LOADING...
Summarize
Risking billions on AI is worth the gamble: Zuckerberg
Zuckerberg is betting big on AI

Risking billions on AI is worth the gamble: Zuckerberg

Sep 19, 2025
02:12 pm

What's the story

Meta CEO Mark Zuckerberg has said he is willing to spend big on artificial intelligence (AI) research and development. In a recent episode of the Access podcast, Zuckerberg said that while an AI bubble is "quite possible," the bigger risk for Meta would be not being aggressive enough in its pursuit of this technology. He emphasized that missing out on superintelligence could be more detrimental than overspending by a few hundred billion dollars.

Tech foresight

Superintelligence arrival could be a game-changer

Zuckerberg stressed that if a company moves too slowly and superintelligence comes sooner than expected, it could be out of position. He believes this would be a major setback for what he considers the most important technology enabling new products, innovation, and value creation in history. "The risk, at least for a company like Meta, is probably in not being aggressive enough rather than being somewhat too aggressive," he said.

Financial commitment

$600B investment in US data centers by 2028

Meta has committed to investing at least $600 billion in US data centers and infrastructure by 2028. This was revealed by Zuckerberg earlier this month and later clarified by Meta's CFO, Susan Li. The amount covers all of Meta's US data center buildout as well as investments supporting its US business operations, including new hires.

Strategic shift

Meta slows down recruitment process amid AI race

Despite the massive investment, Meta has also shown signs of caution. The company recently slowed down its recruitment process after offering huge signing bonuses to secure talent in the competitive AI race. This move comes as Wall Street has been closely watching Meta's labor costs and warning that stock-based compensation could dilute shareholder value without clear innovation gains.

Business sustainability

'We're not at risk of going out of business': Zuckerberg

Zuckerberg also compared Meta with other AI labs such as OpenAI and Anthropic, which depend on fundraising to pay their huge compute bills. He said, "We're not at risk of going out of business," unlike these private companies. Their future depends not just on their performance and the future of AI but also on macroeconomic factors, he added.

Strategic focus

Concentrating top talent to prepare for superintelligence

Zuckerberg revealed that Meta is gearing up for superintelligence by concentrating top talent in a small, flat "superintelligence" lab. This setup has no top-down deadlines to reflect the research-heavy nature of frontier AI. He also said that the company is making "compute per researcher" a competitive advantage by outspending rivals on GPUs and custom infrastructure needed to power them.