Big data, micro work: a labour geography of AI's invisible infrastructure workers in Canada and Tunisia
Type: article
Abstract: Microwork is the systematically hidden "clickwork" that trains and sustains the algorithms that drive Artificial Intelligence.This work is outsourced via digital labour platforms to "anyone, anywhere" for pennies per click in a model that mirrors outwork of the industrial era.Scientific management seems at its peak in microwork where, in training the algorithms that will eventually replace them, microworkers are simultaneously training the algorithmic managers of their work: by inscribing their affective judgments they refine the very statistical engines poised to automate that judgment, performing a form of self-erasure disguised as co-creation.While AI systems are increasingly framed as agential-capable of decision-making, adaptation, and problem-solving without human intervention-microworkers' agency is almost entirely eroded.This study brings together geographers' mappings of the structures of power in the data economy with concerns for the geographically and historically situated experiences of microworkers in a labour geography of microwork in Canada and Tunisia.Considering the different colonial histories and global economic positions of these sites, this study examines the forms of precaritization that bring people to this work as well as the precarity of the work itself, paying particular attention to how these factors cut across gender, race, class, as well as spatial scales.While narratives of the death of distance invoke a flattening of space through the internet, data infrastructures rely on sociospatial differentiations in a coproduction of uneven spaces on and offline.In this light, this study explores workers' digital spatial praxis.In Canada, workers' engagement in online forums act as infrastructures of networked perseverance: they help workers overcome platform opacity yet simultaneously recast microwork as a productive hobby, normalizing participation despite its limited returns.In Tunisia, supported by shared commiseration in online spaces and family support in their home spaces, many workers exercise a kind of networked escape, wherein they reject microwork altogether for failing to meet their employment needs.Yet, these modes of agency remain precarious and unevenly distributed within both groups.
Source: Summit (Simon Fraser University)
Author: Sioufi, Véronique
Institution: Simon Fraser University
Language: English
Subfield: Sociology and Political Science
Field: Social Sciences
Domain: Social Sciences
Open Access status: green
OpenAlex ID: W7139552556