In today’s world, people are increasingly using “AI” (Artificial Intelligence) for both work and learning, widely embracing it to enhance daily life convenience. However, the use of AI has led to a significant release of carbon dioxide (CO2), contributing to the worsening of “global warming” — an issue that should not be overlooked.
“ChatGPT,” a popular AI tool with over 164 million monthly users, is known for its capabilities and intelligence, making it widely used. However, the latest data from KnownHost reveals that “ChatGPT” emits nearly 300,000 kilograms of CO2 equivalent (KgCO2eq) per month, or approximately 3 million kilograms (KgCO2eq) per year. This adds to the intensifying global warming problem.
Every time a user views the ChatGPT page, it releases about 1.59 grams of CO2. While this is a minimal amount, it is offset by the large volume of users. Research shows that the CO2 released by ChatGPT is equivalent to the emissions of 260 flights from New York City to London.
How Does ChatGPT Release CO2?
“AI” models like ChatGPT require high-performance computer resources and need to operate continuously for long periods. Processing large volumes of data requires vast amounts of electricity. In addition to the electricity used for processing, the storage of massive amounts of data also consumes significant energy. Thus, the CO2 emissions increase in proportion to the electricity used. In simple terms: “The more users there are, the more data needs to be processed, which increases the electricity consumption, resulting in more CO2 emissions.”
How to Minimize the Environmental Impact of Using “AI”?
The primary reason for CO2 emissions from using AI is the high energy consumption. The easiest solution to reduce this impact is to “use renewable energy” instead of electricity from fossil fuels. Additionally, some AI tools can reduce energy consumption by writing more efficient code, and managing data centers efficiently also helps save energy.
Reference