Research

My research currently focuses on mitigating catastrophic risks from advanced AI, looking particularly at improving AI lab security, implementing better controls, and expanding U.S. government access to AI talent and research. I've previously published studies on air and missile defense, nuclear forces, military space, and defense industry at RAND and CSIS, where I aimed to reduce the risks of war and improve U.S. and allied defense decisions. My research methodology is reading a lot, writing a little, and asking smart people (and now LLMs) to review and improve my work.