Datacentres are a ticking time bomb. We must make sure AI’s benefits outweigh the costs | Nicki Hutley
TL;DR
Nicki Hutley’s Guardian opinion piece ties the AI boom to the datacentre surge: these sites consume large amounts of power and water, emit heat, and are often framed by politicians as neutral infrastructure. The article says there are more than 10,000 active datacentres worldwide, with numbers expected to grow 3.5-fold. The projected investment bill is about US$7tn.
Nauti's Take
This is not an anti-AI piece; it is a overdue challenge to the infrastructure romance around AI. Anyone selling datacentres as a future project needs to show who benefits, who pays and what concrete local value is created.
Productivity hopes are not enough if power prices, water stress and fossil generation rise at the same time. The better line: speed up AI where it measurably helps, and slow down datacentres where they mainly turn platform power into concrete.
Briefingshow
The piece hits a weak spot in the AI economy: datacentres are built locally, while much of the value often flows to global platforms. When power, water, land and cooling capacity become scarce, households, municipalities and grids carry part of the bill. AI infrastructure therefore needs the same public scrutiny as other major projects, including emissions, jobs, import dependence and local benefit.
Sources