Meta to boost AI spending as Zuckerberg pushes for 'Superintelligence'
What's the story
Meta CEO Mark Zuckerberg has announced plans to invest heavily in artificial intelligence (AI), as part of a bid to make the company a leading player in the field. The move comes amid an intensifying AI arms race with competitors like OpenAI, Google, and Microsoft. During its third-quarter earnings call on October 29, Meta raised its full-year capital expenditure forecast for 2025 to between $70 billion and $72 billion, up from its previous range of $66 billion to $72 billion.
Financial outlook
Projected expenses for 2025
Along with the capital expenditure forecast, Meta also projected total expenses for 2025 to be between $116 billion and $118 billion, up from $114 billion to $118 billion. The company expects a year-over-year (YoY) growth of 22% to 24%. This increase is partly due to higher employee compensation amid an aggressive AI hiring spree. The company has been luring top researchers from competitors like OpenAI, Google, Anthropic, and Apple with billion-dollar pay packages.
Market reaction
Investor concerns over AI spending
The aggressive spending on AI has raised fresh concerns among investors over the potential returns on the billions being invested in AI infrastructure and talent acquisition. This has led to a dip in Meta's stock, which fell as much as 9% on October 29. Despite these concerns, Zuckerberg defended the increased spending during the earnings call, saying it is necessary to build "personal superintelligence for everyone" through Meta Superintelligence Labs.
Strategic approach
Zuckerberg on potential superintelligence impact
Zuckerberg emphasized the importance of aggressively building capacity to prepare for the most optimistic scenarios. He said, "If superintelligence comes sooner, we will be in a great position for a generational paradigm shift and many large opportunities." He also mentioned that if development takes longer, Meta could use the additional compute to speed up its core business, which can profitably use more compute than it has been able to throw at it.
Strategic shifts
Meta's revamped AI strategy
In recent months, Meta has revamped its AI strategy after a series of setbacks. These include the underwhelming performance of its flagship AI model Llama 4 and the departure of several top researchers. The company invested $14.3 billion in data-labeling start-up Scale AI for a 49% stake in June and appointed its CEO, Alexandr Wang, as chief AI officer.
Organizational changes
Restructuring of AI division and new lab launch
Meta has also restructured its AI division and launched Meta Superintelligence Labs. This new lab houses the company's foundation models, product, and Fundamental AI Research (FAIR) teams. It also has a new lab to build the next generation of its AI models. Earlier this month, the company hired Andrew Tulloch, co-founder of Thinking Machines Lab.