Mark Zuckerberg said Meta Platforms had to commit hundreds of billions of dollars to artificial intelligence infrastructure without knowing if the investment would succeed, according to the Cheeky Pint podcast posted Monday.
"We're going to write this check for like, I don't know, $100 billion or $200 billion," Zuckerberg recalled telling chief AI officer, Alexandr Wang, adding he asked his colleague directly whether he was sure it would work.
Wang said it was "not clear at all that it was going to work," and recalled chief financial officer Susan Li repeatedly warning that the cost of failure would be enormous.
The admission lands as Meta's capital spending has climbed sharply: 2025 capital expenditure guidance rose from an initial $60 billion to $65 billion range to $64 billion to $72 billion, and 2026 guidance, including finance-lease payments, now stands at $130 billion to $145 billion.
Zuckerberg's figure describes the scale of the commitment rather than a single disclosed expenditure.
Meta debuted a new AI model Monday that its leadership said matches OpenAI's ChatGPT and Anthropic's Claude on coding and financial or legal reasoning tasks, alongside Muse, a personal assistant for tasks like shopping and scheduling, about which Zuckerberg said expectations were modest.
Investors have grown impatient with the spending; Meta's shares fell 8% after it reported its lowest free cash flow in years last quarter.
The company reports third-quarter results after markets close on 21 October.
