AI Models and Their Worlds: Investigating Data-Driven, AI/ML Ecosystems Through a Work Practices Lens

When we invoke the “future of work,” to whose work do we refer? This paper considers everyday work practices through which contemporary artificial intelligence (AI) and machine learning (ML) ecosystems are made possible. The “future of work” is often talked about in relation to the anticipated domain settings where AI/ML systems might be implemented and the labor conditions such implementations might re-configure (foreseeing or noting changes in medical/health, legal, or manufacturing work, for example). This paper turns our attention to the various forms of labor that must be undertaken to conceive of, train/test, deploy, and ongoingly maintain AI/ML systems in practice. In particular, this paper draws on an ongoing ethnographic endeavor in a large, global technology and consulting corporation and leverages a work practices lens to examine three themes: curating datasets (everyday work practices of data pre-processing); tending models (everyday work practices of training, deploying, and maintaining predictive models); and configuring compute (everyday work practices of back-end infrastructuring, commonly called “DevOps”). This paper considers the value of a work practices lens in studying contemporary sociotechnical labor ecosystems. By locating the work practices through which AI/ML systems emerge, this paper shows that these technologies indeed require considerable human labor, at the same time they are often talked about as drivers of automation and displacers of work. This extends discourses around the “future of work,” giving light to the various standpoints and experiences of labor such imaginaries implicate and ongoingly re-configure.