Environmental Impacts of AI

By Katherine Abayev, Staff Writer

Artificial Intelligence (AI) is a new technology that has impressed the masses with its incredible ability to imitate human conversation and originality. These AI chatbots, like ChatGPT, can solve equations, give advice for multiple situations, and change their responses based on given feedback. AI is famous for its unique responses. The models are fed massive datasets and can search for patterns and trends to make conclusions based on that information. The AI models are then tested over and over again to make sure their information is accurate.

While having this new technology is incredible, unfortunately, there are drawbacks when it comes to its environmental impact. Training AI models takes massive computing power which outputs tons of carbon dioxide emissions. According to Earth.org, ChatGPT outputs 8.4 tons of carbon dioxide per year alone. Also, at the Massachusetts Institute of Technology (MIT) Sloan mentions that “Since 2012, the most extensive AI training runs have been using exponentially more computing power, doubling every 3.4 months, on average.” This steep progression has a more detrimental effect on the environment.

Junior computer science majors Lucas Hornung and Nodira Kazakova are well-informed about AI and the environmental crisis.

“It is important for people to know that just one hour-long conversation with AI can eat up so much energy and to be conscious of that,” Kazakova said.

“Data centers account for 2% of the global energy unit, the more that people use AI and rely on AI the more that these data centers have to operate and grow, I’m no expert, but I think that most of these companies are not environmentally informed, and they are more focused on their growth,” Hornung added.

Kazakova addressed the classified nature of AI. 

“It is confusing for me to know how much environmental damage these corporations are doing because most of it is kept hidden, so most of our information is estimates. There is a lot we just don’t know about AI.” Kazakova said.

Although the environmental impacts of AI should not be ignored, it’s important to acknowledge that AI has many benefits as Hornung discussed. 

“On one hand, I feel that AI is dangerous. As someone who plans to enter the climate science and academic scene, AI has benefits for climate change and has a positive effect on extreme weather communication. A lot of Americans don’t speak English at a proficient level, so AI can communicate with Spanish-speaking people and inform them about potential climate disasters headed their way. Also, AI is getting more useful for climate modeling scenarios, it can be transformative. AI should be developed and maintained sustainably, so its abilities can shine through.” Hornung shared.

Now, the question must be begged, do the pros of AI outweigh the cons that it creates for the environment?

“It’s a tough question, but I am leaning more towards not worth it because research and predicting is not actively doing anything for the environment, but the greenhouse gas emissions are making the environment worse,” Kazakova said.

Hornung gave a response that was more in favor of the pros of AI.

“That is one that I would need to see the data on, but yes, in climate science, the pros outweigh the cons. A lot of climate models that are run for climate science reasons like NCAR, the National Center for Atmospheric Research, have AI and that is operated by researchers. Because this is their industry they are more informed, however, companies like Apple and Google are more interested in their gain and creating the model, than the environmental effects. I think a differentiation needs to be made. AI has a net positive effect on the world,” Hornung answered. 

Overall, AI has the power to create real change and that cannot be denied, however, something can be good and bad at the same time. The immense emissions that AI can create are detrimental to the environment and the world is in no state to accept any more damage. Unfortunately, the Earth has reached a very weak point where measures have to be taken to limit emissions and pollution to preserve what is left.