This is the amount that you can expect to pay. Capital pouring into AI-driven datacenter projects is staggering. Microsoft, Alphabet and Amazon released their quarterly results last week. 2025 capital expenditures They expect this number to continue rising through 2026. Microsoft spent the most money in its data centers, and on other investments. That’s equivalent to nearly 35 billion dollars, or 45 percent, of Microsoft revenue, last quarter.
This amount of money has never, or hardly ever been absorbed by a technology so quickly. The warnings about an AI bubble have been increasing, but the hype is already changing the US economy. Harvard economist Jason Furman estimated that investments in data centers, software processing and other technologies accounted for nearly all of US GDP growth In the first half 2025.
We’ll be looking at the impact of data centers on three key areas today: jobs, public markets and energy.
Cashing out
AI has been a major factor in the US stock market’s boom. AI-related stock prices have increased by over 50% since ChatGPT’s launch in November 2022. 75 percent of S&P 500 returns Michael Cembalest of JPMorgan says this is the case. It is now a question of whether this growth can be sustained as technology firms continue to spend heavily on AI infrastructure.
As financial journalist Derek Thompson noted, at the beginning of 2018, tech giants funded their AI projects primarily with cash on hand. Financial journalist Derek Thompson pointed outIn 2025, the 10 largest US publicly traded companies began with cash margins historically high. Their businesses were so successful that billions of dollars were available to spend on Nvidia GPUs or data centers.
The trend continued through to 2025. Alphabet told its investors, last week, that it expected its capital expenses to be $93 billion this year, up from $75 billion. It also announced that revenues were up by 33 percent compared to last year. Silicon Valley both spends and earns. This means that everything is okay, right?
But not quite. It appears that tech giants are using accounting tricks They may try to inflate their numbers. Nvidia is the main beneficiary of AI funding. new versions It replaces its GPUs every 2 years. Microsoft and Alphabet estimate that their GPUs will last for six years. If they need to upgrade sooner to stay competitive—a likely possibility—that could wind up eating into their profits and weaken their overall performance.
