Climate change is one of the most pressing issues of our time, requiring rapid transformation across virtually every sector of society. In this talk, I describe what this means for research and practice in AI. AI has a multi-faceted relationship with climate change, through a combination of its direct environmental footprint, the impacts of its applications (both good and bad), and the broader systemic shifts it induces. Ultimately, most work in AI has significant implications for climate action, whether or not it is viewed as traditionally “climate-relevant.” Given this, I discuss how the AI community can better align its work with climate action: through the kinds of methods we develop, the kinds of applications we work on, the choices we make while working on these applications, and the ways we communicate with the public about our work.