The Ghost Workers of Artificial Intelligence
Author: Julian Posada, email@example.com
Department: Faculty of Information
University: University of Toronto, Canada
Supervisor: Alessandro Delfanti
Year of completion: In progress
Language of dissertation: English
, social networks
, artificial intelligence
Areas of Research:
Communication, Knowledge and Culture
, labour market
Artificial intelligence does not exist. At least not in the sense of fully intelligent machines that are capable of judgement. Rather, contemporary AI systems have narrow capabilities and rely heavily on human labour outsourced via online platforms. Even virtual assistants and self-driving cars depend on invisible workers that operate from their homes where they provide, label, and correct data destined for machine learning. My research analyzes the work experience of outsourced workers from Canada and Colombia with a focus on the essential support of their social networks, including family, colleagues, and friends.
Labour platforms specializing in AI deny the legal rights conferred by traditional employment to their workers, a condition that aggravates the digital divide between the global north and south, along with the situation of workers struggling with intersectional identities of race, gender, class, or ethnicity. Research further suggests that while platform intermediation alienates workers from their peers, these workers are also embedded in networks of trust as many depend on domestic labour and support. Thus, instead of enduring the problems of labour commodification alone, their social networks could play a vital role in—and outside of—their workplace.
To what degree the workers’ networks configure value production in AI development and their effects on the nature of their experience remains largely understudied. Thus, my doctoral project investigates how these networks impact the labour process and the social reproduction of workers, notably the organization of work and the transmission of social inequalities.