Join us on: Whatsapp
View All Articles on AI
Leading AI companies are spending mountains of cash in hopes that the technology will deliver outsize profits before investors lose patience. Are exuberant bets on big returns grounded in the quicksand of wishful thinking?
The fear: Builders of foundation models, data centers, and semiconductors plan to pour trillions of dollars into infrastructure, operations, and each other. Frenzied stock investors are running up their share prices. But so far the path to sustainable returns is far from clear. Bankers and economists warn that the AI industry looks increasingly like a bubble that’s fit to burst.
Horror stories: Construction of AI data centers is propping up the economy and AI trading is propping up the stock market in ways that parallel prior tech bubbles such as the dot-com boom of the late 1990s. If bubbles are marked by a steady rise in asset prices driven by rampant speculation, this moment fits the bill.
- The S&P 500 index of the 500 largest public companies in the U.S. might as well be called the AI 5. A handful of tech stocks account for 75 percent of the index’s returns since ChatGPT’s launch in 2022, according to the investment bank UBS. Nvidia alone is worth 8 percent of the index (although, to be fair, that company posted a whopping $46.7 billion in revenue last quarter). “The risk of a sharp market correction has increased,” the Bank of England warned this month.
- In September, OpenAI outlined a plan to build data centers around the world that is estimated to cost $1 trillion. The company, which has yet to turn a profit, intends to build several giant data centers in the U.S. and satellites in Argentina, India, Norway, the United Arab Emirates, and the United Kingdom. To finance these plans, OpenAI and others are using complex financial instruments that may create risks that are hard to foresee — yet the pressure to keep investing is on. Google CEO Sundar Pichai spoke for many AI executives when, during a call with investors last year, he said, “The risk of underinvesting is dramatically greater than the risk of overinvesting.”
- Getting a return on such investments will require an estimated $2 trillion in annual AI revenue by 2030, according to consultants at Bain & Co. That’s greater than the combined 2024 revenue of Amazon, Apple, Alphabet, Microsoft, Meta and Nvidia. Speaking earlier this year at an event with Meta CEO Mark Zuckerberg, Microsoft CEO Sataya Nadella noted that productivity gains from electrification took 50 years to materialize. Zuckerberg replied, “Well, we’re all investing as if it’s not going to take 50 years, so I hope it doesn’t take 50 years.”
- AI companies are both supplying and investing in each other, a pattern that has drawn comparisons to the dot-com era, when telecom companies loaned money to customers so they could buy equipment. Nvidia invested $100 billion in OpenAI and promised to supply chips for OpenAI’s data-center buildout. OpenAI meanwhile took a 10 percent stake in AMD and promised to pack data centers with its chips. Some observers argue that such deals look like mutual subsidies. “The AI industry is now buying its own revenue in circular fashion,” said Doug Kass, who runs a hedge fund called Seabreeze Partners.
How scared should you be: When it comes to technology, investment bubbles are more common than not. A study of 51 tech innovations in the 19th and 20th centuries found that 37 had led to bubbles. Most have not been calamitous, but they do bring economic hardship on the way to financial rewards. It often takes years or decades before major new technologies find profitable uses and businesses adapt. Many early players fall by the wayside, but a few others become extraordinarily profitable.
Facing the fear: If an AI bubble were to inflate and then burst, how widespread would the pain be? A major stock-market correction would be difficult for many people, given that Americans hold around 30 percent of their wealth in stocks. It’s likely that the salaries of AI developers also would take a hit. However, a systemic failure that spreads across the economy may be less likely than in prior bubbles. AI is an industrial phenomenon, not based on finance and banking, Amazon founder Jeff Bezos recently observed. “It could even be good, because when the dust settles and you see who are the winners, society benefits from those inventions,” he said. AI may well follow a pattern similar to the dot-com bust. It wiped out Pets.com and many day traders, and only then did the internet blossom.
No comments:
Post a Comment