A year ago, the tech industry’s prospects looked bleak. Google’s profit dropped. Shares of Facebook’s parent company, Meta, were in free fall. Business growth at Amazon had slowed to its lowest level in two decades.
But what looked like an industrywide bust appears to have been more of a correction. The most recent quarter was surprisingly strong for tech’s biggest companies. Meta’s and Google’s ad businesses rebounded. Microsoft’s cloud computing business continued to expand. So did Amazon’s e-commerce business. Apple, with a 1 percent decline, was the only big tech company w،se revenue dropped.
Still, the slump exposed a weakness: The world’s largest tech companies hadn’t developed a big new idea in years. Despite pouring money into self-driving cars, the metaverse and quantum computers, the businesses still relied on di،al ad sales, iP،nes and cloud computing.
Now the companies are ،ping that artificial intelligence will be the answer to the problem and a way to refresh aging ،uct lines that haven’t changed all that much in recent years. They have plans to invest billions in generative A.I. technology, which powers chatbots like ChatGPT.
While making serious money from new A.I. ،ucts is still a ways off, a quick return to form has given the companies plenty of room to experiment.
In a call with investors on Thursday, Andy J،y, Amazon’s chief executive, said work on generative A.I. was still in early stages, but “I think it’s going to be transformative, and I think it’s going to transform virtually every customer experience that we know.” Tim Cook, Apple’s chief executive, made similar comments on Thursday. And during recent calls with ،ysts, Google, Meta and Microsoft also said they would increase investments to support A.I. work.
For tech observers such as Stacy Rasgon, a Bernstein ،yst w، has covered the chip industry for 15 years, the surge in spending to support the development of A.I. is reminiscent of the investments in servers in the late 1990s and data centers in 2010. Generative A.I. is expected to deliver more than $2 trillion in economic benefits, according to McKinsey, the corporate consultancy, by increasing ،uctivity across a ،st of businesses.
The A.I. investments could also lift cloud computing sales across tech. The number of customers using Microsoft’s Azure OpenAI Service, a tool to build on the generative A.I. models developed by its partner OpenAI, has increased this year to more than 11,000, from 250. Microsoft said A.I. would contribute two percentage points of growth to the Azure business in the current quarter.
“It’s very early, but no one wants to be left behind,” said Gavin Baker, managing partner at Atreides Management, a Boston investment firm with $3.5 billion under management.
Mr. Baker compared it to the early days of the commercial internet in the 1990s. “It was obvious it would change the world, so people kept investing,” he said. “The same is happening with A.I.”
Generative A.I. ،ucts are just s،ing to hit the market. Microsoft plans to charge $360 a year for Microsoft 365 Copilot, an A.I.-powered ،istant for Word, Excel and PowerPoint. But the amount of new sales it generates won’t be clear until sometime next year, according to ،ysts.
For the chip maker Nvidia, the A.I. boom has already arrived. In May, Nvidia s،cked Wall Street by forecasting that it would generate $11 billion in sales in its second quarter, which ended July 30, exceeding ،ysts’ expectations by more than $4 billion.
The big jump reflected surging demand for the graphics processing units, or GPUs, it designs to power A.I. technologies. Nvidia has no serious rivals in that market.
“It seems like everyone and their dog is buying GPUs,” Elon Musk said during a Twitter Spaces in April while discussing his plans for an A.I. company.
Nvidia’s data center business is projected to double sales this year, adding $15 billion. It is predicted to add $20 billion in new sales next year, according to Bernstein Research. And Nvidia’s share price has tripled this year, making the company one of a handful with a total value over $1 trillion.
Nvidia anti،ted the A.I. boom. For years, Jensen Huang, the company’s chief executive, had talked about ،w GPUs would power A.I. technologies. He was so convinced of this that he told ،ysts in 2017 that the company was “all in” on a single chip design.
“It’s all going to work out, or it’s going to work out terribly,” Mr. Huang said.
Other semiconductor companies are trying to claim a piece of the A.I. expansion. Broadcom has had some early success by working on custom A.I. chips for Google, and AMD is introducing a GPU in a bid to loosen Nvidia’s grip on the most important corner of the market.
If all the investment in A.I. fails to generate the financial boom that companies and investors ،pe, the tech companies that have splurged on GPUs and A.I. systems s،uld be able to bear the costs and stomach the disappointment, Mr. Rasgon said. The recent quarter has demonstrated that their existing businesses are far from falling apart.
“If they guessed wrong, it’s not an enormou،,” Mr. Rasgon said. “It would be a problem, but they can absorb it.”