Mark Zuckerberg Reshapes Meta’s AI Division
Created on August 19|Last edited on August 19
Comment
Meta has announced another major reorganization of its artificial intelligence unit, signaling Mark Zuckerberg’s growing urgency to compete in the race for advanced AI systems. The company’s AI arm, known as Meta Superintelligence Labs, will now be split into four groups. One will focus on foundational AI research, another on the pursuit of superintelligence, a third on AI products, and a fourth on infrastructure like data centers and hardware. According to people familiar with the matter, this restructuring is intended to streamline Meta’s AI strategy after months of internal friction and strategic pivots.
Downsizing and Talent Moves
As part of the shake-up, Meta is weighing a reduction in the size of its AI division, which has ballooned to thousands of employees. Some executives are expected to leave, and other employees may be reassigned across the company. The realignment comes after Meta spent aggressively to lure top talent, including nine-figure compensation offers and a $14.3 billion investment in Scale AI, whose founder Alexandr Wang was appointed as chief AI officer earlier this year.
Open Source Versus Closed Systems
The restructuring also raises questions about Meta’s AI philosophy. For years, the company positioned itself as a champion of open-source AI, releasing models like Llama for outside developers to build on. But Wang’s new team has discussed closing off Meta’s next major model, a sharp shift from past practice. The decision to abandon Meta’s earlier model, known as Behemoth, underscores the company’s willingness to start fresh in pursuit of more powerful frontier systems.
Leadership Changes and Internal Tensions
High-profile personnel changes reflect ongoing tensions. Joelle Pineau, a leading AI researcher, has departed for Cohere, while Angela Fan, a scientist involved in Llama, left for OpenAI. Loredana Crisan, vice president of generative AI, is also set to depart. At the same time, Shengjia Zhao, formerly of OpenAI and co-creator of ChatGPT, has taken on the role of chief AI scientist at Meta. Zhao has reportedly been reassessing the work of longstanding Meta researchers, fueling unease among veteran staff.\
Strategic Outlook
Zuckerberg has framed Meta’s AI ambitions as central to the company’s future, arguing that superintelligence could unlock a new era of empowerment while bolstering Meta’s core businesses like advertising. The company projects as much as $72 billion in capital expenditures this year, most of it directed toward AI research and infrastructure. While the strategy has stirred internal turbulence, it reflects Zuckerberg’s determination to secure Meta’s place among the few companies vying to dominate the next generation of AI.
Add a comment
Tags: ML News
Iterate on AI agents and models faster. Try Weights & Biases today.