Mark Zuckerberg's Meta to invest hundreds of billions on data centers after poaching top AI talent

Core Insights - Meta plans to invest hundreds of billions of dollars in building massive AI data centers aimed at achieving superintelligence, reflecting a competitive push for top AI talent [1][4] - The company has reorganized its AI initiatives under a new division called Superintelligence Labs, which is expected to create new revenue streams from various AI applications [2] - Meta's CEO Mark Zuckerberg emphasized the strength of the company's core advertising business to support this significant investment, despite investor concerns about potential returns [4][5] Investment and Infrastructure - Meta is on track to develop a 1-gigawatt-plus supercluster for training advanced AI models, with plans for multiple multi-gigawatt data center clusters [5][6] - The first data center, named Prometheus, is expected to be operational by 2026, while another, Hyperion, will have the capacity to scale up to 5 gigawatts [6] - The company has increased its capital expenditure forecast for 2025 to between $64 billion and $72 billion to strengthen its competitive position against rivals like OpenAI and Google [9] Talent Acquisition and Leadership - Zuckerberg has led an aggressive recruitment strategy for the Superintelligence Labs, which will be headed by former Scale AI CEO Alexandr Wang and ex-GitHub chief Nat Friedman [8] - Meta has also made moves to acquire a minority stake in funds managed by Friedman and tech investor Daniel Gross through a tender offer [9]