Illusion of infinite growth: Is the world in danger of collapsing in the artificial intelligence "bubble".
Just two years ago, the world was in ecstasy over ChatGPT and the promise of "overall economic transformation". Today the rhetoric is changing: in business publications and analytical reports, the word "bubble" sounds more confidently. Is the AI industry really on the verge of a massive collapse, or are we dealing with another wave of panic?
The skeptics point to some systemic problems. First, there is a huge gap between investments and real income. Sequoia Capital estimates that the industry has already invested more than a trillion dollars in artificial intelligence infrastructure, while the annual income from generative models barely exceeds tens of billions. Second, energy costs are increasing exponentially: data centers of large corporations consume electricity at the level of small countries, and this trend is only getting stronger. Third, more and more signs of "fatigue" from artificial intelligence products are appearing in the market: corporate customers are cutting budgets for pilot projects, and not seeing a quick return.
The AI debate is divided. New York University professor Gary Marcus says: "We are seeing a classic bubble, where expectations are orders of magnitude ahead of the true value of the technology. The correction is inevitable, and it will be painful." The founder of DeepMind, Demis Hassavis, partially agrees with him, noting that "some of the companies whose valuations were based solely on hype will indeed disappear from the market."
However, there is also an opposite position. Goldman Sachs analysts emphasized in the latest report that "AI is not speculation, but a fundamental change in productivity, and any comparison with dot-com is superficial." And Elon Musk, whose statements traditionally balance between euphoria and pessimism, noted that "there is overheating, but it concerns financial models and not the technology itself."
Sign up for SOBKOR.