ALGOFEED Project

In the contemporary digital society, consumer cultures and practices have been deeply re-mediated and re-configured by the complex sociotechnical systems of digital platforms. In particular, the consumption of cultural entertainment products such as movies and music has almost entirely been re-mediated by the platformed infrastructures of Netflix, Spotify, YouTube, and the like (Poell et al., 2022). These feature AI models and algorithms that “learn” from users’ consumption patterns in order to recommend, filter and rank content in dynamic and highly adaptive ways (Airoldi 2022). Since platform users largely rely on recommender algorithms to decide what to watch, listen or read next, and recommender algorithms analyse users’ datafied behaviour to produce automated recommendations, such digital platforms see the proliferation of “feedback loops” – that is, recursive systems whose outputs affect the inputs of a new iteration (Wiener 1989). Machine learning models attempt to capture users’ preferences without accounting for the effect of their recommendations and, as a result, input data are systematically “confounded” by output results (Chaney et al. 2018). Platform-based feedback loops are likely to have disruptive consequences on our increasingly “algorithmic” consumer culture (Hallinan and Striphas 2016). Scholars indicate that user-machine feedback loops can lead to the path-dependent reinforcement of past consumption patterns, thus resulting in “filter bubbles” (Pariser 2011) and forms of “automation of taste” (Barile and Sugiyama 2015). Still, there is a lack of sociological research providing an empirical assessment of these largely speculative hypotheses, due to epistemological issues in the study of proprietary algorithms and their social implications (Pasquale 2015).

The ALGOFEED project aims at filling this gap. Employing a mixed-methods research strategy, its overarching goal is to shed light on the socio-cultural effects of platform-based feedback loops. The main research question the project asks is: how does the automated recommendation of content on digital platforms affect users’ individual and collective cultural consumption patterns over time? As a result, the three main objectives the ALGOFEED project will achieve are: 1) to produce unique empirical findings about platformized cultural consumption in Italy, 2) to develop a new methodological and theoretical framework for the sociological study of algorithmic consumer culture, and 3) to increase the algorithmic awareness of Italian consumers by engaging in dissemination and impact activities according to a transformative paradigm.