Al Jazeera warns AI data-center boom may fuel economic bubble risks

This essay captures a growing unease: the AI infrastructure boom now looks big enough to matter macroeconomically, not just within tech. The author walks through satellite imagery of sprawling US data centers, Nvidia’s brief stint as a $5 trillion company, and estimates that AI-related capex accounts for the overwhelming majority of recent US GDP growth. The argument is that we may be replaying an old pattern—canals, railways, fiber, crypto—where genuine technological revolutions are financed in a way that creates systemic financial risk.

For the race to AGI, this is a reminder that “more compute at any cost” is not a free option. If hyperscalers and AI labs overbuild capacity based on optimistic assumptions about AGI-driven returns, we could see a painful correction that chills capital for frontier R&D just as systems approach transformative capability. On the other hand, the piece concedes that if AI systems do start generating commensurate economic value, the same infrastructure could entrench a small set of firms with overwhelming power. Either way—overbuild and bust, or explosive success—the stakes transcend model benchmarks and speak to financial stability and industrial policy.