Environmental Cost of AI

Definition

The comprehensive ecological footprint of machine learning systems spanning silicon manufacturing, energy consumption, carbon emissions, and e-waste.

Key Characteristics

• Embodied carbon in specialized accelerator hardware.
• Grid carbon intensity dynamics during training and inference.
• Resource extraction in global supply chains.

Applications

• Environmental humanities
• Ethical AI frameworks

Mentions in Source

• “Situates AI resource extraction within broader infrastructure theory.” — karen-hao-empire-of-ai-water-footprint