According to The Guardian, Google's increasing reliance on energy-intensive data centers to support its new artificial intelligence products is at risk of undermining its goal of reducing its carbon footprint.
Data released by Google on Tuesday showed that its greenhouse gas emissions have grown by 48% over the past five years. The company attributes the growth to the energy consumption of data centers and supply chain emissions. In its annual environmental report, Google also revealed that emissions in 2023 increased by 13% compared to the previous year, reaching 14.3 million tons.
Google, a company that has invested heavily in artificial intelligence, stated that achieving net zero emissions by 2030 "is not easy." The company acknowledges that achieving this goal involves "major uncertainties," including the "uncertainty of the environmental impact of artificial intelligence in the future, which is complex and difficult to predict."
Data shows that Google's emissions have increased by nearly 50% since 2019, which is the baseline year for the company's net zero emissions goal, requiring that the amount of carbon dioxide emitted is equal to the amount removed.
The International Energy Agency estimates that the total electricity consumption of data centers will double from 2022 levels by 2026, reaching 1,000 terawatt-hours (TWh), equivalent to Japan's electricity demand level. According to calculations by research firm SemiAnalysis, by 2030, artificial intelligence will cause data centers to account for 4.5% of global energy generation.
Data centers play a crucial role in training and running artificial intelligence models, such as Google's Gemini and OpenAI's GPT-4, which power the ChatGPT chatbot. Microsoft acknowledged this year that energy use related to its data centers is threatening the company's "moonshot" goal of achieving carbon negativity by 2030. Microsoft President Brad Smith admitted in May that "the moon has changed" due to the company's artificial intelligence strategy.
Large technology companies have become major buyers of renewable energy to achieve their goals of reducing carbon dioxide emissions. However, the commitment to reducing carbon emissions is now in conflict with their heavy investment in artificial intelligence products that require a lot of energy to train and deploy in data centers, as well as the carbon emissions associated with manufacturing and transporting the computer servers and chips used in this process. Water usage is another environmental factor in the boom of artificial intelligence; a study estimates that by 2027, artificial intelligence could consume up to 66 billion cubic meters of water, nearly two-thirds of England's annual water usage.
Key points:
- Google's greenhouse gas emissions have grown by 48% over the past five years, mainly due to the energy consumption of data centers and supply chain emissions.
- Google's "ambitious" goal of achieving net zero emissions by 2030 faces significant challenges due to the uncertainty of the future environmental impact of artificial intelligence.
- The energy consumption of data centers, as well as the carbon emissions from manufacturing and transporting related equipment, make it a conflict for large technology companies to reduce carbon emissions and promote the development of artificial intelligence.