Is Generative AI Bad for the Climate?

Generative AI is powerful, but it consumes a lot of electricity. This has raised concerns about how it will impact the environment. Some technical leaders do not worry much, but companies that overlook excellence need to seriously think about this.

Power-Hungry Technology

Training and operating AI models requires significant computing power. This results in higher electricity consumption, which could lead to additional CO₂ emissions, especially when derived from fossil fuels. While Indian tech companies are investing in green energy, the sun and wind are not always reliable, and there is a concern about the use of batteries.

Training vs. Using AI

At first, training AI models was seen as the main energy drain. But now, using AI daily also takes a lot of power, especially with millions of people making queries. So, the environmental cost of just running AI is now a big issue too.

Finding the Carbon Footprint

Large companies do not share much data about the energy usage of their AI. But if you’re using open-source AI, there are tools like CodeCarbon and benchmarks like AI Energy Score to help estimate the footprint. Even for closed systems, estimates can be made using calculators for transparent study.

Final Thought

Generative AI is exciting, but we must balance progress with care for the environment. As developers and companies, it makes sense to look at how much energy our tools consume. Today’s small changes can help avoid big problems tomorrow.