creators

Lilian Weng - OpenAI researcher, author of Lil'Log

Lilian Weng returned to OpenAI in July 2026 to lead work on recursive self-improvement. She built OpenAI's safety systems team and writes the Lil'Log blog.

Lilian Weng is an AI researcher at OpenAI. She took an undergraduate degree at Peking University and a PhD at Indiana University Bloomington in 2014, supervised by Filippo Menczer in network science and complex systems, then worked at Dropbox and Affirm before joining OpenAI in 2018. At OpenAI she started on the robotics team, which built a robotic hand that solved a Rubik's Cube, helped build the applied AI research team from 2021, and after GPT-4 shipped was asked to create a safety systems group, which had grown past 80 people by the time she left. She became VP of research and safety in August 2024 and her last day was 15 November 2024. In February 2025 she joined Mira Murati's founding team at Thinking Machines Lab. On 27 July 2026 Weng announced she was leaving Thinking Machines, writing that "the amount of consistent stress and workload have pushed me beyond what my health can sustain physically" and that "a scoped role in a more predictable place, and not as a cofounder" would suit her better. Within days OpenAI confirmed she was returning to lead a team on recursive self-improvement; her X profile now reads "MTS @OpenAI; Author of Lil'Log". She has written Lil'Log, a technical blog on deep learning, since 2017.

Specialization

AI safety, large language models, reinforcement learning, technical writing

Country

United States