Artificial Intelligence Is Consuming More and More Energy

Every time we ask a virtual assistant to write a text, generate an image or analyse a document, thousands of servers start working somewhere in the world. It is an invisible process, but far from immaterial, as artificial intelligence requires an ever-increasing amount of energy to power increasingly large and powerful data centres.

There is no doubt that AI offers tremendous opportunities. At the same time, however, it brings an unavoidable consequence: as the demand for computing power grows, so does the demand for electricity.

This is already shaping the energy strategies of many countries, accelerating investments in power generation, electricity grids and digital infrastructure.

For years, energy efficiency has mainly been seen as a way to reduce costs. Today, the issue has become much broader: using energy intelligently means helping to build a more sustainable system capable of meeting steadily growing demand.

In this context, buildings can no longer focus solely on consuming less energy; they must learn to consume it more intelligently. Knowing what is using energy, when, where and why has become strategic information. Equally important is the ability to automatically adapt building systems to users' actual needs, avoiding waste that until a few years ago might have seemed insignificant.

The paradox is clear: while artificial intelligence promises to make businesses and processes more efficient, it also risks increasing overall energy demand. This is why the challenge also concerns those who design, manage and operate the buildings in which we live and work. Will we be able to use the energy we already have in a smarter way?