Understanding human sociality: From brain to behavior to society


Speaker


Abstract

There is increasing concern that the proliferation of AI-driven automation—particularly in areas dealing with labor markets, education, and criminal justice—may perpetuate and even amplify preexisting biases and social inequities facing certain groups of individuals. However, despite the rich social scientific literature on these topics, we are still far from methods and tools that can recognize, quantify, and correct social biases at the scale necessary to address these societal challenges. Here we connect our work in computational neuroscience of social behavior with machine learning approaches, and develop models that are able to (i) make accurate out-of-sample predictions of social decision-making in both lab and field settings and (ii) scale and generalize to societally relevant challenges.