Aditi Sajwan
9 out of 10 internet users in India use AI in some way.
AI models rely on large data centres that consume significant energy and water.
Training an AI model like GPT-3 can consume up to 700,000 litres of water, equivalent to the annual usage of a few households.
AI relies on water to cool its servers and generate the power it needs... but at what cost?
Among the 15 countries with the highest number of data centres– India, Italy, and Turkey –face high to extremely high water stress
Companies can reduce AI's water footprint by optimizing cooling, using renewable energy, and designing sustainable data centers.