Welcome to linking:debate where we share perspectives that shape progress. In this episode, The Green Illusion: who's counting technology's footprint, the discussion focuses on how AI and cloud computing have shifted digital technology from an apparently immaterial service into a form of heavy industry.

The speakers outline how AI differs from earlier digital platforms because each use requires substantial computation, energy and water. This removes the assumption of near zero marginal cost and drives continual expansion of data centres and specialised chips.

They compare current AI energy demand with established industrial sectors, noting that AI and associated data centres already account for several percent of global energy use and are growing faster than any other sector.

The conversation examines the production of GPUs and semiconductors, including fabrication plants, material extraction and toxic chemical use. Short hardware lifecycles and weak recycling systems contribute to rising e-waste and further resource extraction.

Attention then turns to data centre construction, highlighting the carbon intensity of concrete, steel and copper alongside local impacts on electricity grids and communities.

The speakers describe how technology companies use carbon neutrality claims, offsets and power purchase agreements to present growth as sustainable. These mechanisms are shown to shift rather than reduce emissions and can slow wider grid decarbonisation.

Claims that AI will solve climate change are challenged. Examples include efficiency gains, climate modelling and speculative futures such as fusion energy which remain uncertain and distant.

The discussion closes by considering responses at different levels, including regulation, protest, repair cultures, longer device use and more intentional limits on digital consumption as ways to reduce impact and counter greenwashing narratives.

Speakers:

Alistair Alexander is a researcher on the social and ecological impact of technology. Recent projects include researching regenerative futures for AI and digital infrastructures with Bath Spa University, teaching a seminar on Ecologies of technology at the University of Europe for Applied Sciences and creating the Connection Matters display banners for the exhibition Invisible Networks in Berlin. He engages diverse groups with work on the Doughnut Economic Model for Tech and how social networks could be more like funghi networks.

Benjamin Johnson finished his PhD in physics in 2010 at the Technische Universität Berlin with a thesis on thin layer solar cells. He later worked on catalytic materials for alternative fuels. In 2015 he began research in science history, studying technological progress, which led to his book Making Ammonia (Springer, 2022). He now works to increase public understanding of the energy transition by combining natural sciences with history, policy and civil society perspectives.

Introduction by Deborah Causton

Music 'Good Balance' by AO