Mark Zuckerberg's Meta Superintelligence Lab has brought together a team of top AI experts from rival firms like OpenAI, Google DeepMind, and Anthropic. The team is led by Alexandr Wang, former CEO of Scale AI, who has been appointed as Meta's Chief AI Officer. Wang will be joined by Nat Friedman, ex-GitHub CEO, who will co-lead the lab and focus on AI products and applied research.
The team includes several notable members from OpenAI, such as Trapit Bansal, who pioneered reinforcement learning on chain of thought and co-created o-series models. Other OpenAI alumni include Shuchao Bi, Hongyu Ren, Jiahui Yu, and Ji Lin, who have all made significant contributions to various AI projects.
The team also features experts from Google DeepMind, including Jack Rae, who led pre-training tech for Gemini and reasoning for Gemini 2.5. Pei Sun, another DeepMind alum, has worked on post-training, coding, and reasoning for Gemini. Huiwen Chang, also from DeepMind, co-created GPT-4o's image generation and invented MaskIT and Muse text-to-image architectures.
Additionally, the team includes Joel Pobar from Anthropic, an inference expert with 11 years of experience at Meta. Other notable members include Johan Schalkwyk, former Google Fellow, and Shengjia Zhao, co-creator of ChatGPT and GPT-4.
Meta's aggressive recruitment drive, with reported signing bonuses of up to $100 million, demonstrates its commitment to developing superintelligence. The company's goal is to deliver personal superintelligence for everyone, leveraging its vast resources, experience in AI hardware, and company structure that allows for big swings. With this talented team in place, Meta is well-positioned to make significant advancements in the field of AI.