Wall Street experienced a state of panic following a new report from the Massachusetts Institute of Technology (MIT) warning that most corporate investments in artificial intelligence yield no tangible returns. The report, issued by the Nanda AI project at the university, clarified that nearly 95% of institutional investments in generative AI technologies have not resulted in clear financial outcomes, despite spending between $30 and $40 billion.
This warning quickly reflected on the stock markets, where technology company shares fell sharply; Nvidia's shares – the giant valued at $4 trillion – decreased by 3.5%, while Palantir lost about 9% of its value. Japanese company SoftBank's shares fell by 7%, amid investors' concerns about potential massive losses due to its large bets on OpenAI.
The report revealed that half of corporate AI projects ended in failure, and 80% of companies experimented with the technology, but only 40% were able to use it effectively, while only 20% reached the testing stage, and just 5% made it to actual production. It also indicated that employees resort to consumer tools like ChatGPT at their own expense, while expensive corporate tools fail to deliver the promised value.
Even Sam Altman, the CEO of OpenAI, acknowledged in private statements that investors are "overly enthusiastic," warning that some may lose "huge amounts." This admission came after the launch of the ChatGPT-5 model, which disappointed many users who described its improvements as limited.
Despite these concerns, some analysts believe that the market is still on an upward trajectory. Dan Ives, an analyst at Wedbush Securities, asserted that "the skeptics will prove to be wrong once again," noting that the tech upcycle could last for at least two or three more years.
However, with Morgan Stanley's forecasts of $3 trillion in spending on data centers in the coming years to fuel the AI boom, the question remains: Are we facing a new tech bubble akin to the internet explosion at the turn of the millennium, or is the current correction merely a "temporary bump" on the road to an AI revolution?$ETH