By Taarika Sethee (she/her), Youth Volunteer Coordinator
In today’s world, we use AI in our daily lives, and this usage will only go up as technology develops and we become further dependent on it. However, as an environmentalist, I only recently discovered the negative impacts this new technology has on our environment. In this article, I’ll explain some of AI’s major harmful impacts on the environment and finally, what we can do about it.
AI systems use an incredible amount of energy to run, and often they must run continuously. One form of energy used is electricity, which emits various harmful pollutants and greenhouse gases; especially because fossil fuels are how electricity is generated in many parts of the world. The reason AI uses so much energy is because of the incredible amount of computing power and data processing it uses. In addition, the training needed for an AI model can emit hundreds of tons of carbon into our atmosphere. Furthermore, the hardware needed for AI requires rare elements and critical materials. These resources are often mined unsafely, and the production of the metals also leads to soil erosion and overall pollution.
The electronic waste from AI is devastating. The materials from the stages of production, maintenance, and disposal of the hardware from AI create a lot of electronic waste. These are often not recycled properly, which furthers its impact on our environment. The materials these devices use can cause soil and water contamination if not disposed of properly. The MIT Technology Review estimates 5 million tons of e-waste will be produced by 2030 which accounts for a percentage of the global amount. Although there is only a sliver of global waste at 60 million metric tons right now, the concern is that this number will only exponentially increase as AI becomes increasingly popular.
So what do we do about this? Well, obviously, considering AI’s electricity and energy impacts, advocating for a transition to clean energy would majorly help. Also, it’s important that companies disclose the negative environmental impacts their company is making through their AI services. Researchers are studying more chip designs, some prominent ones include in-memory computing, analog methods, and photonics to make the systems more efficient and make AI consume less power. Hardware and software collaboration is essential to develop more environmentally friendly solutions. In conclusion, a transition to clean energy would protect us from the brunt of AI’s consequences, which is just another motivation to say no to fossil fuels.