2026/02/21 by Ayaj Rana · 1 voice
Social Sciences · #Digital Economy and Work Transformation #Political Economy and Marxism #Geographies of human-animal interactions
paper · pdf · doi:10.1177/00197939261426904
openalex publication_date 2026/02/21 · openalex created_date 2026/02/22 · openalex updated_date 2026/07/15
Advances in artificial intelligence (AI) have sparked intense debates about automation and the future of work, yet often overlooked is the human labor quietly fueling these technologies.Feeding the Machine: The Hidden Human Labour Powering A.I. by James Muldoon, Mark Graham, and Callum Cant directly addresses this oversight.The authors deliver a timely exposé of the workers and workplaces behind AI's glossy façade.For scholars of labor and employment, the book's focus on how AI systems rely on global networks of precarious labor makes it a highly relevant contribution.It situates AI not as a purely technical marvel but as a phenomenon deeply embedded in patterns of labor exploitation, a perspective that enriches ongoing discussions in labor studies about the gig economy, platform work, and the political economy of technology.At its core, Feeding the Machine argues that AI is not a futuristic autonomous force but instead an "extraction machine" built on human toil.Drawing on more than 200 interviews conducted across more than a decade of research, the authors reveal a vast, intricate network of labor that underpins AI-a network that spans from low-wage data annotators to well-paid engineers, all operating within a global division of labor reminiscent of colonial-era exploitation.The book connects the hidden workforce of AI to long-standing histories of gendered, racialized, and geographic inequalities, asserting that today's AI boom echoes earlier industrial and colonial labor regimes.Rather than organizing the material by technology or industry, the authors structure the narrative around seven representative human figures in the AI production pipeline.These include the annotator, the engineer, the technician, the artist, the operator, the investor, and the organizer.Each figure is drawn from real-world cases and ethnographic interviews, providing a window into a different link in the AI labor chain.Through these portraits, the book's overarching narrative illustrates how disparate workers are interconnected in serving the AI industry's needs.For example, we meet a Ugandan data annotator performing monotonous data-labeling tasks for pennies.By contrast, an AI engineer in a gleaming tech hub enjoys comfortable pay and perks, yet her creative efforts to build new algorithms ultimately depend on the invisible labor feeding the data sets behind the scenes.The organizer figure, a Kenyan content moderator-turned-activist, exemplifies resistance: After enduring traumatic work filtering toxic online content, he helps spark the formation of Africa's first union for AI data workers.These and other accounts (including an Irish voice artist who discovers her voice was cloned by an algorithm without consent) vividly demonstrate the human faces of AI's supply chain.Crucially, the authors show that the actions of one set of actors can drastically affect others: A Silicon Valley investor's push for profit and speed trickles down to tighter deadlines for annotators, and an Amazon warehouse operator is managed by AIdriven systems designed by faraway engineers and sanctioned by corporate strategists.The central thesis is thus clear and powerful: AI is a socio-technical system sustained by exploited human labor around the globe.Feeding the Machine is a valuable and provocative study, but it is not without its shortcomings.A first critique is that several chapters devote excessive attention to technical and systemic descriptions at the expense of fully developing the workers' lived experiences.The authors' background explanations-for example, a detailed digression into Amazon's warehouse 1426904I LRXXX10.