Nonlinear is a nonprofit AI safety research organization dedicated to ensuring that transformative AI systems remain safe, beneficial, and aligned with human values. We provide support for important, neglected, and tractable advocacy-related projects by experienced professionals.