Meta's Mark Zuckerberg Discusses Balancing AI Infrastructure Sales with Internal Development
Mark Zuckerberg outlined the strategic dilemma facing Meta as it builds out its artificial intelligence infrastructure. During an earnings call, he emphasized the challenge of deciding whether to monetize excess computing capacity by leasing it or retaining it for future model development and services. While other major hyperscalers like Amazon and Microsoft have cloud businesses, Meta does not yet offer such a service despite significant capital expenditures. Zuckerberg noted that companies are offering premium prices for Meta's compute resources, with potential partners like Anthropic already in preliminary talks to lease capacity. However, the CEO also highlighted the importance of reserving enough computing power to support internal AI initiatives led by Chief AI Officer Alexandr Wang. These include new models such as Muse Spark 1.1 and broader enterprise services. Meta's financials reflect its heavy investment strategy: free cash flow dropped sharply in Q2 due to rising capital expenditures, which are expected to reach up to $145 billion for the year.
Tickers: $META
Topics: AI infrastructure capex impact on cash flow cloud computing strategy Meta