The spectacular developments in ChatGPT and artificial intelligence (AI) seem to dazzle us all, filling us with a mixed sense of dread and opportunity. Behind the great announcements lies a plot that is all too familiar to those of us who inhabit territories that have been colonised time and again: micro jobs with minimum wages, disregard for social rights, and the growing tendency to treat individuals as if they were machines. A feminist community-based perspective is needed to criticise and expose this model that undermines rights.
Omar (27) is a migrant living in Germany who has been a “data worker” for some years now. To describe the current workspace, Omar uses an image familiar to many: a call centre. The job opportunity came “by chance,” through a network of contacts, Omar recalls, adding how exciting the prospect of “making the internet a better place” seemed then. Now, five years later, the worst thing in Omar’s life is this job. “We are numbers to them and they completely forget there are individuals behind those numbers. We work with sensitive content here and we do not receive the treatment that is needed.”
“Data workers” are individuals who occupy positions connected with data annotation and training platforms for artificial intelligence. To better understand this kind of work profile, we consulted Matheus Viana Braz, a researcher and professor with the Department of Psychology at the University of the State of Minas Gerais (UEMG) and coordinator of the Work, Health and Subjectivation Processes Laboratory (LATRAPS), who conducts studies on the precarious work behind the artificial intelligence production chain. “These workers perform microtasks that include categorising images, classifying ads, transcribing audio and video, evaluating ads, moderating social media content, verifying information, tagging anatomical landmarks, scanning documents, and improving search engine results, among others. The best paid tasks are connected with pornographic and violent content moderation in social media,” Viana Braz notes.
Continue reading at GenderIT.org.