SEATTLE — If 2023 was the tech industry’s year of the artificial intelligence chatbot, 2024 is turning out to be the year of AI plumbing. It may not sound as exciting, but tens of billions of dollars are quickly being spent on behind-the-scenes technology for the industry’s AI boom.
Companies from Amazon to Meta are revamping their data centers to support AI. They are investing in huge new facilities, while even places like Saudi Arabia are racing to build supercomputers to handle AI. Nearly everyone with a foot in tech or giant piles of money, it seems, is jumping into a spending frenzy that some believe could last for years.
Microsoft, Meta, and Google’s parent company, Alphabet, disclosed this past week that they had spent more than $32 billion combined on data centers and other capital expenses in just the first three months of the year. The companies all said in calls with investors that they had no plans to slow down their AI spending.
In the clearest sign of how AI has become a story about building a massive technology infrastructure, Meta said Wednesday that it needed to spend billions more on chips and data centers for AI than it had previously signaled.
“I think it makes sense to go for it, and we’re going to,” Meta CEO Mark Zuckerberg said in a call with investors.
The eye-popping spending reflects an old parable in Silicon Valley: The people who made the biggest fortunes in California’s gold rush weren’t the miners — they were the people selling the shovels. No doubt Nvidia, whose chip sales have more than tripled over the past year, is the most obvious AI winner.
The money being thrown at technology to support AI is also a reminder of spending patterns of the dot-com boom of the 1990s. For all the excitement around web browsers and newfangled e-commerce websites, the companies making the real money were software giants including Microsoft and Oracle, chipmaker Intel, and Cisco Systems, which made the gear that connected those new computer networks together.
But cloud computing has added a new wrinkle: Since most startups and even big companies from other industries contract with cloud computing providers to host their networks, the tech industry’s biggest companies are spending big now in hopes of luring customers.
Google’s capital expenditures — largely the money that goes into building and outfitting data centers — almost doubled in the first quarter, the company said. Microsoft’s were up 22%. Amazon, which will report earnings Tuesday, is expected to add to that growth.
Meta’s investors were unhappy with Zuckerberg, sending his company’s share price down more than 16% after the call. But Zuckerberg, who just a few years ago was pilloried by shareholders for a planned spending spree on augmented and virtual reality, was unapologetic about the money that his company is throwing at AI. He urged patience, potentially for years.
“Our optimism and ambitions have just grown quite a bit,” he said.
Investors had no problem stomaching Microsoft’s spending. Microsoft is the only major tech company to report financial details of its generative AI business, which it said had contributed to more than one-fifth of the growth of its cloud computing business. That amounted to $1 billion in three months, analysts estimated.
Microsoft said its generative AI business could have been even bigger — if the company had enough data center supply to meet the demand, underscoring the need to keep on building.
The AI investments are creating a halo for Microsoft’s core cloud computing offering, Azure, helping it draw new customers. “Azure has become a port of call for pretty much anybody who is doing any AI project,” Microsoft CEO Satya Nadella said Thursday.
Google said sales from its cloud division were up 28%, including “an increasing contribution from AI.”
In a letter to shareholders this month, Amazon CEO Andy Jassy said that much attention had been paid to AI applications, such as ChatGPT, but that the opportunity for more technical efforts, around infrastructure and data, was “gigantic.”
For the computing infrastructure, “the key is the chip inside it,” he said, emphasizing that bringing down costs and wringing more performance out of the chips are key to Amazon’s effort to develop its own AI chips.
Infrastructure demands generally fall into two buckets: First, there is building the largest, cutting-edge models, which some AI developers say could soon top $1 billion for each new round. CEOs said that being able to work on developing cutting-edge systems, either directly or with partners, was essential for remaining at the forefront of AI.
And then there is what’s called inferencing, or querying the models to actually use them. This can involve customers tapping into the systems, such as an insurer using generative AI to summarize a customer complaint, or the companies themselves putting AI directly into their own products, as Meta recently did by embedding a chatbot assistant in Facebook and Instagram. That’s also expensive.
Data centers take time to build and outfit. Chips face supply shortages and costly fabrication. With such long-term bets, Susan Li, Meta’s finance chief, said the company was building with “fungibility.” It wants wiggle room to change how it uses the infrastructure, if the future turns out to be not exactly what it expects.
This article originally appeared in The New York Times.
