Meta Platforms is making an aggressive and strategic push into the realm of artificial intelligence, with CEO Mark Zuckerberg announcing the formation of a new ‘Superintelligence’ team. In an internal memo obtained by WIRED on June 30, 2025, Zuckerberg introduced nearly two dozen researchers, many of whom have been successfully poached from rival AI powerhouses such as OpenAI, Anthropic, and Google.
The new organization will be known as Meta Superintelligence Labs (MSL), encompassing Meta’s existing foundations, product, and FAIR (Fundamental AI Research) teams. A key component of MSL will be a new lab dedicated to developing the next generation of Meta’s AI models. This initiative underscores Meta’s commitment to advancing AI capabilities, aiming to build systems that push the boundaries of current technology.
Leading this ambitious endeavor is Alexandr Wang, CEO of Scale AI, who has been appointed Meta’s ‘chief AI officer’ and head of MSL. Wang’s expertise will be complemented by Nat Friedman, former CEO of GitHub, who will co-lead the new lab alongside Wang, focusing on AI products and applied research. This leadership structure signals a strong emphasis on both foundational research and practical application.
The recruitment drive has been described as a ‘poaching frenzy,’ attracting some of the most sought-after talent in the AI landscape. Notable hires include:
- Trapit Bansal (formerly OpenAI): Pioneered reinforcement learning on chain of thought and co-creator of o-series models.
- Shuchao Bi (formerly OpenAI): Co-creator of GPT-4o voice mode and o4-mini, previously led multimodal post-training.
- Huiwen Chang (formerly Google Research): Co-creator of GPT-4o’s image generation, and inventor of MaskIT and Muse text-to-image architectures.
- Ji Lin (formerly OpenAI): Instrumental in building 03/o4-mini, GPT-4o, GPT-4.1, GPT-4.5, 40-imagegen, and the Operator reasoning stack.
- Joel Pobar (formerly Anthropic): Focused on inference, with a long prior tenure at Meta on various performance and machine learning projects.
- Jack Rae (formerly DeepMind): Pre-training tech lead for Gemini and reasoning for Gemini 2.5, also led early LLM efforts for Gopher and Chinchilla.
- Hongyu Ren (formerly OpenAI): Co-creator of GPT-4o, 4o-mini, o1-mini, o3-mini, 03, and o4-mini, and previously led a post-training group.
- Jiahui Yu (formerly OpenAI/Gemini): Co-creator of 03, 04-mini, GPT-4.1, and GPT-4o, and previously led the perception team at OpenAI and co-led multimodal at Gemini.
- Shengjia Zhao (formerly OpenAI): Co-creator of ChatGPT, GPT-4, all mini models, 4.1, and 03, and previously led synthetic data.
This influx of top-tier talent follows Meta’s significant investment of $14.3 billion in Scale AI, further solidifying its commitment to leading the next wave of AI innovation. The aggressive talent acquisition has not gone unnoticed by competitors, with OpenAI executives reportedly reacting with dismay, stating, ‘Someone has broken into our home’ and indicating they are ‘recalibrating comp’ to retain their own researchers.
Mark Zuckerberg, who co-founded Facebook in 2004 and rebranded it to Meta in 2021 to reflect a broader focus on the metaverse and immersive experiences, continues to drive the company’s strategic direction. As of July 1, 2025, Forbes estimates his net worth at $254.6 billion, making him the world’s third-richest person, providing Meta with substantial resources for its ambitious AI ventures. This concentrated effort to assemble an unparalleled AI team signals Meta’s intent to be a dominant force in the development of superintelligent systems, potentially reshaping the future of technology and human-computer interaction.
