Recent work explores a convergence of world generation and embodied intelligence through Genie 3, a world model for richly interactive environments that was recognized in TIME’s Best Inventions of 2025, and SIMA 2, a Gemini-powered agent that can reason and learn in virtual 3D worlds.
Previously, I was a postdoctoral research and teaching fellow at the University of British Columbia and the Vector Institute, supervised by Prof. Jeff Clune. During this time, we explored how modern foundation models can drive open-ended learning and autonomous scientific discovery through projects such as Intelligent Go-Explore, Automated Design of Agentic Systems, Automated Capability Discovery, and The AI Scientist. The AI Scientist was featured in the State of AI Report for two consecutive years, and The AI Scientist-v2 later generated the first fully AI-written paper to pass peer review at an ICLR workshop.
I received my PhD from the University of Oxford under the supervision of Prof. Michael A. Osborne and Prof. Yee Whye Teh, where I worked on offline reinforcement learning, including generalization to unseen tasks, uncertainty quantification, learning from pixels, and diffusion synthetic data. My PhD thesis gives a fuller account of that work.
