Hong Kong, China – 2025/03/01: This photo illustration shows perplexity’s artificial intelligence (AI) apps DeepSeek and ChatGPT on a smartphone screen.
Sopa Images | Light Rocket | Getty Images
As companies pour billions of dollars into artificial intelligence, HSBC Chief Executive Georges Ergederi warned on Tuesday of the mismatch between investments and returns.
Speaking at the Global Financial Leaders Investment Summit in Hong Kong, Elhederi said the scale of investment poses a challenge for companies, saying that while computing power for AI is essential, current revenue profiles may not justify such large expenditures.
Morgan Stanley predicted in July that global data center capacity will increase sixfold over the next five years, and that data centers and their hardware alone will cost $3 trillion by the end of 2028.
McKinsey said in an April report that by 2030, data centers capable of handling AI processing loads will require $5.2 trillion in capital investment to keep up with computing demand, while capital investment to power traditional IT applications is projected to be $1.5 trillion.
Mr. Elhederi said companies will be cautious because consumers are not ready to pay that price and the productivity benefits won’t materialize in a year or two.
He said: “These are kind of five-year trends, so this increase probably means that we will start to see the real revenue benefits and the real payment provisions for it, probably slower than investors expected.”
General Atlantic Chairman and CEO William Ford, speaking on the same panel, agreed: “Longer term, it’s going to create a whole new set of industries and applications, and there will be productivity rewards, but it’s a 10-20 year game.”
Big Tech companies Alphabet, Meta, Microsoft and Amazon have all raised guidance on capital spending, predicting that total spending will reach more than $380 billion this year.
OpenAI, which sparked an AI frenzy with the launch of ChatGPT in November 2022, announced nearly $1 trillion worth of infrastructure deals with partners including Nvidia, Oracle, and Broadcom.
Ford said the huge amount of spending being poured into this area shows that people recognize the long-term impact of AI. But the sector is initially capital-intensive and “you have to kind of pay it forward for future opportunities,” he said.
Ford warned that there could be “misallocation of capital, disruption, overvaluation… (and) irrational exuberance” in the early stages, adding that it may be difficult to discern winners and losers at this point.
“This is a broad technology, like railroads and electricity, that has had a huge impact over time and reshaped the economy, but we think it’s been very difficult to predict exactly how it’s going to change in the first few years.”
