Mark Zuckerberg announced that Meta has named former OpenAI researcher Shengjia Zhao as the Chief Scientist of its new Meta Superintelligence Labs (MSL). Zhao will guide the scientific vision of MSL under its CEO, Alexandr Wang, and work to accelerate breakthroughs in advanced AI models.
From OpenAI to Meta Superintelligence Labs
Shengjia Zhao helped develop major milestones at OpenAI, including ChatGPT GPT-4 and the early AI reasoning model known as o1. His track record of pioneering efficient scaling techniques made him a prime candidate to cofound and steer Meta’s most ambitious AI research effort. Zhao has been embedded at MSL since its inception, and his formal title reflects Meta’s commitment to cutting‑edge science.
Leadership alongside Alexandr Wang
Alexandr Wang leads MSL as CEO, bringing product and startup expertise. Despite lacking a formal research background, Wang has attracted one of the strongest teams in AI. With Zhao as Chief Scientist, the lab balances technical vision with execution skills. Together, they will define priorities from AI reasoning to multimodal models that learn across vision, language, and code.
Building a world‑class research team
Meta has recruited top researchers from OpenAI, Google DeepMind, Anthropic Safe Superintelligence, and other leading organizations. The team includes experts in AI safety model scaling and generative systems drawn from Meta’s own FAIR lab and its generative AI groups. This talent infusion positions MSL to compete directly with the best labs in the field.
Investment in next-generation infrastructure
Zuckerberg confirmed that by next year, MSL will gain full access to Prometheus, a one‑gigawatt cloud cluster under construction in Ohio. This massive computing power will support the training of frontier AI models that require vast energy and parallelism. The investment signals Meta’s plan to match rivals in raw compute while optimizing its own hardware and software stack.
Defining the frontier of AI research
Chief Scientist Shengjia Zhao will determine the research agenda to form reasoning models, safety mechanisms, and new training paradigms. He will take the initiative of driving forward beyond existing standards and examine the use cases that integrate vision, language, and decision-making. The work of MSL is supposed to support the work that is being carried out by the Meta FAIR lab, which is set on a 10-year research horizon.
Now, Meta Superintelligence Labs, with Zhao in charge of scientific strategy, can potentially compete with any other AI-related institution in leadership and resources. The future will show whether research plan developed by Zhao, along with the implementation as suggested by Wang in the coming months will lead to the next phase of AI innovation.
