AI is everywhere—from chatbots answering your queries to curated playlists on streaming apps 🎧. But have you ever paused to think about the massive behind-the-scenes machinery powering these services? Data centers, where AI models live and breathe, are placing big demands on our planet. Let’s break it down.
1. Powering up: A massive electricity appetite ⚡️
So far in 2025, global data centers have gulped roughly 200 terawatt-hours (TWh) of electricity—enough juice to keep a medium-sized country going for a month. The more AI goes mainstream (think real-time translations, smart assistants, image generators), the more energy these centers slurp up. In South Asia and Southeast Asia, countries like India and Indonesia are seeing a surge in hyper-scale facilities to support booming digital economies.
2. Water woes: The cooling challenge 💧
Keeping servers cool is critical. Traditional air cooling can require millions of liters of water daily. Imagine cooling that half a football field–sized warehouse—you get the idea. In water-stressed regions such as northern India or parts of the Philippines, this can add pressure to local supplies.
3. Carbon count: Emissions on the rise 🌍
When the power mix includes coal or gas, each megawatt-hour burned can pump hundreds of kilograms of CO₂ into the atmosphere. Data centers are estimated to produce over 200 million metric tons of CO₂ annually—comparable to the aviation sector.
4. Green turns: Solutions in the pipeline 🌱
Good news though! Cloud providers and startups are pushing green initiatives: some data centers in Tamil Nadu and Singapore are already operating on 100% renewable power. Liquid cooling tech, which uses special non-conductive fluids, can cut water use by up to 80%. And AI itself is being used to optimise energy loads, shutting down idle servers or shifting heavy computing to off-peak hours.
As young pros and students in South Asia and Southeast Asia, we live in a streaming, gaming and AI-driven world. Next time you fire up your favourite app, remember there’s a planet behind the pixels—and we can all advocate for a greener AI future. 💪
Reference(s):
cgtn.com
