Warren Buffett (워런 버핏), chairman of Berkshire Hathaway, said big tech companies' large-scale investment in artificial intelligence facilities cannot be dismissed as a simple bubble, and described it as a long-term infrastructure buildout.

On Aug. 1, IT outlet TechRadar reported that Buffett said in a CNBC interview in July that the recent surge in AI investment was "real money" and that current spending was not an experiment but a process of building physical infrastructure.

Major hyperscalers including Meta, Microsoft, Amazon and Alphabet have recently been pouring huge sums into AI data centres, semiconductors and network infrastructure. The industry also forecasts that AI-related spending by the four companies could reach up to $750 billion this year and up to $4.5 trillion in 2030.

Some in the market are raising concerns about overheating and the possibility of a bubble in AI investment because the profitability of AI businesses has not yet been sufficiently proven despite massive spending.

Buffett drew a line against that view. He compared the current scale of AI investment to construction of the U.S. railroad network in the 19th century, saying that kind of money did not go into the railroad business. He said AI investment, too, is not a short-term fad but a process of building infrastructure that will support future industries.

Buffett's view can also be seen in Berkshire Hathaway's investment in Alphabet. After newly buying 17.85 million Alphabet shares in November last year, Berkshire Hathaway also invested an additional about $10 billion in June this year. He was reported to have rated highly Alphabet's ability to secure enough cash flow to sustain massive AI infrastructure investment.

He assessed that in the AI era the importance of physical infrastructure such as data centres and computing resources is rising sharply from software-centred business structures. He has also acknowledged that his past judgement was wrong, having viewed software companies as asset-light businesses and been reluctant to invest, and the point is interpreted as meaning that in the AI era infrastructure competitiveness will determine corporate value.

He particularly said the scale of AI investment is on a different level from expansion of existing cloud businesses. Big tech companies have expanded data centres for cloud services, but the computing power, electricity and facility investment required by AI are far larger in scale, he said.

Buffett's remarks came as debate continues over overheating in AI investment. The AI industry is a field that requires massive capital spending on large-scale data centres, semiconductors, power and cooling facilities. As a result, the market is watching how long big tech companies can maintain an aggressive investment stance and whether such spending can translate into actual profitability.

Buffett said that, at least at this point, he views expanded AI investment as a process of building the foundation for next-generation industries rather than a bubble, and suggested that competitiveness will matter more for companies with the financial capacity to bear it.