'AI race has increased environmental pollution'

Although Google has announced to achieve the goal of Net Zero by 2030, the company has indicated that the challenge has been added to the achievement of the goal.

It is mentioned in the company's report that the production of carbon dioxide gas, called the carbon footprint of Google alone, increased by 13 percent from the previous year to about 14.3 million tons in 2023. Google's report is just one example, various studies have shown that the demand for energy will increase by 160 percent by 2030 due to AI competition spread globally.

As a reflection of the challenge that AI infrastructures will add to the goal of reducing global climate change with the increasing demand for energy, the public Google environmental report has come out last week . Computing infrastructure used by AI platforms like ChatGPT consumes 33 times more energy than ordinary computers, and as the number of such companies increases, the challenge of greenhouse gas emissions will increase, according to a research by machine learning researcher Sasha Lucioni.

The International Energy Organization estimated that the energy consumed by AI-related data centers was 460 terawatts in 2022, and this figure could reach 1000 terawatts in 2026. This number is equal to the energy consumed by Japan, which has a population of about 130 million and is developed in terms of infrastructure. A study published in March stated that energy demand in the UK alone will increase by almost six times over the next 10 years, and that AI will be one of the main reasons.

The race of AI is not limited to Google and ChatGPT, the world's largest Silicon Valley companies Meta, Amazon, X, along with various technology companies from China and European countries are in the race for dominance in the AI ​​market in their own way.

A study shows that the same model used for AI language emits 626,000 pounds of carbon dioxide annually. This is equivalent to the annual emissions of 63 petroleum-fueled vehicles.

Similarly, along with the physical and other infrastructure required for AI companies, the daily increasing investment and demand is also increasing energy and toxic gas emissions. The news that the other facilities are being reconsidered was published by Alaska's Anchorage Daily .

By the end of 2024, the global market of AI will reach about 184 billion US dollars and the number will continue to grow by more than 20 percent in the next 7 years. Researcher Lucioni estimates that the big problem of energy demand and emissions will be a challenge for the US as most of the major AI companies and data centers are located in the US, where there is an AI market of over 50 billion US dollars.

Google's report indicates that although the energy consumed by data centers in Europe and America is mainly of renewable nature, data centers in Asia, Africa and other regions will increase carbon emissions.

'People may not realize it, but every query entered into AI, every data stored in the cloud and every cryptocurrency activity is increasing the digital carbon footprint and energy consumption,' says Sasha. are .' Along with global regulation and market management, AI companies have to think of environmentally friendly alternatives to pollution and wasteful energy consumption.