of algorithmic imaginary, and I analyze 103 TikTok videos to explore how people attempt to politically engage with algorithms to position themselves within audiences. Although algorithms and audiences are proprietary, TikTokers believe they can reassert public control over audience commodities to engage in counterpublic world-making and to re-position themselves within imagined communities. While these practices are impactful, they have conceptual and practical limits; these same tactics are used to reprivatize audience commodities and to reinscribe the neoliberal capitalist underpinnings. This article raises questions for future researchers about the opportunities and limits of sociotechnical beliefs.