Microsoft’s AI chief Mustafa Suleyman has warned that staying competitive in the race to develop frontier artificial intelligence will require investments of “hundreds of billions of dollars” over the next decade.
Speaking on the Moonshots with Peter Diamandis podcast, Suleyman described the scale of resources needed to build advanced AI systems, comparing Microsoft to a “modern construction company” deploying vast numbers of workers to assemble gigawatts of CPUs and AI accelerators.
He noted that the costs extend beyond infrastructure, with salaries for top researchers and engineers adding significantly to the bill.
Microsoft, valued at $3.54 trillion, reported $77.7 billion in revenue for the quarter ending September, surpassing analyst expectations.
Suleyman said his mission is to make the company “self-sufficient” in developing frontier models and to establish “an absolutely world-class superintelligence team.”
He added: “We’re absolutely pushing for the frontier. We want to build the best superintelligence and the safest superintelligence models in the world.”
The executive has previously spoken of building a “humanist superintelligence” aligned with human interests.
However, he acknowledged that the enormous costs involved make it difficult for startups to compete with Big Tech.
“The ambiguity is what’s driving the frothiness of the valuations,” he said, suggesting that if an “intelligence explosion” occurs, multiple players could reach breakthroughs simultaneously.
Microsoft is not alone in its pursuit of artificial general intelligence (AGI) and superintelligence. Meta CEO Mark Zuckerberg recently said he would rather risk “misspending a couple of hundred billion” than fall behind in the race.
He argued that failing to act quickly could leave companies “out of position on what I think is going to be the most important technology that enables the most new products and innovation and value creation in history.”
The push for frontier AI has already driven massive spending on data centres and cloud infrastructure. Microsoft, Meta, Google, and Amazon have all ramped up investments to train and deploy increasingly complex models.
Analysts say the escalating costs highlight both the promise and the risks of pursuing superintelligence, with only the largest firms able to sustain the financial burden.
As Suleyman’s comments underline, the next decade of AI development will be defined not only by technological breakthroughs, but also by the capacity to fund them at unprecedented scale.
Sources: Business Insider, CNBC, The Verge, Bloomberg
