In her public letter, Weng Li clearly stated that the reason for leaving was 'the intensity was too high, my body couldn't take it.' Over the past 7 months, she got sick more times than at any other point in her life, even taking sick leave during the Inkling sprint while battling guilt. She considered switching to a role with narrower responsibilities to reduce anxiety, but believed that 'if I can't give 100% to what I'm responsible for, then I'm not doing the job well,' ultimately choosing to leave.
This decision contrasts sharply with her track record at OpenAI: after joining in 2017, she worked on robotics projects (teaching a five-fingered robotic hand to solve a Rubik's cube), founded the Applied AI Research team, built the Safety Systems team, and became a research VP for safety, witnessing OpenAI's transformation from a non-profit research institute to the ChatGPT era. Additionally, she used her spare time to write the blog Lil'Log, which has over 52,000 Google Scholar citations, and her formula 'Agent = LLM + Memory + Active Planning + Tool Use' has been widely disseminated.
Weng Li's resignation letter did not mention common reasons like strategic disagreements, equity, or KPIs, focusing entirely on personal health. She stated that she still loves AI and enjoys reading papers and following cutting-edge developments, but hopes to go to a more predictable place with clearer boundaries, and no longer serve as a co-founder.