AI’s Effect on the Environment
An AI-incorporated future is inevitable, regardless of people’s opinion on whether it will be beneficial or not. While AI has some potential environmental benefits, such as monitoring the environment and helping businesses make sustainable decisions, if changes are not made soon, then AI will be added to the extended list of man-made creations that are destroying our planet.
AI’s Effect on the Environment
When I typed in: “The environmental impacts of AI,” ironically enough, Google’s new AI feature overview was the first thing I saw. AI has slowly been taking over the world over these past few years. Its development rapidly increasing within just the last 5 years. There is no doubt that AI will be unavoidable in the future. This is a terrifying prospect for many people, and fairly so. People’s jobs, hobbies, and humanity itself are threatened by AI. I mean, has Detroit: Become Human taught us nothing? The list of ethical dilemmas involving AI goes on and on — one of these unexpected dilemmas being the negative impact AI has on the environment.
It’s an unexpected connection. How would AI, a computer program, become detrimental to the environment?
Raw Materials Are Being Drained In Hardware Production
To create AI, you need to produce its hardware such as servers, data centers, and microchips. The production, transportation, maintenance, and disposal of said hardware components require a considerable amount of energy and natural resources. The natural resources needed to create microchips are scarce and difficult to obtain; their extraction process is rarely safe for the environment, often leading to other issues like soil erosion or pollution. The servers and data centers as well as the electronics they contain require an astounding amount of raw materials – a standard computer requiring a whopping 800 kg.
These electronics are usually not disposed of properly, contributing to the issue of pollution and growing e-waste. This waste is also generally hazardous or toxic, containing substances like mercury or lead.
Water Waste
AI requires a shocking amount of water to cool itself, a process known as AI server cooling. Data centers use an abundance of water in their cooling towers to cool off air mechanisms that dissipate heat. Nine liters of water are evaporated every kWh of energy used, and it is predicted that by 2027, AI will have used 6.6 billion cubic meters of water. This is severely concerning, considering the current global paucity of freshwater, with a quarter of our population not having easy access to clean water. Water-stressed regions, such as Argentina and Chile, will be particularly affected, as droughts are prevalent in those areas.
Energy Usage and Carbon Emissions
Unsurprisingly, powering these electronics requires a surplus of energy. Asking Chat-GPT a question consumes 10 times more energy than a Google search. This, in turn, produces more carbon emissions since many of these energy sources come from burning fossil fuels. OpenAI researchers stated that, since 2012, the amount of energy needed to train AI programs has doubled every 3.4 months. Researchers at the University of Massachusetts conducted a study to determine exactly how much energy that is. They found that AI training can produce around 626,000 pounds of carbon dioxide. To put this into perspective, that’s 5 times the amount of carbon emissions an average car will produce in its lifetime and the equivalent of 300 plane rides from New York to San Francisco.
Positive Environmental Impacts?
Macbook with ChatGPT open Photo by Emiliano Vittoriosi on Unsplash The future with AI isn’t all bleak. AI could potentially help the environment by tracking patterns and anomalies in data. This could then be used to help monitor the environment by tracking things such as air emissions, and help businesses make environmentally friendly decisions by determining the best way to conserve energy during manufacturing and supply chains.
Recap
An AI-incorporated future is inevitable, regardless of people’s opinion on whether it will be beneficial or not. In addition to being unethical, AI also poses various environmental threats such as raw material drainage, pollution, freshwater waste, high energy usage, and sizable carbon emissions. While AI has some potential environmental benefits, such as monitoring the environment and helping businesses make sustainable decisions, if changes are not made soon, then AI will be added to the extended list of man-made creations that are destroying our planet.
References
https://hbr.org/2024/07/the-uneven-distribution-of-ais-environmental-impacts
https://environment.yale.edu/news/article/can-we-mitigate-ais-environmental-impacts#:~:text=Besides%20its%20energy%20usage%2C%20AI,Industrial%20biology%20and%20 Sustainable%20Systems