The (Brazilian) factory floor of artificial intelligence: Data production and the role of communication among appen and Lionbridge workers

Abstract

The article analyzes Brazilians' work in global artificial intelligence (AI) platforms, especially Appen and Lionbridge, through exploratory research. We highlight the role of communication and data production for AI, outlining labor conditions and complexities as the factory floor. Thus, we interviewed workers and conducted observations in WhatsApp and Facebook groups. We divided the findings into five categories according to workers' perception: hiring processes, time tracking and difficulties proving hours worked, lack of infrastructural reliability, work strategies and definition, and, finally, understanding of AI. Contrary to digital labor universalism, there is geopolitics in AI platform labor. The material conditions of data production and the role of communication among workers stand out as emerging and complex solidarities, as fissures and cracks, to devise strategies that deal with AI platform labor.