Beyond the Call for Transparency. YouTube's Recommandation Algorithm Reconsidered.

For some years now, YouTube's algorithm has been criticized for its lack of transparency. A considerable body of research has taken up this criticism in an attempt, from a media-educational perspective, to uncover the functioning and ideological biases that govern the selection of targeted outputs on users' screens. Starting from this imperative of transparency, i.e., of making the algorithm’s operating principles visible, this article aims to trace the contours of another take on the algorithm, one that no longer calls for the vizibilization of its mechanisms, but reconsiders it, beyond its opacity, as a producer of visibility of new iconological links between the materials it organizes.