In recent years, artificial intelligence technology has developed rapidly, bringing many conveniences to people’s lives. However, there are huge energy consumption and environmental problems hidden behind this technology. A new study by KnownHost reveals the staggering carbon footprint of artificial intelligence tools. Taking ChatGPT as an example, its monthly carbon emissions are equivalent to the combined emissions of hundreds of transatlantic flights, which raises concerns about the sustainable development of AI technology. worries.
Recently, a new study by website hosting provider KnownHost revealed the staggering carbon footprint of artificial intelligence (AI) tools. Data shows that ChatGPT alone attracts more than 164 million users every month and generates up to 260,930 kilograms (about 260 tons) of carbon dioxide, equivalent to the carbon emissions of 260 New York to London flights. This highlights the huge impact that AI technology has on the environment while developing rapidly.
With the popularity of AI, the energy demand of data centers is also rising sharply. Research from Berkeley Lab shows that energy demand in U.S. data centers will more than double between 2017 and 2023, while a report from the International Energy Agency predicts that by next year, technological advances such as data centers, AI and digital currencies will use The amount of energy is likely to double again, with a third of that growth coming from the data centers themselves. This trend has prompted the Biden administration to propose solutions to data center energy consumption, calling for the use of "clean energy" to operate these facilities.
According to reports, ChatGPT consumes more than 500,000 kilowatt hours of electricity every day, which is equivalent to the daily electricity consumption of more than 17,000 American households. Data scientists from the Dutch National Bank predict that by 2027, the annual electricity consumption of the entire AI industry will reach 85 to 134 terawatt hours. Tesla CEO Elon Musk also warned that power shortages will become the main factor restricting the development of AI in the next two years.
AI’s staggering energy consumption is considered its “dirty secret.” Leslie Miley, technical advisor to Microsoft's chief technology officer, pointed out that the energy consumption of generative AI even exceeds that of ordinary cloud services. Research from Harvard University and UCLA further confirmed that carbon emissions from U.S. data centers have tripled since 2018, now accounting for 2.18% of total U.S. carbon emissions and consuming approximately 4.59% of the nation’s total energy.
There is a sharp contradiction between the rapid development of AI technology and the growing energy consumption and environmental problems. How to balance the development of AI technology and environmental protection has become an important issue before us. In the future, the industry and the government will need to work together to explore a greener and more sustainable AI development path to ensure that AI technology benefits mankind without harming the earth.