Meta Centralizes AI: Superintelligence Labs is Born | Contact Meta WhatsApp | Facebook | Meta WhatsApp | Turtles AI

Meta has established “Superintelligence Labs,” led by AlexandrWang and NatFriedman, to accelerate the development of advanced AI. Massive initiative: billion-dollar investments, top recruiting, and enhanced infrastructure to compete in the AGI race.

Key points:

• Creation of internal Meta Superintelligence Labs division, with the merger of FAIR, GenAI, and AGI research teams.
• Led by AlexandrWang (ex-ScaleAI) and NatFriedman (ex-GitHub), with separate roles on research and product.
• Massive investments: $14.3B in ScaleAI, datacenter expansion, compute upgrades to support compute-intensive workloads.
• Recruiting push: 11 new researchers from OpenAI, DeepMind, Anthropic; million-dollar offers, bonuses up to $100M.

Meta has decided to consolidate all of its AI activities under one strategic umbrella, integrating its basic research components (FAIR), the Llama and GenAI teams, and new groups dedicated to the development of advanced AI models; this operation has led to the creation of the laboratory called Meta Superintelligence Labs, which will be led by Alexandr Wang as Chief AI Officer and co-led by Nat Friedman for applications and products. This reorganization follows the lukewarm response to the Llama 4 model, and the rapid advancement of competitors – Google, OpenAI and DeepSeek – which has pushed Zuckerberg to focus on long-term goals such as the creation of general AI (AGI) and the enhancement of Meta AI applications, advanced advertising tools and smart glasses. Meta has dedicated significant resources to the acquisition of Scale AI – 49% for 14.3 billion USD, valuing the company at 29 billion USD – precisely to ensure Wang’s leadership in the new project; in addition to the financial investment, the construction of data centers for over 1GW of power is added, with capital expenditure expected between 64–72 billion USD by 2025. In parallel, Zuckerberg has orchestrated an intense direct recruitment campaign, also via WhatsApp, hiring eleven researchers from OpenAI (including Lucas Beyer, Alexander Kolesnikov, Xiaohua Zhai, Jiahui Yu, Shuchao Bi, Shengjia Zhao, HongyuRen), DeepMind (JackRae, PeiSun) and Anthropic (Joel Pobar), offering packages that – according to some sources – would have included bonuses of up to 100 million USD. This momentum has generated enthusiasm among investors, contributing to the achievement of a record value for Meta shares – 738.09 USD, with a consequent capitalization close to 1.86 billion USD. However, some observers note that the path to AGI remains uncertain: the massive investment is reminiscent of previous hardware spending (e.g. Reality Labs, with over 60 billion USD in losses), and figures such as Yann LeCun warn that current technologies are unlikely to lead to superintelligence, while Masayoshi Son estimates that the breakthrough could come in a decade.

The birth of Superintelligence Labs signals Meta’s willingness to consolidate its AI capabilities into a single strategic path, supported by investment, talent and infrastructure, without anticipating the outcome of this path.