We all know that AI has become the buzzword amongst Silicon Valley giants. The world’s biggest tech companies are spending billions to power the AI revolution. However, some experts are saying that this rapid build-out might be heading towards a bubble.
If we go by the recent Bloomberg report, five major tech players, including Microsoft, Google, Amazon, Meta, and Nvidia, are expected to spend around $371 billion. This year, on data centers to train and deploy large AI models. McKinsey projects that the figure could jump $5.2 trillion by 2030.
However, some investors and analysts are questioning whether AI will ever generate enough revenue to justify these costs. Bloomberg cites estimates from Bain & Co., which suggest that Big Tech would need about $2 trillion in new annual revenue by 2030 just to offset data center spending, a number that currently looks far out of reach.
Harris Kupperman, founder of Praetorian Capital, believes the current AI spending frenzy “is a bubble,” pointing out that even this year’s infrastructure build-out alone would require $480 billion in extra revenue to break even. Much of the technology, including GPUs that power AI systems, depreciates quickly, making long-term returns even harder to achieve.
Meanwhile, studies from MIT Media Lab and McKinsey show that most companies using AI tools are still struggling to see clear financial gains. About 95% of AI projects, according to MIT, have produced no measurable return so far.
On the other hand, there are AI optimists who argue that this is just the early phase of a long-term shift. And big tech executives like Sam Altman and Mark Zuckerberg continue to recall AI as the next platform shift. So for now, the debate remains open, and only time will tell who is going to win.
