Ryan Othniel Kearns is a DPhil student at the OII, advised by Professor Adam Mahdi in the Reasoning with Machines Lab. Ryan is also an affiliate student with the Laboratory for Human-Centered AI, part of the Institute for Ethics in AI in Oxford’s Faculty of Philosophy. Ryan researches metrology—the scientific study of measurement—as it concerns frontier AI systems. AI evaluations, or benchmarks, are our best resource for characterising AI system capabilities. Ryan’s research develops statistical and methodological techniques to make AI evaluations trustworthy and reliable accounts of real system behaviours.
Ryan’s research is supported by a Clarendon Scholarship. His work has appeared at top AI publication venues including ACL, EMNLP, ICML, and NeurIPS.
Previously, Ryan studied Computer Science (BS) and Philosophy (BAH) at Stanford, followed by Social Data Science (MSc) at the OII. He was also a Founding Data Scientist at Monte Carlo Data, has presented at the Snowflake Summit and the Oxford Generative AI Summit, and wrote three chapters of the O’Reilly’s Data Quality Fundamentals textbook.
Natural Language Processing, Computational Linguistics, Multi-Agent Systems, Trustworthy AI, Explainable AI, Social Epistemology, Social Cognition, Computational Social Choice, Network Science, Network Dynamics.
