OpenAI reasoning chief Jerry Tworek quits after 7 years: Here’s why

Jerry Tworek, the prominent research lead at OpenAI, has officially resigned after an impactful seven-year tenure. Announcing his departure through a post on X (formerly Twitter), Tworek expressed that his future research ambitions no longer aligned with OpenAI's current focus, making the decision a particularly tough one for him. He reflected on the groundbreaking achievements of the company in the AI landscape, emphasizing that OpenAI's influence will persist and grow over time. Joining OpenAI in 2019, Tworek played a crucial role in the development and scaling of GPT models, laying the groundwork for sophisticated Large Language Models (LLMs). In 2022, he ascended to the position of Vice President of Research, overseeing key products such as ChatGPT and the latest GPT-4 iterations. Among his notable contributions were the OpenAI o1 and o3 models, which demonstrated Ph.D.-level proficiency in various scientific disciplines, including physics, chemistry, and biology. Additionally, Tworek led the development of Codex, the technology that underpins GitHub Copilot, and was integral to the post-training processes for the GPT-4 model. His departure marks a significant moment for OpenAI, as Tworek's expertise has been pivotal in the company’s advancements in AI systems and research. In his statement, Tworek mentioned, "I am leaving to try and explore types of research that are hard to do at OpenAI." His exit raises concerns for OpenAI, particularly in light of recent trends where several researchers have transitioned to competing firms. As the landscape grows more competitive, the company faces the challenge of maintaining its research momentum without key talent like Tworek.