Meta, Apple, Microsoft, and Amazon all reported quarterly earnings this week, and there was a common thread tying them together: a boom in AI spending and plans to increase it even more, beyond analyst expectations.

Although capital expenditures above expectations often don’t tend to make investors particularly happy, it had pretty much the opposite effect this week, especially for Meta and Microsoft, both of which saw a pop in their stock following the releases.

And for Microsoft, which posted its largest ever quarterly capital expenditure forecast, the boost in shares led the tech giant to become the second-ever company to hit $4 trillion market valuation, when it briefly breached the threshold on Thursday.

The move was largely because both Meta and Microsoft finally had the revenue to show for their investments.

Meta’s ad revenue, which is a huge moneymaker for the tech giant, for the past quarter came in a couple billion dollars ahead of Wall Street expectations, and CEO Mark Zuckerberg attributed that to the deployment of artificial intelligence in the ad system. Zuckerberg went on to assure investors that this surprise increase in revenue was bound to continue, saying that his multi-billion dollar investment into building a team dedicated to creating “superintelligent” AI will lead to even more payoffs for its advertising business.

Microsoft reported that sales were up 18% from last year and that revenue for its cloud computing platform Azure had surpassed $75 billion this fiscal year, up 34% from last year. Revenue from the company’s productivity and business processes segment also exceeded expectations, and company executives shared that the business software sales were boosted partially thanks to widespread adoption of its AI product Microsoft 365 Copilot.

All the news combined brings to mind one question: Is Silicon Valley’s AI bet finally starting to pay off?

The AI spending boom

Meta has been in the midst of a multi-billion dollar AI push after Zuckerberg admitted that the company had fallen behind competitors in the AI race. The push has been marked by high profile strategic talent hires, and particularly the poaching of OpenAI employees tempted by multi-year deals worth millions of dollars.

In the meantime, the company is going all in on data centers as well. Last month, Zuckerberg said that Meta would be investing hundreds of billions of dollars into AI data centers. The company’s first of multiple multi-gigawatt data centers is to be unveiled next year, and Zuckerberg said in a post on his Threads account that just one of these data centers “covers a significant part of the footprint of Manhattan.”

This week, Meta said it’s expecting to shell out between $66 billion and $72 billion this year, and that it expects to spend even more next year on data centers and hiring.

Microsoft, on the other hand, said that it is expecting to spend more than $100 billion next year, with much of it going toward AI. This upcoming quarter alone, the company is eyeing $30 billion in capital expenditures, again mostly for AI, in what is a record forecast for the company.

Apple also posted better than expected revenue on its earnings report this week, but that was mostly attributable to iPhone sales. Despite that, CEO Tim Cook told investors during the company’s earnings call that the tech giant was planning to “significantly” increase its investments in AI to catch up with rivals and was open to acquisitions to do so.

Is AI demand finally catching up?

One of the biggest concerns when it comes to AI is regarding spending. Even though Silicon Valley is pouring in countless dollars—over $300 billion this year alone, according to numbers from the Financial Times—not everyone believes demand for AI will scale up accordingly. And if it does not, it would cause a major problem for the industry.

In a paper published earlier last month, the Federal Reserve claimed the biggest challenge with generative AI was not the potential of the tech itself but rather getting people and businesses to actually use it. The technology isn’t necessarily adopted widely outside of tech, science, and finance fields, and is deployed mostly by large firms. 

As the technology gets better, demand for AI is bound to increase, too, but by just how much is a mystery. If that demand does not grow as expected, the Fed paper warns, it could have “disastrous consequences,” much like the railroad overexpansion of the 1800s and the economic depression that followed.

The answer to whether or not AI demand will scale up to the level of investment is still not a definitive yes or a no, but this round of earnings gave a substantial dose of hope to the AI bulls.

But the risk for overspending is still there, as the tech giants continue to make record pledges of investment: if the increase in investments isn’t followed by a tangible increase in demand and revenue, especially for the companies’ core businesses, then the possibility of “disastrous consequences” is still there.