Jiahui Yu Becomes Key Meta Prize in AI Talent War
By James Moreau | 09 Jul, 2025

AI engineer Jiahui Yu shaped generative image software and more complex thought processes in LLM technology for Google and OpenAI until his recent defection to Meta.

AI engineer Jiahui Yu made headlines with his move in late June from OpenAI to Meta.  His hiring by Zuckerberg was a loud shot fired in the ongoing machine learning talent war.  Yu is among seven key OpenAI members recently recruited by Meta which is reportedly budgeting up to $300 million over four years in compensation packages to recruit top AI talent.  Most of the compensation is in the form of stock options whose value is keyed to length of employment and company share performance.

Since joining Open AI in 2023 Yu contributed to the evolution of its core AI models, including GPT-4, GPT-4.1, and GPT-4o. He led the Perception team which develops the “senses” of a Large Language Model allowing it to collect, interpret, and understand data, including images, audio, and sensor readings. Yu also contributed to the “o1” reasoning model to enable more “thinking time” to deliver more robust and strategic responses.

Prior to OpenAI, Yu co-led Gemini’s multimodal development at Google DeepMind to improve cohesion and understanding between senses. His work at Google from 2020 to 2023 enabled Gemini to natively process images, audio, and video in addition to text.

In 2017, while interning at Adobe, Yu developed project “Deep Fill” which creates correct facial features when masked, remove unwanted elements like people from scenic photos, and create unique features within masked regions. This project was a direct contributor to Photoshop’s Generative Fill which launched in 2023. A Chinese native Yu graduated Hefei’s School of the Gifted Young at the University of Science and Technology of China before earning a Ph.D. in computer science from the University of Illinois in 2019. Yu’s published works on Google Scholar have been cited over 34,000 times.