Algorithmic Hollywood: Data, Production, and the Cinematic Imagination 2008- Today

Abstract

This dissertation investigates the centrality of algorithms in contemporary Hollywood cinema, exploring issues of industry, production, spectatorship, and narrative. Understanding the algorithm as an essential mathematical element of computation, as both a pragmatic process for problem-solving and as the central emblem of a broader episteme and ideology, the dissertation is structured according to case studies of different films and techniques in Hollywood that explore the wider significance of this technology and logic for cinema. The bourgeoning of algorithmic technologies since the turn of the 21st century has resulted in a cultural milieu shaped by tendencies towards prophecy and prediction, avoidance of risk, numeric rationality, personalization of consumption, enumeration of data, and a general sense of opacity in the way all these things operate. All these characteristics are found in Hollywood production today; furthermore, Hollywood has been instrumental to the emergence and shape of this milieu. In the first chapter, I explore the Netflix Recommendation Engine, an algorithmic method for personalizing and categorizing films to individual viewers, an apparatus which has engendered distinctive modes of exhibition and spectatorship. The second chapter chronicles the prehistory of the Recommendation Engine in 1940s experiments in data-centric audience research, arguing as to a genealogy of contemporary Algorithmic Hollywood in the Studio System. The final chapter further investigates the links between these two periods, arguing that the advent of algorithmic technologies has not caused a revolutionary disruption to the American film industry as was previously predicted, but has further consolidated historically persisting monopolies, practices, and styles.