# Lilian Weng

> Lilian Weng is an AI researcher and writer who helped build OpenAI’s robotics, applied research and safety systems, co-founded Thinking Machines Lab, and returned to OpenAI in 2026 to lead work supporting research acceleration. Her technical blog Lil’Log has documented her learning since 2017.

- **Role:** Research leader at OpenAI
- **Organizations:** OpenAI, Thinking Machines Lab, Fellows Fund, Affirm, Peking University, Indiana University Bloomington
- **Education:** Bachelor’s degree, Peking University, PhD, Indiana University Bloomington
- **Known for:** Contributed to OpenAI’s robot-hand Rubik’s Cube project., Founded and led OpenAI’s Applied AI Research team., Led the Safety Systems group, which she said had more than 80 members in 2024.

## Career timeline

- **2014** — Completed doctoral research at Indiana University Bloomington on information diffusion and social networks.
- **2017** — Began publishing Lil’Log learning notes.
- **2018** — Joined OpenAI and worked on robotics research, including the Rubik’s Cube robot hand.
- **2021** — Began leading OpenAI’s Applied AI Research team.
- **2023** — Became head of OpenAI’s Safety Systems team and published her essay on LLM-powered autonomous agents.
- **2024** — Served as OpenAI vice president of research, safety, then left the company in November.
- **2025** — Co-founded Thinking Machines Lab with Mira Murati and others.
- **2026** — Left Thinking Machines Lab in July and returned to OpenAI to lead a team supporting research on recursive self-improvement.

## Achievements

- Contributed to OpenAI’s robot-hand Rubik’s Cube project.
- Founded and led OpenAI’s Applied AI Research team.
- Led the Safety Systems group, which she said had more than 80 members in 2024.
- Listed as applied research lead in OpenAI’s GPT-4 contributions.
- Created Lil’Log, a long-running technical blog explaining AI research.
- Co-founded Thinking Machines Lab.

## Latest updates

- **2026-07** — Returned to OpenAI to lead a team supporting internal research acceleration and recursive self-improvement work.
- **2026-07** — Published “Harness Engineering for Self-Improvement” on Lil’Log.
- **2026-06** — Published “Scaling Laws, Carefully” on Lil’Log.

## Links

- Website: https://lilianweng.github.io/
- LinkedIn: https://www.linkedin.com/in/lilianweng
- Twitter/X: https://x.com/lilianweng
- GitHub: https://github.com/lilianweng

---

Profile page: https://yespress.io/lilian-weng
Published by YesPress — https://yespress.io
Last updated: 2026-09-29