Meta CEO Mark Zuckerberg has declared that massive, aggressive spending on artificial intelligence is essential, stating it’s better to “overbuild now than regret later.” He is willing to risk hundreds of billions of dollars to ensure Meta leads the AI race.
Key Takeaways
- Massive Investment: Meta plans to invest at least $600 billion in US data centres and infrastructure by 2028.
- Aggressive Stance: Zuckerberg believes the risk of being too cautious far outweighs the risk of overspending.
- Financial Advantage: Unlike startups, Meta’s financial scale protects it from funding droughts.
- Superintelligence Goal: A new, dedicated lab is focusing on achieving artificial superintelligence.
Embracing the AI Bubble Risk
In a recent podcast interview, Zuckerberg acknowledged the current AI boom could be a bubble. However, he argued that the greater danger for a company of Meta’s size is not being aggressive enough. “If we end up misspending a couple of hundred billion dollars, I think that is going to be very unfortunate,” he said. “But the risk is higher on the other side.”
Learning from Past Tech Booms
Zuckerberg compared the situation to previous technological cycles where overbuilding led to a bust. He noted that even after those downturns, the infrastructure that was built often became the foundation for the next wave of innovation. For Meta, the primary fear is missing the arrival of the next big leap: .
A $600 Billion Bet on Infrastructure
This philosophy is already translating into action. The company recently revealed a plan to invest a staggering $600 billion in data centres and broader US business operations by 2028. Meta’s CFO, Susan Li, clarified this encompasses all domestic data-centre projects and investments, including new hires, signalling an unprecedented scale of commitment.
Meta’s Financial Edge Over AI Startups
Zuckerberg also highlighted a key advantage Meta holds over rivals like OpenAI and Anthropic. As a trillion-dollar company, Meta does not face the same financial constraints. “We’re not at risk of going out of business,” he stated, warning that private AI labs could struggle to afford their infrastructure costs if investor funding dries up.
The Superintelligence Lab and Compute Power
To achieve its ambitious goals, Meta has established a dedicated “superintelligence lab” that brings its top researchers together. This team operates with minimal hierarchy and no rigid deadlines. Furthermore, the company is focusing on dominating a new metric: “compute per researcher.” By pouring billions into GPUs and custom AI infrastructure, Meta aims to give its researchers unparalleled computing power to accelerate breakthroughs.
