Monocle Editor Questions AI Doom Warnings, Warns of Bubble Burst

In a Monocle opinion piece, editor Josh Fehnert argues that the existential warnings about artificial intelligence from tech CEOs have a credibility problem because many of those issuing the alarms are simultaneously building and selling the technology. He points to Anthropic's leaked preparations to go public, reported the previous day, as an example of how such warnings can serve commercial interests. Fehnert notes that the companies calling for regulation are also positioned to sell the solutions. He compares the current moment to pre-crash 2007 and the final months of the dot-com bubble, suggesting the realistic short-term danger is not runaway robots but AI companies overpromising. While acknowledging genuine risks such as cyberattacks on critical infrastructure and rogue bots harming the economy or healthcare, he argues the industry needs to move beyond sci-fi tropes of global destruction and articulate concrete benefits. He questions the goal of the AI race—whether it is US geopolitical advantage, investor profit, or societal benefit—and observes that large language models remain clumsy, error-prone, and of dubious utility. If the gamble fails, he warns, confidence will wane, the AI bubble will burst, and heavily indebted governments and shaky economies will pay the price. The piece ends with a note that Fehnert is Monocle's editor and a subscription prompt.