Is Goldman Sachs Not Afraid of an AI Selloff? $125 Billion Construction Plan Revealed—Is the AI Bull Market Not Over Yet?

As AI tech stocks plunge across the board, many investors have begun questioning everything, but Goldman Sachs is moving in the opposite direction, continuing to raise its forecast for U.S. data center construction spending and expecting it to reach $125 billion in 2028.
What does that mean? While the stock market is still debating whether AI is a bubble, the data center industry is already discussing where to build the next batch of facilities!
The main reason Goldman Sachs is bullish is that cloud service providers’ demand for computing power remains strong, and many projects are backed by a backlog of orders.
This means that even if AI stocks face a selloff in the short term, underlying infrastructure construction may not hit the brakes right away.
But there’s an easily overlooked issue here: strong demand doesn’t mean every company can make money.
Building data centers requires huge amounts of capital, and projects take time to go from planning to operation. Any change in interest rates, regulation, power supply, or financing conditions could put developers under enormous pressure.
Companies that borrow to build data centers especially fear not a lack of demand for computing power, but that loan bills will come due before their facilities start making money.
So the AI industry may still offer long-term opportunities, but the market is shifting from storytelling to scrutinizing financial results.
The real winners in the future probably won’t be the companies shouting the loudest about AI, but those able to turn computing power into cash flow.#BTCETF资金持续流出