Deutsche Bank notes that a needle is dangerously close to the AI ​​bubble

The money seems to be constantly circulating in AI space, whether Nvidia announcing an investment of $ 100 billion in Openai or Openai to build more massive data centers before the first $ 500 billion project is even completed. Finally, some of the people who have paid money into these companies will expect to see a return on investment. According to a research note for Deutsche Bank, it becomes more and more difficult to see how it will happen.

Fortune reports that a note written by George Saravelos de Deutsche Bank warned that spending in the AI ​​sector was “parabolic”. In fact, it is so vast, said the researcher, that he could alone support the American economy. “The AI ​​machines – in a fairly literal sense – seem to save the American economy at the moment,” he wrote. “In the absence of spending related to technology, the United States would be close or in the recession this year.” This checks: earlier this year, the Wall Street Journal said that capital expenses for AI have contributed more to the growth of the US economy than all consumer spending combined this year.

If you want to reduce it even more, you can. Saravelos underlined the manufacturer of Nvidia fleas specifically and says that the company “currently carries the weight of American economic growth”. All that must happen so that the company continues to exhaust the entire economy on the back is that its growth is constantly infonvial. No problem, right? “The bad news is that for the technological cycle to continue to contribute to GDP growth, capital investment must remain parabolic. It is very unlikely,” said Saravelos.

To pull.

Now you don’t really need to be an economist to know that it’s usually a bad idea to have all your eggs in a single basket. But just in case, here is one to tell us: Torsten Sløk, chief economist of the Apollo asset management company, wrote last week that “the essential is again that there is an extreme level of concentration in the S&P 500, and investors in equity are considerably overexposed to AI.”

To put figures there, here are some mathematics of a recent report published by the Bain & Company consulting company. The company says that “IAcalculation The demand grows More than double the rate of Moore’s law “, and by 2030, the computing power required to meet the demand for AI would have a cost of two dollars per year.”The The world is always $ 800 billion short has keep Rhythm with demand, ”writes Bain & Co.

Combine this with the report that came out of MIT earlier this year, which noted that only 5% of companies that adopted generating AI tools have managed to achieve “rapid income acceleration”, while the others fell flat, and it becomes a little difficult to see the case for profitability that takes place right at the corner of the street. But hey, who knows, maybe we are only a different small investment of $ 500 billion far from clicking everything. Takers?