Meta Names Shengjia Zhao Chief Scientist of AI Superintelligence Division
On Friday, Meta’s CEO Mark Zuckerberg announced that Shengjia Zhao, a former OpenAI researcher, will lead research initiatives at the company’s new AI division, Meta Superintelligence Labs (MSL). Zhao was crucial in several key advancements at OpenAI, including ChatGPT, GPT-4, and the initial AI reasoning model, o1.
“I’m excited to share that Shengjia Zhao will be taking on the role of Chief Scientist for Meta Superintelligence Labs,” Zuckerberg noted in a Threads post on Friday. “Shengjia co-founded this lab and has been our lead scientist from the start. As our recruitment efforts progress and our team takes shape, we thought it was the perfect moment to formalize his leadership.”
Zhao will establish the research strategy for MSL under the guidance of Alexandr Wang, the former CEO of Scale AI, who has recently been appointed to lead the lab.
Despite Wang’s unconventional background for leading an AI lab, Zhao’s recruitment, acknowledged for his leadership in advanced AI model development, strengthens the leadership team. To further enhance the unit, Meta has hired multiple senior researchers from OpenAI, Google DeepMind, Safe Superintelligence, Apple, and Anthropic, alongside talent from Meta’s existing Fundamental AI Research (FAIR) lab and generative AI team.
In his announcement, Zuckerberg underscored Zhao’s contributions to vital advancements, including a “new scaling paradigm.” This likely pertains to Zhao’s work on OpenAI’s reasoning model, o1, where he is recognized as a key contributor alongside OpenAI co-founder Ilya Sutskever. At present, Meta does not have a rival to o1, positioning AI reasoning models as a major focus for MSL.
The Information reported in June that Zhao would join Meta Superintelligence Labs with three other distinguished OpenAI researchers — Jiahui Yu, Shuchao Bi, and Hongyu Ren. Additionally, Meta has engaged Trapit Bansal, another OpenAI researcher who collaborated with Zhao on AI reasoning models, as well as three researchers from OpenAI’s Zurich office who worked on multimodality.
Zuckerberg is dedicated to ensuring MSL’s success and has initiated an aggressive recruitment campaign, which includes sending personalized emails to researchers and inviting candidates to his Lake Tahoe residence. Reports indicate that Meta has offered some researchers lucrative compensation packages, reaching eight or nine figures, with many offers having quick expiry dates.
Techcrunch event
San Francisco
|
October 27-29, 2025
Meta is also enhancing its investment in cloud computing infrastructure, which will support MSL in executing the extensive training runs needed to develop competitive cutting-edge AI models.
By 2026, Zhao and MSL’s researchers will have access to Meta’s 1 gigawatt cloud computing cluster, Prometheus, located in Ohio. Upon becoming operational, Meta will be among the first tech companies to have an AI training cluster of Prometheus’ scale — capable of supplying energy for over 750,000 homes. This infrastructure will help Meta conduct the rigorous training necessary for creating advanced AI models.
With Zhao’s addition, Meta now has two chief AI scientists, including Yann LeCun, head of Meta’s FAIR lab. In contrast to MSL, FAIR is geared toward long-term AI research, exploring techniques likely to emerge in five to ten years. The collaboration among Meta’s three AI units remains to be seen.
Nevertheless, Meta now appears to boast a robust AI leadership team ready to compete with OpenAI and Google.
