Machine Learning as a State-building Experiment

Abstract New initiatives of machine learning (ML) have become sites of controversies about technoscience, innovation, and global inequalities in Africa. As part of the burgeoning area of artificial intelligence for development, ML is often framed as a technological answer to long-standing debates about development and modernization in African societies. This chapter examines the embedded assumptions and political imaginaries of Africa’s globally marginalized epistemic communities in understanding the social impact of ML in the continent. Looking at these debates through the analytical lens of sociotechnical imaginaries, decoloniality, and alternative modernities, the chapter argues for an understanding of ML as a state-building experiment in postcolonial Africa. From this perspective, this chapter begins to articulate a different theorization of sociotechnical practices of ML from a Majority World perspective. It also complements critical literature on the sociology of ML by integrating perspectives from science and technology studies and contemporary African studies.