Algoritmos de feeds fazem texto? Um ensaio semântico-enunciativo sobre o Youtube

Abstract

Feed algorithms are responsible for the digital distribution of billions of media contents across the internet, primarily through major platforms. By processing massive amounts of data and variables, such as a user’s digital traces combined with data generated by the browsing of billions of other users, these algorithms ultimately create unique media displays for each user session of every registered user, with noticeable trending patterns. Although algorithms are computationally automated, their outcomes depend on other agents, such as the programmers who regulate their parameters, and the users themselves. What if, then, each of these unique displays could be considered as text or enunciation? What would be the implications and new possibilities for linguistic understanding? By choosing YouTube as its object, this essay seeks to describe the dynamics present on the platform and understand the possible meanings that algorithmic curation can produce. The theoretical foundation for the analysis was Semantics of Enunciation. Other theoretical perspectives, such as Discourse Analysis and Textual Linguistics, were also used for comparison and to explore potential convergences. This multifaceted approach was deemed necessary particularly due to the complexity of informational technology development in recent decades and its corresponding emerging social impacts.