It is reasonable to keep the energy use of AI systems under scrutiny, but so often the discussion of the topic feels overblown (AI drive brings Microsoft’s ‘green moonshot’ down to earth in west London, 29 June). The reason is that the emissions that are likely to be generated by powering AI are so much smaller than those created by other sectors such as construction or transport.
So the current debate is framed in the wrong way. AI is already proving to be a critical enabler in helping to tackle the climate and nature crises. And, in some cases, the power use for AI is much lower than the traditional computing that it replaces – for example, in energy-intensive advanced climate models.
Meanwhile, AI can unlock significant decarbonisation opportunities. In my own sector – the built environment – it can help enable the shift away from polluting materials such as concrete, reduce the carbon impact of energy and transport systems, and support the reuse and repurposing of existing buildings.
We need to keep improving the efficiency of AI development and ensure that it is developed safely and responsibly. But as important is harnessing AI to help accelerate the green transition.
Dr Will Cavendish
Global digital services leader, Arup