Fylom
Back to Tech & AI
The energy and water cost of AI

Tech & AI

The energy and water cost of AI

11 min

Training and running AI models consumes staggering amounts of electricity and water. Explore how the data-center boom is straining power grids and local resources, and the scramble to power the AI era.

Listen on the app, request early access:

Show notes

Training GPT-three consumes ten gigawatt-hours of electricity, matching the annual usage of one thousand American households.

A single data center can consume five million gallons of water daily, rivaling a city of fifty thousand.

Data center power demand is projected to reach nine percent of the total grid by twenty thirty.

Cooling fans at data facilities reach eighty-five decibels and raise local temperatures by up to nine degrees.

Fifteen coal and gas plants have postponed their retirements to meet the rising energy demands of AI.

Eighty percent of potable water used for data center cooling is lost to evaporation.

In this episode

  1. 1Intro1 min
  2. 2The Physical Footprint of a Digital Prompt2 min
  3. 3Grid Strain and the Heat Dome Crisis3 min
  4. 4The Geography of Scarcity2 min
  5. 5The Scramble for New Power3 min
  6. 6Outro1 min

Sources

Fylom generates episodes like this on any topic you're curious about.

Fylom episodes are researched and written by AI. Automated checks help catch inaccuracies, but episodes aren't reviewed by a human and AI can still get things wrong. Treat them as a starting point, not a source of record — more in our accuracy disclaimer.

The energy and water cost of AI — Fylom