Por Vinícius Jonas de Aguiar (CFCUL/GI3).
This paper investigates to which extent and in which ways data sciences, more specifically recommender systems powered by Artificial Intelligence (AI), are transforming music aesthetics. The interplays between music and AI are not new in history. But while historically, from Athanasius Kircher to Magenta, the focus has been the automatization of creativity, I argue that a new horizon of interplays between music aesthetics and AI has emerged with the massive popularization of streaming platforms powered by recommender systems. I provide an overview of how the task of music recommendation powered by AI is transforming music aesthetics in three dimensions: epistemological, normative, and phenomenological. I argue that music recommender systems are characterized by an interplay between (i) theory and modeling of aesthetic objects, (ii) generation of aesthetic outputs, and (iii) the eliciting of aesthetic experiences. I conclude with some brief remarks about the limits of a music aesthetics based on inductive-deductive inferences and about the lack of abductive reasoning in music recommender systems.
Transmissão via Zoom
