Lead research projects end to end, designing and executing large-scale experiments to measure automated AI R&D capabilities and identify risks from recursive self-improvement. Build prototypes and production-grade infrastructure, and communicate findings to forecasters, policymakers, and the AI safety community.
Develop and refine models of AI takeoff, identify the assumptions most influential to forecasts, and estimate key parameters using public data and large-scale experiments. Collaborate with engineers and external communities to run experiments on frontier models, propose ways to safely pace automated AI research and development, and publish research findings.
This is an expression of interest for potential non-technical research roles supporting P-Zero Research’s mission to improve the long-term trajectory of AI. Applicants should explain how they could contribute; the organization will retain applications and contact candidates when a suitable need arises.
