If you do not have access to the 11th Floor for in-person attendance, please check in at the 10th floor main reception desk to register as a visitor, and someone will escort you to the conference room location.
AI today is often framed as a dichotomy between closed and open models, yet in both paradigms, a single team develops AI top-down, with centralized and monolithic architectures. I envision participatory AI, where modular AI systems are built bottom-up by diverse stakeholders who shape them with their own data, expertise, and priorities. Participatory AI requires rethinking the AI lifecycle: in this talk, I will demonstrate how to train models to advance the systems they join with Compositional RL, align models through peer competition and judgment with Sparta Alignment, collaboratively adapt released models into stronger experts with Model Swarms, and deploy AI systems built from hundreds of researchers’ contributions with Scaling Participation. Together, these works lay the technical foundation for a modular and participatory AI future, one built, improved, and governed by the many.