Artificial intelligence has enormous potential to boost productivity, but investors and businesses expect too rapid results from it. My analysis shows that the real economic impact of AI implementation will not become apparent for at least several years. Deutsche Bank's Head of Macroeconomic Research Jim Reed rightly notes that in his entire career, there has been no technology with such potential, yet companies will need time for full integration.

After the rapid growth of AI company stocks, a correction is inevitable — this is a normal market reaction. However, the technology's long-term impact will extend beyond cycles. The world will enter a period of adaptation, accompanied by high volatility and a reassessment of expectations. At the same time, fears of mass job losses due to LLMs are, in my view, exaggerated. Previous technological breakthroughs did not reduce overall employment levels, and AI will become a tool for enhancing labor efficiency, not a replacement for humans.

Investments are growing, returns are lagging

Notably, even with productivity growth, AI will not necessarily reduce inflation. As historical experience over 250-300 years shows, the adoption of new technologies improved efficiency but did not lead to sustained price declines. On the contrary, the development of AI may create additional inflationary risks.

Analysts at Deutsche Bank emphasize in their annual World Outlook review that the current pace of technology development will almost certainly lead to noticeable productivity growth, but real results will only become visible after 2026. The Wall Street consensus forecast for capital expenditures by the largest technology companies this year stands at $527 billion — higher than the $465 billion expected at the start of the third-quarter earnings season.

Снимок экрана — 2026-07-07 в 17.07.56

However, the scale of investment is currently outpacing measurable economic impact. Goldman Sachs estimates that AI spending supports investments in equipment and infrastructure, but its contribution to GDP growth remains limited. In 2026, it will add about 0.3 percentage points to real economic growth and only 0.1 percentage points to measurable growth.

My expert assessment: The market is overheated with expectations, and a correction is inevitable. But those who can weather this period of adaptation and wait for the real integration of AI into business processes will gain a significant competitive advantage. For now, we are witnessing a classic bubble of expectations that requires time to deflate and then grow based on real data.