Ex-Gemini, GPT-4o Researcher Jiahui Yu Leaves Meta to Found AI Startup
Jiahui Yu, a prominent Chinese AI researcher, announced his departure from Meta on August 14 to launch his own startup. The move marks the end of a roughly one-year tenure at Meta’s Superintelligence Lab, where he previously served alongside other high-profile talent recruited directly from OpenAI and Google DeepMind. In a post on X, Yu stated he intends to address a significantly underexplored challenge that he believes will profoundly shape the future of technology, though he withheld details regarding the new company’s name, leadership team, funding, or specific research direction. Yu’s career trajectory has consistently centered on advancing multimodal artificial intelligence. He began at Google DeepMind in late 2022 as a lead for multimodal vision research, contributing to the development of the Gemini foundation model. In October 2023, he transitioned to OpenAI as head of the perception team, where he spearheaded work on visual reasoning and image generation for high-profile releases including GPT-4o, o1, o3, and o4-mini. His expertise in scaling vision-language systems and reasoning architectures positioned him as a key contributor behind OpenAI’s unified multimodal capabilities. Upon joining Meta in June 2025, Yu joined the newly formed TBD Lab under the Superintelligence Lab umbrella. He directed the company’s multimodal initiatives, leading the development of the Muse Spark platform alongside subsequent image, video, and voice modeling tools. Muse Spark, launched in 2026, represents Meta’s push toward native multimodal systems capable of processing text, audio, and visuals while executing complex tool-use tasks. Yu credited the experience as highly formative, citing collaboration with executives including Mark Zuckerberg and Scale AI founder Alexandr Wang. Educated at the University of Science and Technology of China and holding a Ph.D. in computer science from the University of Illinois Urbana-Champaign, Yu’s foundational research focused on computer vision and generative architectures. Early work at institutions such as Adobe and Microsoft Research Asia established his reputation in image restoration and neural network design before he transitioned to large-scale system development. Yu’s departure underscores a continuing trend of top-tier AI talent migrating between the industry’s largest laboratories to pursue independent ventures. While his immediate next steps remain undisclosed, his track record in multimodal reasoning, perception, and generative modeling suggests his forthcoming enterprise will likely target foundational capabilities at the intersection of vision, language, and autonomous agency. The AI research community will be monitoring his transition closely to identify the unresolved technical challenges that prompted his exit from established corporate research environments.
