These days everyone uses AI, but have you ever wondered what impact does AI leave on the environment?
According to the London School of Economics and Political Science, data centers currently contribute about 1% of global energy-related greenhouse gas emissions. “By 2035, increased data center energy use could lead to an additional 0.4–1.6 gigatons of CO2 equivalent (GtCO2e) emissions.”
It is really hard to understand how AI data centers manage to pollute the earth with C02, so LSE further notes that the improvement of artificial intelligence models requires a large amount of computational resources. This means that a lot of electricity is being used, just for improvement. There is more electricity being used for maintaining the AI.
How does electricity pollute the earth?
According to an article by kunak, emissions from the energy industry include: carbon dioxide, carbon monoxide, sulfur oxides, nitrogen oxides, and many more. This is because fossil, nuclear, and even renewable energy plants emit these chemicals. These chemicals, in turn, negatively impact global warming.
How to improve AI to be more sustainable?
There are not many ways you can contribute to solving AI’s pollution because the software is run by large corporations. A mass AI cutback is very unlikely, but it would contribute to solving the problem. Data centers would go bankrupt and be forced to close, thus solving the data center pollution problem.
Conclusion
To summarize, AI data centers cause a need for electricity, which in turn forces chemicals out, polluting our air.
