In recent years, companies developing artificial intelligence have invested heavily in the construction and equipment of data centers as well as chips that speed up artificial intelligence calculations. However, these large-scale investments in AI also divert attention and resources from other technologies.
This year’s big projects of five leading AI companies are expected to reach around 700 billion dollars, as written by The Washington Post. Spending by Amazon, Google, Microsoft, Meta and Oracle, which equates to roughly three-quarters of the US military’s annual budget, will be nearly twice as much in 2025.
As Korean website The Chosun reported, Meta’s capital expenditures this year will come in at between $115 billion and $135 billion — up 60 percent to 90 percent from last year and about 60 percent of its annual sales. Amazon plans to invest $200 billion in AI, a 50 percent increase over last year.
According to some critics, these high initial costs will only be recouped for companies if artificial intelligence significantly transforms work and the economy, for example. Last year, JPMorgan calculated that the tech sector would need to generate $650 billion more in annual revenue to make a reasonable return on their investment. Now, this value would be even higher, as spending on AI development continues to rise. For example, OpenAI, the company behind ChatGPT, expects to make over $100 billion by the end of the decade.
It turns out that these large-scale investments in AI create a lack of resources for the development of other technologies. For example, Apple told investors last week that it is having trouble buying enough chips needed for iPhones and Macs. However, AI companies also need the same chips to equip their data centers.
AI projects also “steal” labor
Higher demand is thus also reflected in the price of chips, and therefore also in the prices of consumer electronics, such as mobile phones or computers. According to IDC analyst Francisco Jeronimo, manufacturers of smartphones and other electronics will either increase in price by at least five percent this year, or more will launch less powerful devices. However, such price increases can last for years and can drive smaller manufacturers out of the market.
In addition to chips, AI companies are also usurping labor. The lack of specialized workers, for example in the construction of factories or medical facilities, is beginning to show. The construction of data centers is often better paid, so some professions move to them.
OpenAI told the White House in the fall that its planned data centers would need about 20 percent of the existing workforce of skilled tradesmen, such as electricians and mechanics, in the coming years, and called on President Trump’s administration to strengthen their training system. It is estimated that half a million workers will be out of the U.S. construction industry next year.
Stock exchanges also reacted to the announcement of large-scale investments by companies. For example, Amazon shares fell eight percent on Friday after announcing a $200 billion investment in AI.
According to DA Davidson analyst Gil Luria, investors’ caution shows they are “healthily cautious,” Yahoo Finance reported. “That skepticism is probably healthier than in any previous cycle I’ve experienced,” he said. Investors are looking more closely than before at how tech giants are getting returns on their AI infrastructure investments, he said, especially after fears of a potential investment bubble intensified in the second half of 2025.
date:2026-02-08 13:22:00