Stay up up to now with free updates
Simply register for Large technology firms myFT Digest – delivered straight to your inbox.
In the race to construct the infrastructure to support artificial intelligence, major technology firms have increased their capital spending by 50 percent this yr, to greater than $100 billion, at the same time as skepticism grows on Wall Street concerning the return on these unprecedented investments.
Microsoft, Alphabet, Amazon and Meta all announced massive spending increases in the primary six months of 2024 of their latest quarterly reports – totaling $106 billion – while their bosses ignored stock market jitters and announced further investment increases over the following 18 months.
“At this point, I’d slightly take the chance of constructing the capability before it is required than too late,” Meta CEO Mark Zuckerberg said this week, predicting that capital spending at Facebook's parent company could reach $40 billion this yr.
According to their joint forecasts, AI-related investments by Big Tech firms could greater than double by year-end. Analysts at Dell'Oro Group now expect as much as $1 trillion to flow into infrastructure comparable to data centers inside five years, although the businesses have to date did not persuade investors that their customers are willing to spend big on AI services and products.
“Technology management teams are betting on more spending,” said Jim Tierney, head of U.S. growth at AllianceBernstein. “Investors are still not clear on all of the business models and earnings. This creates a 'trust us' environment that is just not exactly reassuring given the general spending.”
The latest round of earnings reports from major technology firms collided with a general decline in sentiment on Wall Street, because the Nasdaq entered a correction on Friday because of weaker U.S. jobs numbers. Shares of semiconductor firms, including leading AI chipmaker Nvidia, were particularly volatile this week as investors became more sensitive to comments from major technology firms on their spending plans.
Nvidia gained or lost about $200 billion in three consecutive trading sessions, while Intel, which has not yet been in a position to attract the majority of AI infrastructure spending because of a scarcity of competitive products, lost greater than 1 / 4 of its value on Friday after announcing massive layoffs.
But at the same time as shares of Google, Microsoft and Amazon sold off immediately after the discharge of their earnings reports, executives at major technology firms made no apologies for his or her buying spree.
Zuckerberg estimated that the computing power required to coach the following big language model shall be “almost 10 times higher” than the previous version, but he acknowledged that it can take “years” before a few of its AI features, comparable to the Meta AI chatbot, start earning profits “on their very own.”
“When you undergo these sorts of transitions within the technology industry, the chance of underinvesting (in AI) is dramatically higher than overinvesting,” said Google CEO Sundar Pichai.
After Google's parent company Alphabet last week reported a 90 percent increase in capital spending to $25 billion in the primary two quarters of 2024, Microsoft responded on Tuesday with a 78 percent increase to $33 billion. Amazon's investments in real estate and equipment in the primary half of the yr – which include spending on its vast e-commerce and logistics network – rose 27 percent to $32.5 billion, the corporate announced on Thursday.
Amazon said in a subsequent statement that it expects a “significant increase” in total capital spending in 2024, most of which can go toward latest cloud infrastructure, in line with Chief Financial Officer Brian Olsavsky. Generative AI is now a “multi-billion dollar business” for the corporate, he added. But amid the general market slump and signs of weakness in the buyer business, Amazon shares fell 8.8 percent on Friday.
Google's management had pointed to stable promoting revenues – a rise of 11 percent to $64.6 billion within the second quarter alone – as evidence that the corporate's core business was healthy enough to bear the burden of expenses.
While this didn’t stop Alphabet's share price from falling, investors in Microsoft seemed more reassured after Chief Financial Officer Amy Hood said the corporate's data centers were long-term assets that might be “monetized over 15 years or more.”
If demand for AI services and products is lower than expected, Microsoft may slow the pace at which it fills its data centers with expensive AI hardware, Hood said. Last quarter, cloud growth was limited by a scarcity of capability, not customer demand, Microsoft said.
Much of the investment by major technology firms goes into buying land and constructing latest data centers for his or her cloud computing businesses. Huge sums are also spent on hardware, including special chip clusters – mainly from Nvidia – needed to coach and run the massive language models on which chatbots are based.
Demand for cloud services has surged as firms test generative AI services to automate processes and increase productivity, even when most of those experiments haven’t yet moved into production. At the identical time, startups like OpenAI, Anthropic, Elon Musk's xAI and France's Mistral are competing for scarce computing resources to coach increasingly advanced LLMs.
This week's price swings follow a historic bull run. The technology-dominated Nasdaq 100 index has risen about 70 percent since early 2023, when enthusiasm for AI first began to rise, making Apple, Microsoft, Nvidia, Alphabet and Amazon the five largest publicly traded firms on the earth.
“I feel the natural comparison that involves mind for a lot of investors is the telecom bubble of the late Nineties and early 2000s,” says Morningstar analyst Michael Hodel. “Most of the businesses involved in that buildout went bust. This buildout seems similar in some ways… The key difference, nonetheless, is that the businesses doing many of the buildout have enormously profitable existing businesses and fortress-like balance sheets.”
