The environmental cost of datacentres is rising. Is it time to quit AI?
TL;DR
Global datacenter power demand is growing four times faster than all other sectors combined, according to the International Energy Agency.
Key Points
- By 2030, datacenters could consume more electricity than Japan – a country of 125 million people.
- The 'QuitGPT' movement is gaining momentum, with users questioning whether AI boycotts can reduce environmental impact.
- The Guardian's sustainability column explores whether individual opt-outs from AI tools make a meaningful difference.
Nauti's Take
'QuitGPT' is an understandable reflex, but it is not a climate strategy. Celebrating personal AI abstinence while datacenters run on coal is self-deception at scale.
What the industry actually needs are mandatory transparency requirements for energy and water consumption per model query – not moral appeals to individual users. It would also be more honest if AI providers disclosed their environmental costs openly rather than burying them in glossy ESG reports.
Sources