Tencent has brought on a second former researcher from the U.S. AI developer known for ChatGPT to join its large language model division, according to a company insider.
This new team member, Tian Yonglong, will focus on developing visual-language models, the source revealed yesterday. Last December, Tencent, headquartered in Shenzhen, appointed Vinces Yao, a former researcher from OpenAI, as the chief AI scientist and leader of the AI Infrastructure and Large Language Models departments.
Tian holds a Ph.D. in Electrical Engineering and Computer Science from MIT, with primary research interests in computer vision, visual representation learning, and generative models. Before Tencent, he worked at DeepMind, an AI company backed by Google, and afterward at OpenAI, where he collaborated with Yao.
Yao has played a significant role in streamlining processes at Tencent, according to senior executive Dowson Tong in an interview last month. The company’s large language model, Hunyuan, has also seen substantial advances.
Having led AI agent development at OpenAI, Yao reports directly to Tencent’s president, Martin Lau. He has overseen the development of Hunyuan, releasing the Hy3 preview in April and the full Hy3 version earlier this month, while also working on integrating the language model with consumer products like Yuanbao.