On January 6, 2026 (Beijing Time), tech website The Verge reported that Jerry Tworek, OpenAI's Vice President of Research, announced his departure on Monday, concluding nearly seven years of service at the company. Tworek was a key figure in building GPT-4, ChatGPT, and OpenAI's first AI programming models, playing a crucial role in pushing the boundaries of new model scales. His 'Reasoning Models' team focused on developing AI systems capable of handling complex logical reasoning and was also a core member of the teams behind the o1 and o3 models, which underpin many of OpenAI's recent AI advancements. Tworek stated in an internal memo and on social platform X that his departure was to 'explore types of research that are difficult to pursue at OpenAI.' As a 'legendary figure' at OpenAI, his exit has sparked widespread speculation about the company's internal research direction and commercialization pressures. Tworek's departure is not an isolated case; previous exits by core members such as Dario Amodei and Ilya Sutskever reflect the clash between idealism and practical interests during OpenAI's commercial transition.