Microsoft AI CEO says there’s a huge price to pay for staying in the AI game.
Mustafa Suleyman said in an episode of the “Moonshots with Peter Diamandis” podcast published Wednesday that it’s going to cost “hundreds of billions of dollars” to compete at the frontier of AI over the next five to 10 years.
“Not to mention the prices that we’re paying for individual researchers or members of technical staff,” he added.
Suleyman compared Microsoft to a “modern construction company,” where hundreds of thousands of workers are building gigawatts of CPUs and AI accelerators.
The scale of investment that is required is huge, and “clearly there’s a structural advantage by being inside a big company,” he said.
Microsoft, which has a market capitalization of $3.54 trillion, brought in $77.7 billion in revenue for the quarter ending in September, surpassing analysts’ estimates.
Suleyman said his mission is to make Microsoft “self-sufficient” in developing its frontier models and to build “an absolutely world-class superintelligence team.”
“We’re absolutely pushing for the frontier,” Suleyman said. “We want to build the best superintelligence and the safest superintelligence models in the world.”
Suleyman said last month that his team is “trying to build a humanist superintelligence” — one that is aligned with human interest.
With the high cost required to keep up with AI, Suleyman said on the podcast that “it’s hard to say” if startups could compete with Big Tech.
“The ambiguity is what’s driving the frothiness of the valuations,” he said. “If suddenly we do have an intelligence explosion, then lots of people can get there simultaneously.”
The high cost of AI
Microsoft is one of several tech giants chasing artificial general intelligence, or superintelligence, and executives across the industry have been blunt about how expensive that pursuit will be.
Artificial general intelligence, or AGI, refers to AI systems that can match human intelligence across most tasks. Superintelligence goes a step further — systems that surpass human abilities.
Meta CEO Mark Zuckerberg said in September he’d rather risk “misspending a couple of hundred billion” than fall behind in superintelligence.
If superintelligence arrives earlier than expected and a company moves too slowly, it’ll be “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 and history,” Zuckerberg said.
Billions of dollars have also been poured into AI data centers. In recent months, Big Tech firms like Microsoft, Meta, Google, and Amazon have ramped up spending on cloud and compute infrastructure to train and run frontier models.
