Datacentres are a ticking time bomb. We must make sure AI’s benefits outweigh the costs | Nicki Hutley
TL;DR
Nicki Hutley argues in the Guardian that the AI boom is colliding with the climate crisis through rapid datacentre expansion: more than 10,000 active sites worldwide, expected 3.5x growth and an estimated US$7tn investment wave. Australia is the case study: 286 active or planned datacentres, interest from AI players such as Anthropic, and governments treating the boom too casually as infrastructure.
Nauti's Take
Teams buying or building AI workflows should add infrastructure to the checklist: where the model runs, which data region applies, and how clearly providers report energy and water use. The source is a Guardian opinion piece, not a neutral market report.
The durable test is simple: AI value has to be weighed against operating costs and local infrastructure impact.
Briefingshow
This pulls AI out of the app layer and back into the physical world: power grids, water, heat, imported hardware and local jobs. That is where the real test sits: whether AI infrastructure creates broad public value or shifts private resource costs onto everyone else. Europe should care because the same questions are coming with cloud regions, sovereignty debates and energy prices.
Sources