The Duke Initiative for Science & Society:  

The Duke Initiative for Science & Society is an interdisciplinary initiative that examines the wide-ranging and integral role of science in social institutions and culture. Science & Society integrates and fosters innovation in related research, education, and engagement at Duke by adopting a pluralistic approach to understanding how science and human endeavors intersect with a specific focus on ethical, legal, and social implications for science. For more information, visit our website at http://www.scienceandsociety.duke.edu.   

The Fellowship: 

The Ethical Technology Fellowship is a residential fellowship that provides an opportunity to become involved in the activities of Pratt School of Engineering, Science & Society, Duke Law School, the Sanford School of Public Policy, and Duke University, while pursuing independent scholarship at the intersection of technology, law and policy. Candidates are offered a one-year position, with a possibility of renewal for a second one-year position.  Fellows are provided with office space, a competitive stipend, and benefits. Applicants should have a doctoral level degree in a relevant area to the Fellowship responsibilities listed below. The fellowship will be overseen by Pratt Engineering Professor, Michael Zavlanos and Lee Tiedrich, the Distinguished Faculty Fellow in Ethical Technology.

Fellowship Responsibilities: 

This Fellowship will study and report on data governance and risk management and impact assessments for artificial intelligence.   Research topics will likely include the following:

·         Designing and evaluating metrics for AI impact assessments, including how relevant impacts should be designed and measured, how results can be reported and shared, how frequently impact assessments should be conducted, and techniques for mitigating materially adverse impacts;

·         Designing and evaluating practices for data governance and enhancing data quality and interoperability, including methodologies for tracking data provenance and lineage and improving traceability, methods for measuring and reporting on data quality, and evaluating how to leverage synthetic data and possibly other advanced data techniques to improve data quality and to report on such efforts in a data governance model; and

·         Helping develop a data commons that will provide a platform for testing impact assessments, risk management, and data governance methodologies.

The successful candidate should have strong writing and communications skills and enjoy collaborating with experts, including in other disciplines, such as law and policy.