new york
CNN
—
As tech earnings season approaches, one big question is on Wall Street’s mind: When will we start making real money from artificial intelligence?
Eighteen months after ChatGPT kicked off the AI arms race, tech giants have promised the technology would revolutionize every industry, justifying spending tens of billions of dollars on the data centers and semiconductors needed to run large-scale AI models. Compared to that vision, the products they’ve rolled out so far feel somewhat trivial: chatbots with no clear path to monetization, cost-cutting measures like AI coding and customer service, and AI-enabled search that sometimes fudges the facts.
But big tech companies have spent billions but have yet to see much in the way of big revenue growth from AI or lucrative new products, and investors are starting to get nervous.
Amazon (AMZN) reported disappointing earnings and guidance on Thursday, likely due in large part to concerns that its big investments in AI are not paying off as its core business faces challenges, which led to its stock dropping nearly 9% on Friday. Intel (INTC) shares plummeted 25% on Friday after the company said Thursday night that after investing heavily to adapt to the AI wave, it was trying to rein it in with $10 billion in cost cuts and laying off tens of thousands of employees.
Ultimately, investors’ concerns boil down to whether there is actually value in all of this, or whether it’s just a shiny object that the industry is chasing again dreams of infinite growth before abandoning them and moving on to the next big thing.
“There’s a lot of debate going on in the industry right now about the[capital expenditure]requirements around generative AI and whether the monetization is actually worth it,” Morgan Stanley analyst Keith Weiss said on Microsoft’s earnings call.
UBS analyst Stephen Ju asked Google CEO Sundar Pichai how long it will take for AI to “help generate revenue rather than just reduce costs and create significant value in the long term?”
And a Goldman Sachs report last week raised the question of whether we are “spending too much and reaping too little” on generative AI.
Shares of Google and Microsoft fell after their earnings reports, a sign of investor frustration that the companies’ huge AI investments didn’t translate into better-than-expected results. Meta, which experienced similar shareholder dissatisfaction last quarter, avoided the same fate this time around by showing that its AI investments were at least contributing to its core business, such as making it easier for companies to use its AI tools to create compelling ads.
DA Davidson analyst Gil Luria told CNN that some investors had expected this to be the quarter in which tech giants would start to show signs of scaling back investments in AI infrastructure because “AI isn’t delivering the benefits that we were hoping for.”
The results were the opposite. Google, Microsoft and Meta all signaled they plan to spend more to lay the groundwork for what they hope will be an AI future. Meta now expects full-year capital expenditures of $37 billion to $40 billion, raising the low end of its guidance by $2 billion. Microsoft said it expects to spend more in fiscal 2025 than it has in 2024, up from $56 billion in 2024. Google predicted “more than $12 billion” in capital expenditures per quarter this year. (That’s big numbers, even for the very wealthy: Google’s second-quarter capital expenditures amounted to about 17% of total revenue.)
And technology leaders say what we need is more time, more time.
Microsoft CFO Amy Hood said on the company’s earnings call that the data center investments are expected to help monetize the company’s AI technologies “for the next 15 years or more.”
Similarly, Meta expects “benefits from generative AI will be realized over a longer period of time,” CFO Susan Li told analysts. “Generative AI is a space where we are at a fairly early stage. We do not expect our generative AI products to be a significant driver of revenue in 2024. However, we do expect new revenue opportunities to emerge over time, allowing us to better generate benefits from our investments.”
That timeline is unsettling for many investors who have grown accustomed to Silicon Valley’s roughly steady quarterly revenue and profit growth.
“If you invest now and you’re looking to make a profit in 10 to 15 years, that’s a venture investment, not an investment in a public company,” Luria said. “With a public company, you expect the return on your investment to be much shorter. That’s what’s creating anxiety, because right now we just don’t see the types of applications or revenue from those applications that are needed to justify these investments.”
And some investors are skeptical that AI investments will pay off: “AI technologies are not designed to solve complex problems that are cost-effective,” Goldman Sachs analyst Jim Covello argued in a report last week.
As an example of how long it may take for AI products to become reality, take Tesla’s AI-based “full self-driving” technology. Tesla has touted this driver-assistance technology as key to its business plan since 2015, and has consistently promised that it will be fully functional in a short time. But FSD still requires an attentive human driver who can take the wheel if something goes wrong, and nearly four years after it was first released to Tesla customers, safety concerns remain frequent.
For now, tech CEOs seem to agree, as Google’s Pichai said on an earnings call last week, that “the risks of underinvesting are significantly greater than the risks of overinvesting.” (Meta CEO Mark Zuckerberg echoed that sentiment on the company’s earnings call.) Data centers take time to build, so if there’s a winner in the AI race, no company wants to miss out on the top spot just because they don’t have enough computing power. And they’re making enough money from their core businesses that investors are likely to hold off on spending for now.
But in the near future — Luria predicts it could happen later this year or early next — investor pressure to back off infrastructure spending and catch up with revenue growth will be strong enough to prompt tech companies to pull back.
“The situation right now is, ‘We want to maintain our leadership position, so we all have to show we’re willing to invest as much as necessary.’ But at some point the investment becomes too burdensome and one of the companies will say, ‘Maybe next quarter I’m not going to invest as much,’ and then the rest of them will,” Luria said. “Overall, this level of investment is not sustainable.”
