Scotiabank Market Points Powering intelligence: AI’s energy footprint
Nov 3, 2025
In this engaging discussion, Alex de Vries-Gao, Founder of Digiconomist and researcher on digital sustainability, delves into the escalating energy demands of AI. He highlights the complexity of quantifying AI's power consumption and the gaps in corporate transparency that hinder accurate measurements. Alex notes that AI's energy usage could rival that of the Netherlands, impacting electricity demand and environmental policies. He warns about the hidden costs of unchecked growth and advocates for responsible innovation amidst rising emissions.
AI Snips
Chapters
Transcript
Episode notes
Researcher Journey From Crypto To AI Sustainability
- Alex De Vries began DigiConomist in 2014 focusing on cryptocurrencies and shifted to sustainability of emerging tech.
- His work expanded into academic papers and a PhD on tech sustainability.
Corporate Opacity Around AI Power Use
- Big tech knows AI energy use but rarely discloses it, creating opacity that hinders analysis.
- Alex De Vries reconstructed demand bottom-up because companies refuse to publish AI-specific power figures.
Bottom-Up Reveal Shows Rapid AI Growth
- Bottom-up triangulation using chip makers and packaging capacity reveals large, measurable AI electricity demand.
- De Vries estimates AI already consumed ~20% of global data-center power and may approach half this year.