Dotcom all over again ?

Today, there is a lot of speculation in the news (Wed, Nov 19 2025) about an imminent AI stock market crash.

Google CEO Sundar Pichai, Yesterday, in an interview with BBC News at Google’s California headquarters, acknowledged the current moment as “extraordinary” but noted that elements of “irrationality” had crept into what he described as an otherwise rational boom.

The sentiment is understanding. Early in the book, I speak about the dot-com bubble of the 1990s, and although we are decades later, the same psychology is at play here. Investors don’t want to miss out; they can all sense that we are on the verge of something significant, and that warrants being on the front foot when capitalising on the forthcoming profitable worldwide opportunity across all sectors and markets.

The anxiety in the markets is well-founded. We are in a moment of excitement, and cash outflows to seemingly seed-stage startups are in the 10s of millions of dollars, which is unheard of. The right teams have got investors excited, but like the dot-com boom, will all these start-ups deliver tangible, profitable businesses in the future? No, not all of them will. And several factors are at play here.

In the book, I discuss the convergence of technologies that enable breakthroughs. A startup with a good idea and a good team may well find itself redundant within a year if some other company comes up with a more compelling offering in this fast-moving space.

LLMs themselves are not immune to this; innovation is driving how they work, focusing on efficiency and a better way of working, as demonstrated by China’s Deep-Seek Model. Markets are asking: Will we need all these NVIDIA chips?

Add to this the way the fundamentals of what is being invested in can change; recently, LeCun, one of the most respected leaders in this space, left Meta to launch a new startup, which may call into question the value of the current LLM architecture. This is part of the infrastructure that companies like Meta, OpenAI, Google and others have poured billions into, and it is the backbone of the current boom.

So, will there be a market correction? Yes, most likely.

Will it affect the long-term view? No, it won't.

We are just building the infrastructure for Synthetic intelligence, and, as with the internet boom of the 90s, the AI equivalents of Amazon and Apple that provide actual utility to customers and will be profitable businesses are yet to emerge.