Currently developing a novel AI Alignment agenda that focuses on scalably inducing self-other overlap in state-of-the-art machine learning models in order to facilitate the learning of empathetic and non-deceptive representations and elicit prosocial behaviors.