Colonialities of Machine Learning

Abstract

Abstract This chapter discusses the relationship between machine learning (ML) and coloniality, which refers to enduring colonial-like power imbalances between the Global North and Global South that continue beyond the official end of colonialism. First, it explores how postcolonial theory can illuminate the colonial aspects of ML by showing how ML usage is entwined with data extraction and underscoring the importance of recognizing cultural diversity within the various locales that contribute to ML’s design and implementation across divides between the Global North and Global South. Second, it considers how the field of ML can prompt a re-evaluation of critical elements within postcolonial sociology. The chapter contends that feminist standpoint theory—a foundational influence on postcolonial sociology—attributes an epistemological privilege to oppressed groups. This premise becomes problematic when considering the operational dynamics and structure of ML.