Equality is not equity: Recognizing group differences in AI fairness

Responsible AI often takes an equality-based approach that is centered around ideas of sameness and impartiality. And yet, while mathematically convenient, this approach does not take into account the relevant social context and historical injustices that we might care about. Instead, we should pivot to an equity-based approach that is able to recognize legitimate social group differences. In this talk, I will touch on two lines of work in this vein. The first considers how we can evaluate fairness in a way that does not treat each social group as interchangeable. The second considers how in our rush to replace human participants with LLMs we have forgotten the importance of human participant positionality (i.e., the relevance of social identities like gender and race).

Join us online: Link will be shared the day before the talk

Angelina Wang is a postdoc at Stanford University, and incoming assistant professor of information science at Cornell Tech. Her research is in the area of machine learning fairness and algorithmic bias. She has been recognized by the NSF GRFP, EECS Rising Stars, Siebel Scholarship, and Microsoft AI & Society Fellowship. She publishes in top machine learning (ICML, AAAI), computer vision (ICCV, IJCV), interdisciplinary (Big Data & Society), and responsible computing (FAccT, AIES) venues, including spotlight and oral presentations. She earned her PhD in Computer Science from Princeton University and BS in Electrical Engineering and Computer Science from UC Berkeley.

Host or Publisher Cambridge Social Dynamics Team