Has the AI stock market bubble apocalypse been averted or merely delayed? In recent days concerns have grown. About a third of the S&P 500’s market value sits in the “Magnificent Seven” — Alphabet, Amazon, Apple, Meta, Microsoft, Nvidia, and Tesla — and their AI splurge has done most of the work behind the index’s 15% gain this year.

But global markets have rallied, following an unexpectedly good earnings report on Wednesday from Nvidia, the world’s leading designer of AI-related hardware and software.

According to Carlota Perez, author of Technological Revolutions and Financial Capital: The Dynamics of Bubbles and Golden Ages (2003), stock market bubbles have regularly appeared during the development and diffusion of new technologies like railroads, automobiles, and computers. The bad news is that the restructuring of the economy over several decades can lead to a “frenzy phase” in which valuations are decoupled from fundamentals. Over time — often only after a painful downturn — the widespread deployment of the new technology lifts productivity and ushers in a “golden age” of growth.

If this historical pattern holds, there is good news and bad news. The bad news, for those in search of stability, is that with AI we are likely at the end of the beginning rather than the beginning of the end. Whether the benefits of AI are overhyped or not may not be known for decades. In the meantime, booms and busts linked to AI investment may be common for decades, as over-enthusiastic investors periodically get cold feet.

But there is good news as well. To begin with, the stock market casino and the “real” economy are two different things. Although more than 60% of Americans own some stocks, the majority of stock ownership is highly concentrated. During the dotcom bubble of 2000-2003, the crash contributed to a recession among other factors, like 9/11. But the market bounced back within a few years; a handful of wealthy investors took the hit, and the true tech giants — Google, Apple and others — survived the winnowing and grew stronger, unlike Pets.com.

The most devastating stock market crashes, like those that triggered the Great Depression of the Thirties and the Great Recession of the 2010s, are rare and tend to be caused by widespread banking crises and the collapse of real estate bubbles, like the sub-prime mortgage crisis of 2008. During and after the Covid pandemic, there were unsustainable housing bubbles in some desirable locations in the US to which buyers and renters fled. Today home mortgage and rental prices in a number of former hotspots like Miami and Austin are softening, suggesting a gradual decline to sustainable levels rather than a devastating crash.

And today’s AI company valuations may be more sustainable than those of tech firms in the dotcom bubble era. According to Goldman Sachs, the median price-to-earnings ratio is half that of the seven largest companies preceding the dotcom crash.

The trip to the AI-dominated future in the metaphorical robocar may be a bumpy ride. But for now, fear of an AI stock apocalypse may be as unrealistic as the alarm about a robot jobs apocalypse or the enslavement of humanity by a Silicon God.