In a bold move to regain dominance in the artificial intelligence race, Meta has enlisted Scale AI founder and CEO Alexandr Wang to head its newly launched “superintelligence” research lab—a significant strategic shift aimed at catching up with rivals like OpenAI, Anthropic, and Google.
The initiative comes amid growing pressure on Meta to deliver breakthroughs in AI after recent setbacks, including the underwhelming launch of its Llama 4 model. According to reports by The New York Times and Bloomberg, Meta is not only bringing Wang aboard but is also in talks to invest billions in Scale AI, the startup Wang co-founded in 2016.
Meta’s Superintelligence Lab: A New Frontier
Meta’s new lab will focus on “superintelligence”—a term referring to AI systems that exceed human intelligence across a wide range of tasks. While the term has often been considered speculative, Meta CEO Mark Zuckerberg reportedly views it as essential for staying competitive.
In recent months, Zuckerberg has taken a personal role in AI hiring efforts, holding meetings at his homes in Lake Tahoe and Palo Alto with elite engineers and researchers. Sources say he’s personally assembled a team of 50 AI experts for the new lab, with Wang now positioned as a central figure in that effort.
Who Is Alexandr Wang?
At just 28, Alexandr Wang has emerged as one of the most influential figures in the AI world. Through Scale AI, he built the infrastructure that provides training data—often described as the “fuel” of AI models—for tech giants including Meta, OpenAI, and the U.S. Department of Defense.
Scale’s biggest strength lies in its ability to label massive datasets with high accuracy—critical for improving machine learning performance. Wang’s startup was also a key partner in Defense Llama, a military-grade large language model built on Meta’s Llama 3 architecture.
Why Meta Needs Scale AI (and Wang)
Meta has experienced significant AI turbulence in recent years, including multiple reorgs, top talent departures, and mixed success with its AI models. Despite early promise with the open-source Llama series in 2023–2024, Llama 4’s release in April 2025 was met with criticism for rushed development and weak performance against Chinese rival DeepSeek.
To address these challenges, Meta is taking a page from competitors like Microsoft and Amazon, both of which have used strategic investments to attract top-tier AI talent and acquire access to proprietary tech—without triggering antitrust alarms.
The Stakes: Meta vs. OpenAI, Anthropic, and Google
Meta is reportedly offering seven- to nine-figure compensation packages to lure elite researchers, but even that hasn’t guaranteed wins. Some top candidates have still chosen OpenAI or Anthropic, highlighting the fierce competition in the AI talent war.
Meanwhile, Meta’s long-standing AI research arm, FAIR (Fundamental AI Research), led by Meta chief scientist Yann LeCun, has seen high-profile exits and a shift in focus toward Advanced Machine Intelligence (AMI). LeCun has publicly distanced himself from the hype around AGI, casting doubt on the current LLM-driven path toward human-level intelligence.
This raises questions about whether LeCun’s role will be marginalized as Meta shifts its bets toward Wang’s vision of scalable, data-driven AI progress.
A High-Risk, High-Reward Strategy
Meta’s collaboration with Scale AI and Wang could reshape its AI trajectory—if it pays off. For Wang, the opportunity cements his role as a rising power player in global AI. For Zuckerberg, it’s a high-stakes bet that his company can still leapfrog back into leadership in the race for next-generation AI.
As AI development grows more expensive and complex, Zuckerberg is betting that a tight integration of data, talent, and compute innovation will be Meta’s best shot at achieving true AI breakthroughs.