
Designing AI for Societal Good
Qian Yang, an assistant professor at Cornell University and a human-computer interaction researcher, presents this Stanford CS547 HCI Seminar talk on how AI systems might be designed to serve the public good rather than narrow commercial aims. She uses generative AI chatbots and their effects on public mental health as a worked case study, examining how design choices in these systems shape user wellbeing at scale. From there she proposes a framework she calls designing AI as planting seeds, an approach aimed at nudging better business models, policy, and political outcomes rather than engineering direct solutions. Yang walks through early research exploring this design practice as an emerging research agenda, discussing what it would mean for HCI researchers and designers to think past individual products toward the institutions and incentives those products feed into. The talk runs just under an hour and sits within Stanford's ongoing HCI seminar series, aimed at a graduate research audience interested in responsible AI design.