AI frenzy tests Big Tech’s newfound cost discipline

For tech investors, generative artificial intelligence has fuelled a powerful rally this year, helping pull the entire US stock market higher. While the sector’s companies have been only too willing to predict great things from the technology, one thing has been notably lacking: precise forecasts of exactly when generative AI will have an impact on the industry’s finances.

That uncertainty has loomed large in recent days as some of the biggest companies have reported their latest earnings, bringing a barrage of questions about the likely effect of the technology on costs and revenues.

Second-quarter earnings from the first batch of the biggest US tech companies — Alphabet, Amazon, Apple, Meta and Microsoft — beat expectations. But, along with other groups hoping to get a lift from generative AI, they have not been prepared to predict when new AI features will show up in their products and services or what they hope to charge for the technology, let alone the financial implications.

On Thursday, after Amazon and Apple completed the latest round of quarterly earnings reports, Brent Thill, an analyst at Jefferies, summed up the mood among investors: “The hype is here, the revenue is not.”

Despite the frustration, the earnings season at least helped to underpin this year’s big share price rises, with Meta, Amazon and Apple having risen more than 50 per cent. Uncharacteristically for a group of companies that often seem fixated on growth, cost discipline has played a central part in the earnings outperformance, following a rare round of jobs cuts earlier this year. Amazon has cut some 27,000 positions, while the lay-offs at Google led its headcount to fall by nearly 9,000 in the second quarter.

Amazon in particular finally showed signs of getting its ballooning costs under control, after a surge during 2022. Its operating profit margin climbed to 5.7 per cent in the second quarter, two points higher than the opening months of the year and its best performance in two years. The surprise recovery boosted its shares nearly 9 per cent in after-market trading, adding about $120bn to its stock market value.

Apple, which reported a 1 per cent slip in revenue as consumer spending on gadgets remained weak, also leaned heavily on cost-containment as a response to an unsettled economy.

“We know that this has been an uncertain period for the last few quarters,” said chief financial officer Luca Maestri. “We’ve been quite effective at slowing down the spend.” He added that Apple was “very pleased with our ability to decelerate some of the expense growth”, even as it consciously increased its spending on research and development at a faster rate.

Despite the greater attention to costs, however, a new investment wave caused by generative AI is already starting to loom, raising questions about whether Big Tech’s spending discipline will prove shortlived.

“The belt-tightening that happened in late 2022 and early 2023 was a moment in time,” said Jim Tierney, a growth stock investor at AllianceBernstein. “Are the strong margin rebounds in 2023 going to be fleeting?”

Microsoft, which has moved fastest to embed generative AI into its existing products and services, is set to register the pick-up in spending first. Capital spending in its latest quarter, at $10.7bn, was higher than Wall Street had expected. The software company said spending would rise higher still in each quarter of its current fiscal year, which began in July.

Analysts at Morgan Stanley predicted this would lift Microsoft’s capital spending on things such as data centres and the equipment to fill them to $50bn over the next 12 months. That would be more than 50 per cent up on the previous year, and more than double its spending only two years before.

Chief executive Satya Nadella claimed Microsoft would be able to tie this higher spending to its revenue growth as the AI services take hold, describing the planned investment as “replacement capital plus some new capital that is going to drive new growth”.

Other companies were less ready to predict exactly when the AI investment boom will hit, though they pointed to some level of higher spending in 2024. Meta chief Mark Zuckerberg said he couldn’t predict “how quickly our new AI products will grow”, leaving the company without a “clear handle” on how much it would need to invest.

Zuckerberg was among the tech executives to insist in recent days that even as they anticipate an increase in spending, recently discovered cost disciplines would remain in place. At companies such as Amazon and Facebook, that could lead to a period of increasing revenue and expanding margins, said Thill at Jefferies. “It’s growth returning with improving margins, it’s not growth at any cost.”

Pressed by investors to say when they would see a revenue lift from AI, Alphabet, Meta and Amazon all pointed to existing services that rely heavily on investments made in the past. These range from the AWS’s machine learning services for cloud customers to AI-enhanced tools that Google and Meta offer to their advertising customers.

When it comes to the newer generative AI wave, however, predictions were few and far between. Amazon chief executive Andy Jassy said on Thursday that the technology was in its “very early stages” and that the industry was only “a few steps into a marathon”.

Many customers of Amazon’s cloud arm, AWS, see the technology as transformative, Jassy said — though “most companies are still figuring out how they want to approach it, they are figuring out how to train models”. He insisted that every part of Amazon’s business was working on generative AI initiatives and the technology was “going to be at the heart of what we do”.

“Right now, the evidence is a little bit scarce” about what the effect on revenue will be across the tech industry, said Tierney at AllianceBernstein.

Even Microsoft offered only a cautious prediction of when this would show through in the form of higher revenue. Amy Hood, chief financial officer, told investors on an earnings call last week that the revenue impact would be “gradual”, as the features are launched and start to catch on with customers. The caution failed to match high expectations ahead of the company’s earnings, wiping 7 per cent from its stock price over the following week.

“Behind the scenes, the whole industry is scrambling to figure out the business model [for generative AI]: how are we going to price it? How are we going to sell it?” said Thill.