Artificial intelligence is widely promoted as a promising solution to healthcare challenges, yet its clinical benefits still remain uncertain and its risks persistent. Drawing on extensive fieldwork with researchers developing machine-learning applications for healthcare, this talk demonstrates how practitioners sustain AI’s legitimacy by framing their work as serving the common good. It is argued that “moralization”—the work of aligning technological projects with values such as equity, care, and wellbeing—generates consensus and momentum for technological solutions to social problems amid uncertainty. At the same time, the moralization of machine learning risks obscuring technical limitations, diffusing accountability, and diverting attention from alternative approaches. Using healthcare AI to illuminate broader transformations in scientific authority as knowledge production becomes increasingly applied, interdisciplinary, and problem-oriented, it is shown how technical authority is established not through claims of neutrality or detachment, but by building relationships across institutional and professional boundaries.
Bio: Shira Zilberstein is a sociologist whose research investigates the moral and organizational foundations of knowledge production, with particular attention to how science and technology are used to define and address social problems. Her current work focuses on the development and governance of AI models for healthcare. Using qualitative methods, she studies how researchers and practitioners justify and seek to ensure the social value of their work. She is especially interested in how research communities navigate uncertainty and sustain commitments to technological innovation even in the face of contested outcomes and limitations.
Shira holds a Ph.D. in sociology from Harvard University, where she was a Science and Technology Studies Fellow and a BA in sociology and history from Northwestern University. Her research has been supported by the National Science Foundation, the American Sociological Association, the Russell Sage Foundation, and the National Academies of Sciences, Engineering, and Medicine, among others. She has collaborated with researchers at MIT, Duke, and Yale on projects related to the development and governance of healthcare AI. At Princeton, she is also affiliated with the Princeton Societal AI Initiative.
This talk will not be livestreamed or recorded.
If you need an accommodation for a disability please contact Jean Butcher at butcher@princeton.edu.