Microstock images of artificial intelligence: How AI creates its own conditions of possibility

Abstract

The main goal of this paper is to account for the ‘algorithmization’ of microstock imagery. By this term, the authors refer to a material process implying the chronic use of graphic editors, semi-automatic keywording allowing complex and dynamic proto-classifications, and access to the images via search engines. The algorithmization of microstock imagery also goes along with the exploitation of producers’ labour, so that the authors recognize in it a form of digital labour. Moreover, the term ‘algorithmization’ is meant to underline that this material process has symbolic effects on the image contents as well as on people’s expectations and imaginaries of these contents. The paper analyses, in particular, the case study of microstock images depicting artificial intelligence (AI). By producing hundreds of thousands of visual representations of AI that spread via the Web and beyond it, algorithmized microstock imagery also produces its own symbolic conditions of possibility, that is, the expectations and imaginaries that contribute to the success of AI beyond its concrete effectiveness. The paper is structured into three sections. In the first section, the authors account for the existing literature on stock imagery. They contend that this literature focuses too much on the symbolic message, and too little on the material processes of production of these images. In the second section, the authors describe an empirical analysis they conducted on Shutterstock images depicting AI. In the third section, they distinguish three forms of digital labour and show that microstock imagery entertains resemblances to and differences from each form. They contend that despite its peculiarities, microstock image production is a paradigmatic form of digital labour due to its convergence towards algorithmization. In the conclusion, the authors show how, for microstock images depicting AI, the algorithmic loop of microstock imagery is complete.