Key takeaways
- Liam Fedus resigns from OpenAI to focus on AI applications in scientific research.
- OpenAI will invest in and partner with Fedus's new venture in AI for science.
- Fedus previously led post-training efforts at OpenAI, enhancing AI model performance.
- His academic background in physics fuels his passion for applying AI in scientific contexts.
- Recent leadership changes at OpenAI highlight ongoing shifts within the organization.
Liam Fedus, vice president of research, post-training at OpenAI announced his resignation on Tuesday.
Fedus, who led OpenAI's post-training team, which is responsible for fine-tuning AI models for public release, shared his decision in a message to colleagues, stating he intends to remain involved with OpenAI as a partner while pursuing new opportunities in AI for science.
He expressed excitement about applying AI technology to scientific research, citing his academic background in physics. “I’ve gotten really excited about AI for science. My undergrad was in physics and I’m keen to apply this technology there,” he said in his post on X.
He also mentioned that OpenAI plans to invest in and partner with his new venture.
“Because AI for science is one of the most strategically important areas to OpenAI and achieving ASI, OpenAI is planning to invest in and partner with my new company,” he said.
Fedus joined OpenAI in 2022 after a career spanning roles at Google Brain and academic research in physics and machine learning. He played a pivotal role in developing ChatGPT and advanced reasoning models like Strawberry.
Fedus thanked OpenAI leadership and colleagues for their support during his tenure, stating, “Contributing to the mission of OpenAI and working with world-class teams to create and improve ChatGPT has been an experience of a lifetime.”
“Thanks to all the leadership who believed in me early on, especially, Sam, Greg, and Mark. Thank you everyone on post-training and to all of our collaborators across research and product.”
OpenAI has experienced the departure of several notable leaders recently. Among the prominent departures, Ilya Sutskever, the chief scientist and co-founder, resigned in May 2024 after being involved in the failed coup against Altman. He launched his own AI startup called Safe Superintelligence (SSI).
Mira Murati, the chief technology officer, left OpenAI in September 2024 after serving over six years at the company and founded her own startup called Thinking Machines Lab last month.
Bob McGrew, the chief research officer, and Barret Zoph, vice president of research, also announced their departures shortly after Murati. Alec Radford, a lead author of the original GPT paper, left OpenAI to pursue independent research.
