Meta CEO Mark Zuckerberg has appointed former OpenAI researcher Shengjia Zhao as the chief scientist of Meta Superintelligence Labs (MSL), the company’s new AI research unit focused on frontier models and reasoning capabilities. Zhao, known for key contributions to ChatGPT, GPT-4, and OpenAI’s reasoning model o1, has been a foundational figure in MSL since its creation. His appointment formalizes his leadership as the lab accelerates its efforts to develop advanced AI systems.
MSL is led by Alexandr Wang, former Scale AI CEO, who, despite a non-research background, is tasked with managing the unit’s operations while Zhao sets the scientific agenda. Meta has recruited top AI talent from OpenAI, Google DeepMind, Safe Superintelligence, Apple, Anthropic, and internal Meta AI teams, making it one of the most competitive AI research groups globally.
Zuckerberg highlighted Zhao’s pioneering “scaling paradigm” research, which has become essential in AI reasoning models, an area in which Meta currently lacks a direct competitor. The hiring comes after Meta aggressively pursued top researchers, offering high-value compensation packages and personalized recruitment efforts.
To support MSL’s work, Meta is expanding its cloud computing infrastructure, including the upcoming Prometheus cluster in Ohio, capable of one gigawatt of power for large-scale AI training. This massive computing resource, expected to be online by 2026, positions Meta as one of the few tech giants with such capacity.
With Zhao joining Yann LeCun, Meta’s long-standing AI leader and head of the FAIR lab, Meta now operates three AI divisions, each with distinct missions. While FAIR focuses on long-term AI advancements, MSL is charged with immediate progress toward superintelligent AI models. How these teams collaborate is still evolving, but the strengthened leadership signals Meta’s intention to compete directly with OpenAI and Google in AI’s next frontier.
