Generative Artificial Intelligence, “perfect” bodies and aestetic coloniality

Abstract

In this article, we address the generation of images by artificial intelligence systems by problematizing the way they tend to maintain a colonial logic in their computational programming and in their generated products, a logic from which are crystallized, for example, some representations of beauty and perfection that are unreal and based on a white, European and hegemonic reference. In order to do so, we mobilize theories about discourse and the body and we discursively analyze six images, all generated by such systems. We discuss the conditions of possibility in which they are generated, and their effects. Taking this analysis as an example of a problematic form of programming and considering that artificial intelligence systems can function as control and manipulation devices against users, we point to the need to include direct and detailed discussion about artificial intelligence in the language classroom, in order to effectively respond to critical linguistic education orientation.