Meta was prepared to write an up to $200 billion AI check before knowing whether its model would work

Mark Zuckerberg was ready to bet as much as $200 billion on AI, and Meta’s finance chief wanted the person building it to understand what failure would mean.

On the Cheeky Pint podcast released on Monday, Meta CEO and Alexandr Wang, Meta’s chief AI officer, recounted conversations from late last year about enormous investments before Meta’s new models had proven themselves. “At the end of last year, we’re like, all right, we are quite excited about the progress.

We haven’t seen the models yet,” Zuckerberg told show host John Collison. “I’m going to, like, you know, we’re going to write this check for, like, I don’t know, like $100 billion, $200 billion.

I’m like, ‘Alex, are you sure this is going to work?’” Zuckerberg said of Wang.

Zuckerberg did not specify the timeframe or exact scope of that contemplated commitment.

Wang said he had encouraging research updates at that point, but success was far from certain.

At an executive gathering in Hawaii, Wang said that CFO Susan Li repeatedly pulled him aside. “Susan pulled me aside many times and was like, ‘You know, I just want to tell you, if this doesn’t work, like, you know, this is going to be really, really bad.

It’s going to be very, very bad in financial terms.’” Wang said of Li.

Meta’s capex over the past few quarters has been substantial, raising eyebrows among investors.

Meta’s latest guidance puts 2026 capital expenditures, including finance lease principal payments, at $130 billion to $145 billion.

In forecasts previously seen by Business Insider, Cantor Fitzgerald analysts expect the spending to reach $200 billion in 2027.

Meta shares fell more than 6% in after-hours trading following April’s earnings report, despite strong results, after the company raised its annual spending forecast to $125 billion to $145 billion.

Inside Meta, the leadership is well aware of the stakes.

In Wang’s words, “this is a fuck ton of risk.” “So I think that in order for it not to work out, it’s not just that we have to fail, AI overall has to fail,” Zuckerberg said. “And then even if that happens, we still have all these other uses to productively use compute.”