My research aims to make artificial intelligence systems usable to human beings, so that human-AI interactions are more productive, enjoyable, and fair. I believe AI systems should be designed to augment human agency, and thus approach this process by considering the capabilities and limits of human intelligence. Before joining Google, I did my PhD research in the User Interface Design group at MIT, where I built "wait-learning" tools to help people practice desired skills in short chunks while waiting, thereby making use of fleeting moments in the day.