This is episode 20 in CASBS's webcast series Social Science for a World in Crisis. Learn about the series here.

Episode partners: the Behavioral Science Policy Association; the Center for Advancing Safety of Machine Intelligence at Northwestern University; the Stanford Institute for Human-Centered AI; the Psychology of Technology Institute; the Regulation, Evaluation, and Governance Lab (RegLab) at Stanford University; and The Rockefeller Foundation.

Artificial Intelligence (AI) technologies are full of promise but fraught with perils. Reports abound of AI technologies compromising human safety and well-being, unfairly discriminating and amplifying biases, undermining human autonomy, and exacerbating polarization and the spread of misinformation. Yet AI is here to stay and can be a tremendous force for good. How do we better reap the benefits while avoiding the harms?

Social science is no less essential than computer science in developing trustworthy AI technologies. Why? Machine learning, applied to appropriate datasets, enables optimizing metrics of interest. But machine learning alone does not tell us how to construct the datasets, select appropriate target metrics to optimize, or manage ethical tradeoffs. AI development teams need tools to better navigate such issues in partnership with relevant end-users and stakeholders. In short, the real-world goal often is not to optimize algorithms, but rather design systems of humans working with algorithms. Designing such socio-technical systems requires ideas from disciplines such as ethics, psychology, organizational design, and behavioral economics. What does such a field look like? What steps will get us there? Join Kristian Hammond, Daniel Ho, and Jennifer Logg in conversation with Jacob Ward as they outline a new field of AI practice whose scientific foundation extends beyond machine learning and embraces social science.