Marco Zaccaria Di Fraia
Virtual Water
Virtual water measures the hidden water embedded in producing goods.
Developed by geographer Tony Allan to track global supply chains, it makes visible what agriculture and manufacturing obscure: the material costs of production. This image extends that framework to AI infrastructure. It shows a crop whose virtual water footprint roughly matches the water consumed to generate this digital image.
The irony is intentional. We call digital images "virtual" as if they're unembodied, but generating them requires real water in real data centres. The exact amount varies by model, infrastructure, and geography - but the parallel holds. The lettuce depicted looks edible but is only pixels - it will never decay, never feed soil, never participate in any cycle. A simulacrum trading material resources for immaterial representation.
The recursive structure - where the system defines the output - echoes conceptual art traditions (Sol LeWitt, Douglas Huebler). The tension inverts Baudrillard's critique: here the simulation reveals rather than obscures its physical infrastructure. Nothing virtual is immaterial.
Marco Zaccaria Di Fraia, PhD, is an engineer, researcher, writer, and founder working across AI, space systems, and sustainability. Based in London, he's worked on problems ranging from deep-space navigation for asteroid proximity operations to autonomous retail, citizen science, and generative AI for creative industries and intelligence analysis. He often speaks on data ethics, DataOps, frontier technology and the future of work.
Marco Zaccaria Di Fraia, PhD, is an engineer, researcher, writer, and founder working across AI, space systems, and sustainability. Based in London, he's worked on problems ranging from deep-space navigation for asteroid proximity operations to autonomous retail, citizen science, and generative AI for creative industries and intelligence analysis. He often speaks on data ethics, DataOps, frontier technology and the future of work.