12.11.2025, Moscow.
Many companies have started rehiring employees they had previously laid off and replaced with artificial intelligence (AI) technologies. That is a clear sign of growing disappointment in digital assistants, Rossa Primavera News Agency‘s IT Desk wrote.
On November 4, the analytics firm Visier, after analyzing employment data for 2.4 million workers, reported that about 5.3% of them had returned to their former jobs. Although this share is small, it has been steadily increasing. According to researchers, employers have encountered a gap between their expectations of technology and the actual performance of AI systems.
It turns out that the technological hype has outpaced the real capabilities of the technology. Developers of neural networks skillfully promoted their products, but reality is asserting itself. The issue is not that digital assistants are useless, but that their integration has proven more expensive and failed to replace many of the functions performed by humans.
As far back as 1995, Gartner introduced the concept of the “Hype Cycle,” describing how new technologies impact markets. According to this model, after reaching the “peak of inflated expectations,” technologies enter the “trough of disillusionment,” followed by gradual recovery and reaching the “plateau of productivity,” which remains below the initial peak level.
For some time, this trial-and-error period will continue, as neural networks are applied across different fields. In some areas, they will perform well; in others, human correction will be constantly required; and in some cases, they will offer no real benefit at all.
However, excessive trust in AI technology could lead to an economic crisis. Too much capital has been invested in building infrastructure for neural networks. Fueled by hype, IT companies have been receiving investments and cheap loans, pouring hundreds of billions of dollars into the field.
Many investors expect the emergence of “true” AI, but it is already clear that current technology cannot deliver it: each new model requires more investment, while the improvements become less significant. When the euphoria ends, the “hangover” will be severe.
The recent $1 trillion decline in the stock value of major US tech giants over the past week is a striking sign that the peak of the AI frenzy is nearing its end.
Source: Rossa Primavera News Agency
