Water or we can say the fundamental element of our life, is not only crucial for living beings to survive. You will find alternatives to everything in this world, but no such matter can replace water. If you think that being a living being, only you have the legal right to consume water, then maybe you need to change your thoughts. What if I tell you everyone's favorite ChatGPT is also in the water line?
While ChatGPT has as of late gone under the spotlight for its significant water utilization, the tech business' propensity for water has for quite some time been a worry, a long time before the artificial intelligence chatbot appeared in November 2022.
Microsoft and Google predicted astounding increases in water use in 2022, with Microsoft reporting a 34% increase and Google reporting a 21% increase compared to 2021. Who is the main villain in this aquatic saga? Data centers and their insatiable thirst for water to keep temperatures stable.
Google's 2023 sustainability report unveils the staggering figure of 5.6 billion gallons as the total water consumption at their data centers and offices in 2022. To put this into perspective, that's equivalent to the annual water requirements of 37 golf courses in the arid southwestern United States. Meta's annual report also stresses that operational data centers account for much of their water consumption.
Interestingly, neither Microsoft nor Google had introduced their ChatGPT competitors, Microsoft Bing Chat and Google Bard, until the first quarter of 2023. However, the computer power necessary to run these AI systems, together with their widespread adoption, has the potential to dramatically increase water consumption rates. Unfortunately, we will have to wait until early 2024 for the figures.
While specific numbers for ChatGPT's water usage remain undisclosed, a forthcoming study by Shaolei Ren, an associate professor of electrical and computer engineering at the University of California, Riverside, suggests that it consumes the equivalent of a 16-ounce bottle of water for every 20-50 queries. According to preliminary data from Ren's research, the extent of this digital desire has mostly gone unnoticed.
According to the study, training GPT-3 in Microsoft's cutting-edge US data centers can directly consume 700,000 gallons of clean, fresh water - enough to produce 370 BMW cars or 320 Tesla electric vehicles. It's a startling revelation about the water footprint of AI models.
Now, why do data centers have such an unquenchable thirst for water? CNBC explains that water is the cost-effective choice for tech companies to cool their servers. In essence, it's the unsung hero of Big Tech, ensuring that servers don't overheat as they tirelessly process the world's digital demands.
Google emphasizes its data centers as the engines that power products used by billions of people worldwide, such as Gmail, Google Cloud, Search, and YouTube. While locating data centers in colder places may promise "free" cooling, such conditions are only available to a small percentage of the world's servers. In actuality, the United States has 2,701 data centers, dwarfing other countries such as Germany, which has 487. With its enormous stretch of Siberian cold, even Russia has only 172 inhabitants.
These US data centers are tapping into freshwater reserves, primarily in densely populated coastal regions, a fact outlined by Virginia Tech. Many of these water sources are already strained, particularly in warm climates, exacerbating the problem of outstripping natural replenishment cycles by hundreds or even thousands of years, as The NYT reports.
The future appears to be devoid of hope. According to a report by Intelligent Computing, the computing power required for AI would double every 100 days and might increase by more than a million times in the next five years.
The gravity of the situation is not lost on tech giants. Microsoft, Google, Meta, and Amazon are all exploring various strategies to address water consumption. One key approach is "replenishment" projects designed to offset their water usage. These programs involve international collaboration with NGOs and organizations to improve clean water supplies, fund research to minimize water consumption, and other similar efforts. While they continue to rely on freshwater reserves, the goal is to do so at a slower and more efficient rate.
For instance, Google currently replenishes 6% of its water consumption and aims to replenish 120% of its water use by 2030. Meanwhile, Microsoft has 27 replenishment projects, expected to recover 35 million cubic meters of water. While these efforts are commendable, they are a drop in the bucket compared to Microsoft's 6.4 million cubic meters of water consumption in 2022.
The tech industry faces a massive issue in a world where water resources are becoming increasingly scarce, and demand is likely to surpass supply by 40% by 2030, as predicted by Google. As we navigate the waters ahead, striking a balance between technical advancement and environmental sustainability will be critical.
Read next: Unveiling the High-Tech Hunt with New York's Finest and Social Media Surveillance
While ChatGPT has as of late gone under the spotlight for its significant water utilization, the tech business' propensity for water has for quite some time been a worry, a long time before the artificial intelligence chatbot appeared in November 2022.
Microsoft and Google predicted astounding increases in water use in 2022, with Microsoft reporting a 34% increase and Google reporting a 21% increase compared to 2021. Who is the main villain in this aquatic saga? Data centers and their insatiable thirst for water to keep temperatures stable.
Google's 2023 sustainability report unveils the staggering figure of 5.6 billion gallons as the total water consumption at their data centers and offices in 2022. To put this into perspective, that's equivalent to the annual water requirements of 37 golf courses in the arid southwestern United States. Meta's annual report also stresses that operational data centers account for much of their water consumption.
Interestingly, neither Microsoft nor Google had introduced their ChatGPT competitors, Microsoft Bing Chat and Google Bard, until the first quarter of 2023. However, the computer power necessary to run these AI systems, together with their widespread adoption, has the potential to dramatically increase water consumption rates. Unfortunately, we will have to wait until early 2024 for the figures.
While specific numbers for ChatGPT's water usage remain undisclosed, a forthcoming study by Shaolei Ren, an associate professor of electrical and computer engineering at the University of California, Riverside, suggests that it consumes the equivalent of a 16-ounce bottle of water for every 20-50 queries. According to preliminary data from Ren's research, the extent of this digital desire has mostly gone unnoticed.
According to the study, training GPT-3 in Microsoft's cutting-edge US data centers can directly consume 700,000 gallons of clean, fresh water - enough to produce 370 BMW cars or 320 Tesla electric vehicles. It's a startling revelation about the water footprint of AI models.
Now, why do data centers have such an unquenchable thirst for water? CNBC explains that water is the cost-effective choice for tech companies to cool their servers. In essence, it's the unsung hero of Big Tech, ensuring that servers don't overheat as they tirelessly process the world's digital demands.
Google emphasizes its data centers as the engines that power products used by billions of people worldwide, such as Gmail, Google Cloud, Search, and YouTube. While locating data centers in colder places may promise "free" cooling, such conditions are only available to a small percentage of the world's servers. In actuality, the United States has 2,701 data centers, dwarfing other countries such as Germany, which has 487. With its enormous stretch of Siberian cold, even Russia has only 172 inhabitants.
These US data centers are tapping into freshwater reserves, primarily in densely populated coastal regions, a fact outlined by Virginia Tech. Many of these water sources are already strained, particularly in warm climates, exacerbating the problem of outstripping natural replenishment cycles by hundreds or even thousands of years, as The NYT reports.
The future appears to be devoid of hope. According to a report by Intelligent Computing, the computing power required for AI would double every 100 days and might increase by more than a million times in the next five years.
The gravity of the situation is not lost on tech giants. Microsoft, Google, Meta, and Amazon are all exploring various strategies to address water consumption. One key approach is "replenishment" projects designed to offset their water usage. These programs involve international collaboration with NGOs and organizations to improve clean water supplies, fund research to minimize water consumption, and other similar efforts. While they continue to rely on freshwater reserves, the goal is to do so at a slower and more efficient rate.
For instance, Google currently replenishes 6% of its water consumption and aims to replenish 120% of its water use by 2030. Meanwhile, Microsoft has 27 replenishment projects, expected to recover 35 million cubic meters of water. While these efforts are commendable, they are a drop in the bucket compared to Microsoft's 6.4 million cubic meters of water consumption in 2022.
The tech industry faces a massive issue in a world where water resources are becoming increasingly scarce, and demand is likely to surpass supply by 40% by 2030, as predicted by Google. As we navigate the waters ahead, striking a balance between technical advancement and environmental sustainability will be critical.
Read next: Unveiling the High-Tech Hunt with New York's Finest and Social Media Surveillance
