Tanya Clement (U Texas), Andi Gustavson (U Texas), Allyssa Guzman (U Texas), Nathan Alexander Moore (CU Boulder), and Lauren Walker (U Texas) have published a white paper entitled “Good Systems Humanist-in-the-Loop: Responsible Data Operations and Workforce Development in Libraries, Archives, and Museums” on the University of Texas at Austin’s Texas Scholar Works repository. This paper describes the “Humanist-in-the-Loop” project, a “first-step attempt to create a workforce development program called ‘Ethical Data in Practice’ for future library and archives professionals that teaches them to engage in the work of ‘responsible data operations.'” As the paper states,
Cultural heritage institutions are increasingly interested in data practices, machine learning (ML), and artificial intelligence (AI) but are far from being able to use these practices ethically, efficiently, and cost-effectively. Without informational professionals trained in ethical data work, libraries and archives run the risk of replicating bias with new data-oriented practices. While graduate students in schools of information and library science are trained in the importance of digital cultural heritage and critical data studies and graduate students in the humanities are trained in the ethics of archival research, neither groups are given first-hand experience providing access to collections and engaging directly with data projects. It is imperative that we train a workforce of information professionals who will, in their future careers, be prepared to make ethical decisions about the data practices that shape our world.
This project and white paper build on the work of Safiya Noble and Ruha Benjamin, who have highlighted the ways in which algorithmic biases “replicate hierarchical and exclusionary practices within archives.” It is of particular interest to DH library workers who engage in data-intensive work or hope to create data-centered services.