Energy-hungry AI is driving up tech giants’ carbon emissions. No one knows how to fix it
![Energy-hungry AI is driving up tech giants’ carbon emissions. No one knows how to fix it Energy-hungry AI is driving up tech giants’ carbon emissions. No one knows how to fix it](https://i3.wp.com/images.theconversation.com/files/605360/original/file-20240708-18-2uji2n.jpg?ixlib=rb-4.1.0&rect=0%2C344%2C5472%2C2736&q=45&auto=format&w=1356&h=668&fit=crop&w=1356&resize=1356,668&ssl=1)
Since the release of ChatGPT in November 2022, the world has seen an incredible increase in investment, development, and usage of artificial intelligence (AI) applications. According to one estimate, the amount of computing power used for AI doubles approximately every 100 days.
The social and economic consequences of the boom have sparked reactions around the world. European regulators recently pushed Meta to suspend its plans to train AI models on users’ Facebook and Instagram data. The Bank for International Settlements, which coordinates the world’s central banks, has warned that adopting AI could change how inflation works.
Environmental impacts have so far received less attention. A simple query to an AI-powered chatbot can consume up to ten times more energy than a typical Google search.
Generally speaking, a generative AI system can use 33 times more energy to complete a task than traditional software. This huge energy demand translates into increased carbon emissions and water consumption, and can put additional pressure on power grids already strained by climate change.
Energy
Most AI applications run on servers in data centers. In 2023, before AI really took off, the International Energy Agency estimated that data centers already accounted for 1 to 1.5 percent of global electricity consumption and about 1 percent of global energy-related CO₂ emissions.
For comparison, in 2022, the aviation sector accounted for 2% of global energy-related CO₂ emissions, while the steel sector was responsible for 7-9%.
How is the rapid growth in AI usage changing these numbers? Recent environmental reports from Microsoft, Meta and Google provide some insight.
Microsoft has invested heavily in artificial intelligence, with a significant stake in OpenAI, the creator of ChatGPT, as well as its own Copilot apps for Windows. Between 2020 and 2023, Microsoft’s reported annual emissions increased by about 40%, from the equivalent of 12.2 million tonnes of CO₂ to 17.1 million tonnes.
These figures include not only direct emissions, but also indirect emissions, such as those caused by the generation of electricity used to operate data centers and those resulting from the use of the company’s products. (These three categories of emissions are called Scope 1, 2 and 3 emissions, respectively.)
Meta is also investing huge resources in AI. In 2023, the company revealed that its Scope 3 emissions had increased by more than 65% in just two years, from the equivalent of 5 million tonnes of CO₂ in 2020 to 8.4 million tonnes in 2022.
Google’s emissions are expected to increase by nearly 50% in 2023 compared to 2019. The tech giant’s 2024 environmental report notes that planned emissions reductions will be difficult “due to growing energy demand from increased AI computing intensity.”
Water
Data centers generate a lot of heat and consume large amounts of water to cool their servers. According to a 2021 study, data centers in the United States use about 7,100 liters of water for every megawatt-hour of energy they consume.
Google’s US data centers alone consumed approximately 12.7 billion liters of fresh water in 2021.
In regions where climate change is increasing water stress, data center water consumption is becoming a major concern. The recent drought in California, where many tech companies are based, has led companies like Google, Amazon, and Meta to launch “water positive” initiatives.
These big tech companies have announced commitments to replenish more water than they consume by 2030. Their plans include projects such as designing ecologically resilient watersheds and improving community water conservation to improve water security.
Climate risk
Data centers located in or near cities may also compete with residents for access to resources during times of shortage. Extreme heat events are one example.
Globally, the total number of days above 50°C has increased every decade since 1980. July 2023 was the hottest month on record.
Extreme heat impacts the health of local populations. A 2022 Lancet study found that even a 1°C increase in temperature is positively associated with increased mortality and morbidity.
On extremely hot days, air conditioning can save lives. Data centers also like to stay cool, so their energy consumption increases with temperature, increasing the risk of power outages and unstable power grids.
And after?
What now? As we have seen, tech companies are increasingly aware of the problem. How is this translating into action?
When we surveyed Australian sustainability professionals in July 2023, we found that only 6% believed data centre operators provided detailed sustainability data.
Earlier this year, we surveyed IT leaders in Australia and New Zealand to get their views on how AI applications are driving increased energy consumption. We found that 72% are already adopting or testing AI technologies.
More than two-thirds (68%) of respondents say they are concerned about the increase in energy consumption related to AI needs. However, there is also significant uncertainty about the magnitude of this increase.
Many IT leaders also lack the skills to adequately manage these sustainability impacts, regardless of companies’ sustainability commitments. There is an urgent need to train IT leaders to understand and manage the sustainability impacts of AI.
News Source : theconversation.com
Gn bussni