It’s a bumpy time for the hyperscalers and the tech companies involved with AI and its build-out. Intel, for example, has experienced its worst drop in share price since 1974. Its lower forecast for laptop and desktop shipments for 2027, due to chip demand for AI data centres driving up the prices of chips for laptops and desktops, didn’t go down well with investors. Alphabet, Google’s parent, also suffered its worst share price decline in over two decades due to news that its AI search mode brings in less advertising money than a classic search, and also that its free cash flow is negative – more cash is going out on operating activities and capital investment than is coming in from regular business. In fact, investors in the tech giants and companies associated with AI are increasingly nervous about whether the massive spending on it and its infrastructure will pay off. Things are volatile, and if, like the Badger, you have been monitoring the AI Data Centre index over time then you will have seen investor nervousness being reflected in share price oscillations.
The Badger’s been musing on the similarity of the AI boom to the 1980s PC and 1995-2001 dot.com booms. Many people, see here for example, assert that AI is different, and that although there will be bumps and corrections, a sudden collapse like that of the dot.com era is unlikely. They have a point. However, having been an IT professional during the PC and dot.com booms, the Badger remains suspicious of the current AI market boom. Objective analysis of indisputable metrics over years is always the best pointer to whether a significant AI market correction is imminent, and so at the end of last week, the Badger listened to a major financial institution’s webinar analysing what’s going on in markets. This included a chart showing all the contributions, quarter on quarter, to annualised US Gross Domestic Product (GDP) since the mid-1970s. Spikes caused by the PC and dot.com eras were plain to see, as was the current AI spike. The shape of all the spikes was similar, with the PC and dot.com peaks being immediately followed by a decline in US GDP. The AI spike’s peak appeared to be flattening, but no clear GDP decline is visible yet. The Badger felt the analysis supported the view that it would be a brave act to assert that a decline will not happen anytime soon.
By the time you read this, the US mid-term elections will be dominating headlines. The world is highly volatile at present, and if you worry about midterm election outcomes and whether the vast sums of money being invested in AI will actually pay off, then there’s one thing that you should really do. That, as Star Trek’s Captain Jean-Luc Picard will be saying from the heavens, is brace for impact…
