Center for Advanced Study in the Behavioral Sciences

Towards a Theory of AI Practice

It is commonly stated that AI needs perspectives from the humanities and social sciences. This is true but does not go far enough.

a young woman stands as a light installation projects streams of blue-white light flowing around her

Attending to humans in the loop does not sufficiently address the need to reflect the strengths and limitations of human cognition in the design of systems that become integral to our workflow and daily lives. It is time to rethink the field of applied AI from the ground up, founding it simultaneously in both the social sciences and computational/statistical sciences. This project facilitates the tools and collaborations needed to develop such a field.

This project is led by Jim Guszcza (CASBS fellow 2020-21) and co-led by Allison Stanger (CASBS fellow 2020-21) with the assistance of CASBS Director Margaret Levi and CASBS Program Director Zachary Ugolnik. The project also includes a core-working group as well as an advisory committee.

Toward a Theory of AI Practice is supported by the Rockefeller Foundation. This project developed from a previous CASBS collaboration with the Rockefeller Foundation on Innovating AI Governance.

This project is affiliated with the CASBS program Humans, Nature, and Machines.

For more information, please contact CASBS program director Zachary Ugolnik (zugolnik@stanford.edu).