Exclusive Insights into Meta's Compact Superintelligence Lab

The superintelligence lab of the technology giant Meta could fit into a Boeing 737 Max and still have unused seats left, as CEO Mark Zuckerberg recently noted with a wink on the podcast "The State of AI with Rowan Cheung." The comparison suggests that the seats in his "boat" are particularly coveted and are occupied with the utmost care. Zuckerberg emphasized that it's not about the number of workers, but about a carefully selected team of 50 to 100 people to achieve the company's ambitious AI goals. Meta's hiring policy aims to form small, talent-dense teams. Employees of this exclusive circle can expect generous compensation packages, unrestricted access to capital for GPU-related costs, and reduced deadlines, as described by Zuckerberg himself. After realizing that Meta's Llama-4 AI model was not taking the desired path, the company initiated an intensive acquisition phase. As part of this venture, Meta invested $15 billion for an almost 50% stake in Scale AI, resulting in their CEO Alexandr Wang moving to lead the superintelligence lab at Meta. To further strengthen the team, Meta secured high-ranking talents from industry giants like OpenAI and Google DeepMind. Zuckerberg emphasized the sensitivity of the project: lack of performance could have serious consequences. Interestingly, once part of the team, the lab's employees are not required to adhere to rigid deadlines. The nature of the research is unpredictable, and the competition among researchers enables them to work at the cutting edge. This trend toward streamlining corporate structures, which has also affected other technology giants such as Google and Microsoft, is a step that Zuckerberg has long advocated. For the superintelligence lab, Zuckerberg also relies on a flat hierarchy to ensure that technical expertise remains consistently at the forefront and is not hindered by outdated management structures.