AI's environmental costs arrive as a thousand cuts, not one crisis

The distributed nature of AI infrastructure—spread across data centers, supply chains, and regional power grids—makes its environmental footprint harder to regulate or reduce than a single point-source pollutant. Instead of fighting one villain, governments and companies face coordination challenges across fragmented systems where small inefficiencies compound into massive aggregate damage, creating perverse incentives to defer responsibility to the next actor in the chain. Environmental costs won't trigger the unified political response that a single dramatic crisis might. Incremental degradation is the actual threat.