Gartner, a US firm of technology analysts, is famous for its “hype cycle”, a representation of the tendency for exciting innovations to experience a “peak of inflated expectations” (a bubble), followed by a “trough of disillusionment” (a crash).
Eventually, according to the oft-cited Gartner graph, of which it produces more than 90 variants each year in different tech domains, this is usually followed by a “slope of enlightenment” and finally a “plateau of productivity” as a technology matures.
The entire cycle can take years to play out, or it can be truncated.
The dot-com bubble in the late 1990s and its subsequent bursting soon after the turn of the millennium was a classic example of a huge hype cycle. The stock market — particularly the tech-heavy Nasdaq — boomed, with start-ups flooding into the market to take advantage of an investor frenzy. EToys.com, Pets.com and many other e-commerce companies like them briefly shone as brightly as the sun before they burnt up.
They failed not because their ideas were bad — the world always was going to move to online shopping — but because they were at least a decade ahead of their time. Their shares were priced for perfection, which never came. Yet today, e-commerce is integral to modern life. I, for one, rarely go into a physical store any more. From groceries to pet food to office stationery and electronics, I buy almost everything online now.
It’s uncanny how accurate Gartner’s hype cycle often is. And the generative AI boom — which OpenAI kicked off with the public release of ChatGPT in November 2022, thereby fuelling hundreds of billions of dollars in investment across the IT industry — appears to be following the curve closely. That’s a worry. Indeed, Gartner warns that we have already passed the “peak of inflated expectations” when it comes to generative AI.
Today the world’s 20 biggest tech firms are worth in the region of $20-trillion, making up nearly a fifth of stock market value globally
Could pioneering AI start-ups (and the venture capital money that’s flowed into them) soon be washed away like their dot-com equivalents of a quarter century ago? And could this drag down the valuations of some of the biggest companies in the world, such as Nvidia, Microsoft, Google, Meta Platforms and Apple, potentially leading to another global stock market crash? It’s impossible to predict the timing of the next crash with any certainty.
Since the last market crash in 2008, the value created by big US tech companies has been nothing short of astonishing: today the world’s 20 biggest tech firms are worth in the region of $20-trillion, making up nearly a fifth of stock market value globally. In rand terms, the numbers become silly: Microsoft, a leader in generative AI through its investment in OpenAI, is worth R57-trillion. All the companies listed on the JSE, by way of comparison — and that includes Naspers, with its exposure to China’s Tencent — have a combined market valuation of “just” R6-trillion.
And then there’s Nvidia, which has become a poster child for the AI bubble. Earlier this year, Nvidia briefly topped Apple and Microsoft to become the world’s most valuable company. It’s in third place now, with a market cap of $2.9-trillion.
For now, Nvidia is still struggling to meet demand for its graphics processing units — chips traditionally used in computer graphics but now retooled for generative AI. Tech giants such as Microsoft and Google continue buying these in boatloads. The big “hyperscalers” have already invested more than $100bn in AI data centres and related physical infrastructure since ChatGPT’s 2022 release. Will they see a return on this investment any time soon?
Like the dot-com bubble, it’s likely that AI adoption will take longer than many pundits think. A trough of disillusionment and a sharp rerating in tech stocks may be inevitable. In the longer term, there’s little doubt that generative AI, and advances in machine learning and robotics, will have a transformative impact on the world of work, and will deliver a step-change improvement in productivity. But like all new technologies, this will take time.
According to Gartner, generative AI is yet to deliver on its expected business value. The technology has already passed the “peak of inflated expectations”, though hype about it continues, the firm said in a report published in June.
Is a market bust inevitable? Though shares are expensive, the Nasdaq isn’t nearly in the same bubble territory as it was in the late 1990s. But one big quarterly earnings miss from Nvidia or Microsoft could trigger a huge sell-off in AI-fuelled stocks.
Buckle up! There could be turbulence ahead.
McLeod is editor of TechCentral
Would you like to comment on this article?
Sign up (it's quick and free) or sign in now.
Please read our Comment Policy before commenting.