In Nevada, a summer of weak international tourism depressed employment and weakened the local economy. But the building boom Data centers The connection to the fast-growing AI industry helped soften the blow.
In the Washington, D.C., area, federal job cuts and the longest government shutdown on record threatened to push the regional economy into recession. But there are also AI-related investments that help offset the damage.
In North Dakota, low oil prices have idled drilling rigs and reduced state revenues, however Data centers Artificial intelligence technologies help bridge this gap.
The US economy in 2025 is divided into two parts: everything related to artificial intelligence is expanding. Almost everything else is not.
AI developers and chip makers are raising hundreds of billions of dollars in investments. Data centers The scale of theme parks is showing across the country. Energy utilities are racing to build new plants and revitalize old ones to meet growing electricity demand.
- “Shock and turmoil”: Elon Musk predicts a future in which artificial intelligence and robots make work “optional”
Workers with the right skills – from the developers who build artificial intelligence models to the electricians who install the equipment that makes them work – earn high wages.
As for the rest of the economy, the scenario is different. Unemployment rates have risen, hiring has slowed, and sectors such as manufacturing and homebuilding have begun to cut jobs. Consumer confidence declined amid rising prices.
The public sector has been under pressure due to federal budget cuts and layoffs. Tariffs, and the uncertainty surrounding them, have hurt international trade and prompted many companies to reduce their investments.
“It’s a two-speed economy,” said Mark Morrow, an economist at the Brookings Institution who has studied the impact of artificial intelligence on local economies. -The AI gold rush is generating all the excitement and covering up the stagnation in the rest of the economy.
- Will artificial intelligence be the end of Google? Learn why it’s not expected to challenge the search engine monopoly
By one measure, investments in computer hardware and software accounted for more than 90% of GDP growth in the first half of the year. While economists caution against taking these numbers too seriously — if not for AI, some of this money would have gone elsewhere — they say there is no doubt that investments in AI help explain the economy’s surprising resilience this year.
But relying on artificial intelligence as a source of growth raises a question for economics: What happens if the gold rush stops?
This threat is most evident in the stock market, which has set record after record in recent months, driven largely by the strength of a handful of companies focused on artificial intelligence. Seven companies, known as the “Seven Wonders”, including Amazon, Microsoft and Google’s parent company Alphabet, now account for more than a third of the value of the S&P 500.
- capital: Pressure is mounting on Brazilian executives to show a return on investment in artificial intelligence
Only one of the so-called Magnificent Seven companies, Nvidia, which makes the chips that run many advanced language models, recently had a market cap — albeit briefly — of $5 trillion.
Such lofty valuations are based on the assumption that recent rapid growth will continue for years, which some investors have warned may be unrealistic. Even Sam Altman, CEO of OpenAI, said in August that he thinks investors are very excited about AI. Nvidia’s strong quarterly earnings report on Wednesday failed to fully dispel these doubts.
The bursting of the bubble – whenever it happens – may have tangible effects. Household consumption in recent quarters has been increasingly driven by higher-income households, which have continued to spend even as many lower-income households have reduced their spending. But if the stock market falters, wealthier households may also cut back on spending.
“If consumption growth is dominated by households that have benefited most from the market performance of AI-related stocks, a stock correction could actually be very painful for the economy,” said Aditya Bhave, US economist at Bank of America. —This creates a certain degree of fragility.
- Jeff Bezos: Investments in artificial intelligence may generate a “bubble,” but they will bring benefits
Low-income families did not see the benefits of the stock market rise. But they can still be harmed by its reflection. If wealthy Americans reduce their spending on restaurant meals, vacations and luxury goods, it could lead to job losses in the service sector.
“If you see a decline in leisure and hospitality spending, in that top group, that will have ripple effects,” said Michael Reed, an economist at RBC Capital Markets. “This is where I question the downward spiral in the job market.”
For now, the boom — and its support for the broader economy — shows little sign of abating. US companies spent more than $60 billion on computing equipment in the second quarter, an increase of 45% from the previous year. They spent another $10 billion on construction Data centersan increase of 35%. Economists and industry experts say AI is likely responsible for most of this growth.
- COP 30: Renewable energy supplies make Brazil attractive for AI data centers, but containing the impact on the environment is a challenge
These investments diffuse and sustain parts of the economy outside the traditional technology sector. Caterpillar, typically associated with its excavators and backhoes, has seen an increase in sales of turbines and power generation equipment used in data centers.
Johnson Controls, another industrial group, has benefited from demand for cooling and fire suppression systems. Eaton, which makes energy management systems, has a growing backlog of projects waiting for its equipment.
Companies are betting on this growth to continue. Eaton is investing more than $1 billion to expand its manufacturing capacity to meet growing demand, and announced this month that it is spending $9.5 billion to acquire Boyd Thermal, which makes cooling equipment used in data centers.
“We are still in the early stages of AI expansion,” Paulo Ruiz, CEO of Eaton, said in an interview.
- I estimated: Investments in data centers could reach US$11.4 billion in 2026
The two-speed nature of the economy is particularly evident in the construction sector, which has been hit hard by high interest rates and tariffs. Nonresidential construction spending fell in August, continuing its downward trend. Housing construction is well below the peak seen during the pandemic.
Even as the construction sector slowed, data center construction increased. This also created the infrastructure needed to operate data centers and distribute information. The American Cement Association released a report in June estimating that AI data centers will use nearly 1 million tons of cement over the next three years.
Data center construction is “the only real driver of nonresidential construction spending growth in the United States,” said Anirban Basu, chief economist for the Builders and Contractors Association, an industry association.
– If one is looking for what drives non-residential spending, this is it Data centers And associated investments in energy generation and distribution.
- Zuckerberg: Meta’s CEO says he will launch a series of giant data centers to boost his AI efforts
This increase has also benefited contractors working on project components Data centers. Associated Builders and Contractors recently said that about one in seven of its members have been hired to work there Data centers.
Chuck Goodrich, CEO of Gaylor Electric, an Indianapolis-based electric utility, said 50% to 60% of his business is now tied to building power plants. Data centers. He estimated that Gaylor Electric’s revenues would increase about 30% this year, to $1 billion, and grow another 20% next year, largely due to the race around. Data centers.
“From a biblical perspective, we live a life of abundance,” he said.
But this abundance does not reach everyone, nor does it reach every place. Most of the companies developing leading AI models, such as OpenAI and Google, are located in the San Francisco area, and their highest-paid employees are concentrated in major cities, which have been the winners in the 21st century economy.
Investments in AI infrastructure are more widespread, but the benefits are much less certain. Some of the biggest Data centers They are being built in rural areas, where land is cheap and jobs have been disappearing for decades. Local leaders in many communities have embraced this idea Data centers and other AI infrastructure as a way to diversify their economies and secure entry into a thriving sector.
- Data centers: Access to the electricity grid and delays in regulation represent obstacles to projects in Brazil
But data centers consume water and electricity voraciously, which can lead to higher costs for residents. While they create jobs during construction, they employ relatively few people once they are up and running, turning the vast campuses into vibrant ghost towns. Microsoft, for example, said that Data centers Employing approximately 50 people per building, 24 hours a day.
Many large companies Data centersknown as “hyperscalers”, received generous tax incentives from local communities, which limited the benefits to residents. While local leaders say the projects will attract other technology companies, it is unclear whether that will actually happen, said Morrow, the Brookings Institution economist.
“Some places have been confused by a lot of messages from hyperscalers saying this is the beginning of the regional tech economy,” he said.
- Chinese power: China wants 100,000 Nvidia chips to build AI data centers in the middle of the desert
Instead, he added: “It’s just huge buildings with a few hundred jobs, which is not a bad thing, but it’s not going to make a real difference.”
However, many economists and industry experts say the infrastructure-building phase of the AI boom is still going strong. Even as data center capacity expands at a record pace, demand is growing just as quickly, and companies are reporting waiting lines that last for years.
Demand for infrastructure is likely to continue to rise, said Paul Ashworth, chief North America economist at Capital Economics.
– Actually, this is just the beginning. He added that the stock market may be close to a bubble, but there is no indication that the industry has bought too many chips or built too many data centers.
- “Vampires Content”: An increasing number of companies are waging a war against AI bots
But for these investments to pay off, AI will have to fulfill its promise not just as a useful tool, but as a transformative technology that leads to massive increases in productivity.
A lot of the investments made may end up being unjustified, said Bhave, the Bank of America economist.