Environmental Cost of Generative AI
Tools like ChatGPT-4o make tasks like image creation fast and easy, but they consume huge amounts of energy.
Every AI interaction uses power from data centres, most of which still rely on fossil fuels.
AI data centres could use up to 10% of global electricity by 2030.
Training a single AI model can emit as much CO₂ as five cars over their lifetime.
India currently meets its AI energy needs, but rapid growth calls for urgent planning.
Need for Transparency
AI companies should disclose:
How much energy their tools use.
The source of that energy.
What steps they are taking to reduce energy use.
Transparency would help researchers find ways to make AI more energy-efficient.
Small Modular Reactors (SMRs) as a Solution
SMRs are compact, scalable nuclear reactors that can supply consistent, zero-carbon power.
Unlike large nuclear plants, SMRs are easier to build and can be placed near data centres.
They offer:
24/7 power supply (unlike solar/wind).
Faster construction and lower cost.
High safety with passive cooling systems.
Flexibility to operate in cities or remote areas.
Challenges and the Way Forward
SMRs need new safety regulations and public support.
High upfront costs and integration with existing renewable energy plans are concerns.
In India, SMR electricity cost could drop significantly after setup, making it cheaper than current electricity rates.
A public-private partnership is essential to develop SMRs alongside renewables and ensure AI growth remains sustainable.
COMMENTS