img

Alibaba Group Holding Ltd.’s Chairman Joe Tsai is constructing a case around infrastructure and tech stating that the break-neck pace of development in datacenter construction could overtake the demand that AI services would make. He distinctively stated that a bubble could form in this particular space.  

Tsai, while speaking at the HSBC Global Investment Summit in Hong Kong on Tuesday claimed that the frenzy from the US and Asia to build server bases is starting to feel out of control and lacks proper vision. He noted how many of these projects were being constructed without any definite clients in the scope.      

Contrary to this statement, we see softbank, medicines, and Microsoft Corp. hurling billions of dollars on Nvidia and SK Hynix Inc. for their AI and tech chips so arbitrarily that everyone seems to be going in both directions. As opposed to what Alibaba stated about going head first into AI in february, they are seemingly restraining their plans for investment and only planning to spend upwards of 52 billion dollars over the next 3 years while server farms are made in India, Malaysia and other regions. Meanwhile, there’s Trump boasting about a Stargate he claims will spend half a trillion dollars on it.

A significant number of analysts in Wall Street have become skeptical of the spending, especially now after the release of DeepSeek's open source AI model, which is claimed to rival US technology, claims. Additionally, critics have noted that AI lacks a sufficient number of practical applications that can be implemented in the real world. The Chinese startup released an update to its V3 model Tuesday that claims to enhance programming capabilities.

To the delegates of the summit, he said, “I start to see the beginning of bubble.” He added that a few of the projected ones started the process of gathering capital with no "uptake" contracts, as he called them. “I’m beginning to be concerned when people are speculatively building data centers. There are a number of people coming up, funds coming out, to raise billions or millions of capital,” Tsai said.

Tsai pointed out specifically US spending, saying Amazon ‘only’ plans to build 4 data centers. This year, however, was notably worse. For AI infrastructure spending, Amazon.com Inc, Alphabet Inc. and Meta Platforms Inc. are making expenditures of 100, 75, and 65 billion dollars sequentially.

But in February, Microsoft has been rumored to be pulling out of other leases for US data center capacity, because TD Cowen analysts emphasized that Microsoft appears to be over spending in AI computing capacity, bringing with it doubts over the long term usefulness.

The company’s leaders have tried to minimize those worries, claiming that Microsoft’s current expenditures, which use to $100 billion on semiconductor and data centers, are at an all-time high.\n\n“Those numbers people are talking about investing in AI in the U.S. is still astounding to me,” Tsai said.\n\n“People are literally talking about several hundred billion dollars to the tune of 500 billion dollars […] While I think there’s some demand that I feel people are investing into for what they’re seeing today is too high, I do think it is too pessimistic.”

 


Read More: How to Use Grok AI on Telegram: A Guide for X Premium Users