Field Notes · Data Center Health
Water, Power, and Noise: Measuring the Real Public Health Footprint of AI Infrastructure
The health effects of AI infrastructure are not only about emissions. Water, power, noise, and heat all reach the people who live nearby.
Each new AI model runs on physical buildings full of servers. Those buildings draw water, power, and land, and the effects reach far past the fence line.
Water
Cooling can use large amounts of water, often during heat waves when supplies are tight. Communities can ask for daily peak use, the source, and plans for drought.
Power
New demand can extend the life of older power plants and add new gas plants. That affects local air quality and can raise household bills.
Noise
Cooling fans and transformers can create a steady, low hum that carries at night. Ongoing noise is linked to poor sleep, stress, and higher blood pressure.
Heat and Light
Large roofs and paved lots add heat to nearby areas. All-night lighting can disturb sleep for neighbors and wildlife.
What to Measure
- Baseline air, noise, and water data before construction
- Ongoing public monitoring after the site opens
- Health data for nearby neighborhoods over time, including asthma and sleep problems
Dr. Allison Alimayu holds a PhD in biomedical informatics. She is a former elected official and a public interest futurist, and she leads Alimayu Studio.