Meta Platforms Inc. Chief Executive Mark Zuckerberg disclosed that the company committed to spending hundreds of billions of dollars last year on the infrastructure to support its artificial intelligence (AI) initiatives before knowing whether the investments would yield success. Speaking on the “Cheeky Pint” podcast, Zuckerberg recounted a moment of uncertainty when he asked Meta’s Chief AI Officer, Alexandr Wang, if the substantial expenditure—potentially reaching $100 billion or $200 billion—would pay off.
The admission highlights ongoing investor concern over Meta’s rapidly increasing expenses tied to ambitious AI projects. Despite Zuckerberg’s assurances that these investments will eventually drive revenue and profit growth, some shareholders remain cautious about the company’s ability to generate returns quickly enough.
During the podcast conversation with host John Collison, both Zuckerberg and Wang noted that Meta’s Chief Financial Officer, Susan Li, had repeatedly pointed out the high stakes and the potential cost of failure. “I think Susan scared you a little bit,” Zuckerberg said, while Wang acknowledged the uncertainty surrounding the outcome, describing the investment as “obscenely large” without guarantees of success. Wang declined to make firm predictions, saying he could only offer a “vibe report” on progress.
On Monday, Meta unveiled a new AI model that it described as a significant advancement over its previous versions. The company asserted the model possesses “frontier capabilities,” positioning it to compete with leading generative AI systems such as OpenAI’s ChatGPT and Anthropic PBC’s Claude, particularly in tasks like coding and complex financial or legal queries. This milestone represents a critical test of whether Meta’s substantial financial commitment to rebuilding its AI team has translated into competitive technological progress.
In addition to the new AI model, Meta introduced Muse, a personalized AI assistant aimed at automating consumer tasks including shopping, purchasing movie tickets, and scheduling appointments. Zuckerberg acknowledged uncertainty regarding consumer reception, noting the company’s expectation that Muse would require iterative improvements.
Meta’s sizable AI investments are part of a broader effort to keep pace with U.S. technology rivals like Alphabet Inc., OpenAI, and Anthropic in the fast-evolving AI landscape. The company ranks among the industry’s largest spenders on AI data center infrastructure.
Investors will have another opportunity to assess the impact of Meta’s AI-driven strategy when the company reports its financial results after market close on October 21. In the previous quarter, Meta posted its lowest free cash flow in years, reflecting the mounting costs of its AI data centers and other ventures such as smart glasses, which precipitated an 8% drop in the company’s shares.