Neil Rimer, the Index Ventures partner who has backed some of Europe and America's defining technology companies, is making a prediction that will comfort skeptics and unsettle froth-chasers: the money pouring into AI is about to start flowing back out. Rimer has argued that the current capital cycle around frontier models and AI infrastructure has outrun its fundamentals, and that history suggests capital eventually rotates out of overheated layers and back toward places where value is actually captured.
Rimer's framing follows a familiar pattern in general-purpose technology waves. Railroads, telecom and the early internet all attracted oceans of capital that built genuine infrastructure while incinerating enormous wealth in the process. The money came back out when the market repriced what each layer of the stack was truly worth. His view is that AI is now in the phase where the gap between narrative and revenue is widest, and therefore where the correction risk is greatest.
The comment carries particular weight because Rimer is neither an AI doomer nor an outsider to the boom; Index has been one of the most successful investors in the application layer, and his firm has profited handsomely from the cycle even as he questions its excesses.
For founders, the implication is practical: raise on conviction, but build as if repricing is coming. For limited partners, it is a reminder that the most expensive moments of a technology wave are usually the least profitable. Whether Rimer is early or right, the direction of his warning marks how far sentiment has shifted in a year once defined by unconditional exuberance.
