That’s what a prominent tech critic believesLast year, Big Tech’s AI capital expenditure touched $400bn. This year, it’s likely to top $700bn. How long can this cash burn last without a revenue spurt? NYU professor Gary Marcus has described the unending AI buildout as the “greatest capital misallocation in history”.Tech critic Cory Doctorow agrees. His book Enshittification , which explores why user experience on online platforms sucks, was a hit last year. Now, he’s back with another book, titled The Reverse Centaur’s Guide to Life After AI .What’s a reverse centaur, and what does Doctorow mean by a post-AI life? Surely, the text and image generators we’ve become so used to aren’t about to vanish? Who’ll write the courtordered 500-word essays for juvenile delinquents then?Centaurs were mythic creatures with a horse’s lower body, and a man’s torso and head.But in automation theory, abranch of computer science, a centaur is a human assisted by a machine. That means every computer user. A reverse centaur is the opposite: a machine’s human assistant.AI, as we know it, wouldn’t have been possible without armies of such reverse centaurs – low-paid workers who annotated images and text to train AI. And there are many other kinds of reverse centaurs as well. In fact, all employers would love to get rid of their human staff, retaining just a handful of reverse centaurs who can help AI do bulk of the work.And that, per Doctorow, is the crux of the whole AI “bubble”. We know that AI firms have been losing money for years, and there’s currently no business model that can recoup their rapidly increasing losses. So, why do mind investors continue pouring money into AI?

Doctorow says investors hope “The AI they are funding will be so good that employers can fire half their workers and replace them with AI, with the proceeds split between AI companies’ customers, and the AI companies themselves.”That’s why, AI is not an ordinary race with a finish line. Rather, it’s a race to death for major firms. In the end, only one player might survive, to become a monopolist. At least, that’s what investors hope, says Doctorow. Which firm will it be? There are as many hopes as there are investors, because each investor believes they have made the right bet.To keep investors convinced, firms must show them more and more powerful AI models. Hence, the relentless buildout with its frightening energy and water demands, “endangering the planet and the lives of millions of people to make several redundant AI tools that are functionally indistinguishable.”Doctorow does not believe in the AI game because of its flawed economics. Before the dot-com bubble burst, he says, the unit cost of adding users, and the cost of each web session, had been falling. But with AI, “Each generation of AI foundation models has been vastly more expensive to train and operate…AI gets more expensive when it adds a user or a feature.”So, the bubble is bound to burst, and considering how much weight AI stocks have in markets, it will hurt ordinary people, pensioners, etc. But after that, we will “get down to assessing what these applied statistics tools can do for us”.Doctorow isn’t saying AI tools are useless, but that their best and viable uses are yet to be found.