Barret is currently the Vice President of Research (Post-Training) at OpenAI. Their team focuses on alignment, tool use, evaluations, ChatGPT, Search, multi-modality, and more.
Previously, he was a Staff Research Scientist at Google Brain, where he worked on training large sparse language models and applying them to various applications. Barret also served as the research team lead for MuM and made significant contributions to AutoML research and product development. His work on Neural Architecture Search (NAS) served as a key impetus in the field.
Before joining Google Brain, Barret worked at the Information Sciences Institute with Kevin Knight and Daniel Marcu on statistical machine translation.