In especially recent years, AI has began to play both a prevalent and controversial role in society. AI, which stands for artificial intelligence, is technology designed to perform actions that humans typically do, but at a faster rate (IBM). People use them for studying, learning, looking things up, and more. A prime example is ChatGPT, which people use for various things like teaching themselves content, doing their math homework for them, and more. But how exactly does it impact the environment? On one hand, AI can be trained to do things at a better rate than we can, potentially making the solutions of environmental damage easier to achieve, but on the other hand, AI does require a lot of energy for its training. Just how good is it for the environment, and vice versa? Is there a way we can combat this or use AI effectively so we don’t damage the environment but rather, help it
How AI Helps the Environment
AI, according to Gavi.org and World Economic Forum, can be trained to do various things that humans cannot achieve in the same amount of time. As the global climate increases, icebergs unfortunately will melt as well. These are particularly difficult to detect, but scientists at the University of Leeds stated that they could detect icebergs in a miraculously short amount of time – less than a second. However, for humans, this is difficult, as the areas surrounding icebergs are quite monochrome. Additionally, another thing both Gavi and World Economic Forum points out that AI can map out deforestation. This is extremely useful as then we can use our resources appropriately for the places that need them the most. According to the UN News, AI can also be used to fine-tune cloth manufacturing, leading to less waste being produced. AI can also improve weather forecast accuracy, so in the event of a hurricane or other weather-induced natural disaster, people are more prepared (UN News). CleanHub also states that AI can review plastic collection – when retrieving pollution from the ocean, it can review data that companies like CleanHub enter to make sure it is accurate. It also can help with recycling plastic waste, using smart recycling bins that detecting the type of waste discarded, and even implementing recycle robots that help make the picking process easier. That’s all well and good – AI, when used for the right reasons, can be really useful! Sadly, there are still cons to them.
How AI Hurts the Environment
There are, unfortunately, some pretty detrimental effects from AI as well. According to Earth.org, the amount of energy required to train AI is huge. Anders Nordgren states in an Emerald Insight journal that emissions from the Information and Technology industry by the year 2024 would reach 14% of global emissions. This is a concerningly large number, and judging based on how much AI is getting to become common use, this will only increase. Additionally, at the University of Massachusetts, a study was held to see how much carbon dioxide actually comes from training multiple common large AI – and it was 626,000 pounds! Enough carbon dioxide emissions are released every year, and this is not helping matters. Besides electricity consumption, water is also an issue with AI – it is necessary to cool the hardware, according to MIT News. This puts a big strain on water supply. Even after training the AI, there are still damaging effects on the environment. Every time you ask AI to do something for you, that consumes energy, and even more so (5 times more) than just looking it up (MIT News).
What Can We Do About This?
The effects of AI are quite bad for the environment, but at the same time, it can be used to help, so we need to find a balance. On a wider scale, what we can do to lessen the impact of training AI is by developing energy efficient hardware, as Earth.org says. Additionally, according to Harvard Business Review, we can also use already existing tools, not constantly develop new ones, and fix the ones we already have. We can make them more precise so as to use less energy. Another thing on a wide scale to do is to use a large model only for actually intensive purposes (Harvard Business Review). We also should be implementing carbon monitoring, as Harvard Business Review says. From there, after improving how we use the AI, we can easily target things in our ecosystem that require our help. In general, as well, companies need to be more transparent about what the effects of AI are and just how many carbon emissions they’re outputting, as MIT News says. This effectively will make people more concerned about their impact on the environment.
On a smaller scale, we need more people to be aware of this. Many people use AI, and they should be considering not just the effects of AI but how we can effectively use it to help the environment. We have to consider that AI can be rather harmful if it is still used in this particular way, but if we make changes to how we implement it, we can easily move forward in fixing the environment. A simple google search (which costs less energy) also helps you learn more about this topic (since to be completely honest, this barely scratches even the surface of possibilities), and thus knowledge is power!
Tell others about how AI can affect the environment, and the more others are aware, the more easily change will come. There is little doubt that we will continue to use it in our futures, but we also must consider our environment at the same time. AI may be the future, but if we have to sacrifice our environment at the same time, there is nothing left for us.