Writing

Energy · 23 October 2024

AI as a Catalyst for Climate Solutions

Forecasting, grids, precision agriculture, CCS and the uncomfortable fact that AI’s own energy use must be governed, not waved away.

Rows of solar panels and a distant turbine under overcast light

AI is already forecasting wind, tightening grids, guiding precision agriculture, and helping carbon-capture plants run. Google’s wind-dispatch work and Copernicus-class climate models are not hypothetical. The same essay insists on the other ledger: training and inference emit, data centres already take a material share of electricity, and “green AI” is a design requirement, not a press line. Climate AI that cannot account for its own load is not a climate strategy.

First published on n-consultancy.blogspot.com. This article was first published on my research blog and is updated here.