Big Tech’s massive spending on artificial intelligence is expected to remain unchecked in 2025 after Amazon has surpassed its rival in more than $100 billion in infrastructure investment this year.
Spending by four major US high-tech companies had already surged to historic levels by 63% last year. Now executives have vowed to accelerate their AI investments and have dismissed concerns about the huge amounts they are betting on new technologies.
Microsoft, Alphabet, Amazon and Meta report total capital expenditures of $246 billion for 2024. This is up from $151 billion in 2023. It remains at the forefront of research into major AI language models.
The scale of their spending ambitions was announced along with fourth quarter revenues – which surprised the market and was triggered by the release of an innovative and inexpensive AI model from Chinese startup DeepSeek in late January It made the sale worse.
Microsoft and Google’s parent alphabets each were wiped out of their $20 billion market value after reporting weaker than expected growth in the cloud computing sector along with a sharp rise in capital spending. Google’s 8% decline on Wednesday was its fifth trading day in the last decade.
“The unlimited enthusiasm across the ‘magnificent Seven’ has been replaced by a pocket of skepticism, creating some ‘show me’ situations,” says Jim Tierney, head of Alliancebernstein’s Intensive US Growth Fund. states. “The concerns I’ve had since this summer are growing.”
In the hype about AI’s transformational potential, shareholders are otherwise repaid in the form of buybacks and dividends by doubling their spending without compromising increases in revenue, and are not starved. I’m worried that I can eat to capital to starve the line.
Google is unclear about its use and revenue from Gemini Chatbot, and companies are hiring Microsoft’s glitches and costly co-pilot “agents” to increase workforce productivity.
“If we see accelerated cloud growth on Google or (Microsoft’s) Azure, or if Copilot ingestion improves, investors will be more comfortable with spending at Alphabet or Microsoft,” Tierney said. Masu. “A cheaper and more commercialized AI model will probably amplify investors’ concerns during that time.”
Deepseek’s R1 model represented such fear. The claims of China’s AI Labs are that they build inference models with similar capabilities to Google and Openai products at just a fraction of the price, and because Nvidia’s most advanced graphics processing units are inaccessible, chip manufacturer stocks are not available. It plummeted 17%, eliminating $600 million. One day, it only recovers partially.
The Big Technology Chief holds nerves. On Tuesday, Google’s Sundar Pichai defended its plan to increase $75 billion by 42% in 2025 from $530 billion last year, saying that the AI opportunity was “as big as it’s coming, and that’s why you’re me.” He said he saw them investing, to meet at that moment.” Deepseek will add to demand by showing how new techniques can make it cheaper and encourage new research lines, he said.
Satya Nadella of Microsoft said in Davos two weeks ago: He reiterated his belief in the stupidity of slowing down early support for startup Openry and not capitalizing.
And on Thursday, Amazon CEO Andy Jassy broke through Google and Microsoft by forecasting more than $100 million in capital spending this year. The majority are directed at Amazon Web Services data centers and servers, and Jassy said it was simply responding to “significant signals of demand.” Stocks fell by 7% in after-hours trading.
“The growth is cooked a little, but the appetite for investment has not been reduced,” said Jeff Pearson, vice-chairman of Cloud Strategy at Consultancy Presidio. “They cultivate first, even if their return on investment seems to be far away.”
Meta received a more aggressive reception for revenue, but for Mark Zuckerberg promised to spend “thousands of billions” more on AI, in addition to the $400 billion invested in 2024. Stocks also rose.
Referring to the use of AI to improve Facebook and Facebook ad targeting, Tierney said: Instagram.
Meta’s success in showing concrete returns from AI investments contrasts with new competitors and Google, facing the challenging task of integrating AI into searches without cannibalizing their core advertising business. It was.
Search giants have introduced a simple answer, or “AI Overview,” at the top of their search results, but these replace the list of links.
Nevertheless, “If you intend to have a crack in Google’s Search Empire, it certainly hasn’t appeared yet,” says Bernstein analyst Mark Shmulik. Alone. “Google hasn’t missed even one search expectation since ChatGpt was launched nine quarters ago.”
Recommended
According to Société, spending on “Magnificent Seven” (including Apple, Nvidia and Tesla) including Apple, Nvidia and Tesla warns the US benchmark S&P 500. Generale. Elite group profits skyrocketed by a third over the same period, with the remaining 5% at 5%.
The existence of spending is not limited to publicly available companies, nor is its deep exploration or fear of the AI bubble slowing down the flow of capital to Silicon Valley startups.
Openai’s Sam Altman has partnered with Softbank and Oracle to invest $100 billion in AI-related US infrastructure, potentially rising to 50 trillion over time. Japanese investors are in discussions to invest $25 billion in startups at a $260 billion valuation.
“Are there any AI winters at some point? Certainly, Rishi Jaluuria, an analyst at RBC Capital Markets, said, “But if you’re in a position to be a leader, you’ll take your foot off the gas. You can’t do that.”