Resolving the problem of Algorithmic Dissonance: An unconventional solution for a sociotechnical problem
In this paper, I propose the concept of “algorithmic dissonance”, which characterizes the inconsistencies that emerge through the fissures that lie between algorithmic systems that utilize system identities, and sociocultural systems of knowledge that interact with them. A product of human-algorithm interaction, algorithmic dissonance builds upon the concepts of algorithmic discrimination and algorithmic awareness, offering greater clarity towards the comprehension of these sociotechnical entanglements. By employing Du Bois’ concept of “double consciousness” and black feminist theory, I argue that all algorithmic dissonance is racialized. Next, I advocate for the use of speculative methodologies and art for the creation of critically informative sociotechnical imaginaries that might serve a basis for the sociological critique and resolution of algorithmic dissonance. Algorithmic dissonance can be an effective check against structural inequities, and of interest to scholars and practitioners concerned with running “algorithm audits”.