What should the classroom of the future look like, and what is the role of AI in these classrooms? Over the last 30 years, the research on how people learn increasingly acknowledges that learning is collaborative, dynamic, and supported by tasks that connect to the needs and interests of the learner. Yet, the prevailing approach to using artificial intelligence (AI) in education has focused on supporting individual students in narrowly defined tasks, rather than open-ended collaborative knowledge building. In this talk, I will describe the work of the Institute of Student-AI Teaming (iSAT) and how we are working to build learning environments and AI technologies that support equitable and productive collaborations. I will also discuss how these technologies might support teachers' orchestration of classroom activities at the small and whole group levels in secondary classrooms.