Being on a college campus, it feels like the use of artificial intelligence is everywhere. I hear about it from my professors, fellow students, and even in my textbooks. Beyond campus, just being a citizen of the world means having AI looming over your head. AI is the tech industry’s shiny new toy and personally, I feel stuck on how so many people have so quickly accepted it.
Students have always found ways to cheat. On assignments, tests, essays, you name it. If there is a will, there will be a way. It is hard to make an argument against AI use in education without bringing this up first. But times have changed from writing down answers to a test on your palm. AI can be used to write entire papers, answer all of the questions on a quiz, and even compose emails. There are AI chrome extensions that can virtually take a quiz for you. No thinking required, just copy and paste. I’ll sit behind other students in class and watch as they instantly copy an assignment into ChatGPT.
This isn’t to say that every college student is using AI on all of their assignments. They’re not. I have had many productive conversations with peers about our grievances and worries about AI. Many of my own complaints are shared, but I remain in the minority. The Digital Education Council’s 2024 Global AI Student Survey showed that 86% use AI in their studies, 24% using AI on a daily basis.
Artificial intelligence has been around for a long time. Although it feels like this is completely new technology, the groundwork for artificial intelligence was laid in the 1950s by scientist and mathematician Alan Turing. What we think of AI today is usually generative artificial intelligence. Generative AI consists of algorithms that are used to create new content. This can range from audio, text, code, images and videos.
One of the main issues that I see with Generative AI being used for schoolwork is that it impacts your ability to synthesize new information. When something is difficult and stumps you a little bit, you learn by doing. Figuring it out, messing up, and getting questions wrong are all essential parts of learning. When you use AI for assignments and start skipping these steps, your learning can be hindered,. A study done by the University of Pennsylvania showed that when AI is used as a crutch for assignments or practice problems, students typically preform worse on their own. The study showed that students who used ChatGPT on their homework scored 17% worse on their final exams than the students who did not use ChatGPT and did the practice work on their own.
There are many arguments that using AI features such as ChatGPT can increase productivity, and be a helper in research. These features are less helpful when asking homework problems, since the program often doesn’t give an explanation to the answer, it will just spit out the answer. While this is helpful in the moment, it can limit your ability to retain said information considering that you did not do the work to find the answer yourself.
In addition to this, there isn’t always a guarantee that the answer ChatGPT will give you will be correct. In my brief stint trying to give ChatGPT the benefit of the doubt, I asked for it to pull out specific quotes from a book that I could use to analyze in a paper I was writing. Every quote that the program gave me was not actually in the book. No matter how many times I specified the book or clarified exactly what I wanted, it still gave me faulty information. The quotes were completely fabricated and found nowhere within the pages.
This experience I had was not a unique one. Researchers at Purdue University found that 52% of programming answers from ChatGPT are incorrect. In this study they looked over 517 questions and analyzed ChatGPT’s ability to answer them. Regarding their findings they wrote, “We found that 52 percent of ChatGPT answers contain misinformation, 77 percent of the answers are more verbose than human answers, and 78 percent of the answers suffer from different degrees of inconsistency to human answers.” Would you trust a search engine that only gives you accurate information less than half of the time?
In addition to being unreliable when it comes to accurate information, generative AI has been found to be potentially harmful to the environment.
There is a lot of power needed to run and train generative AI models. These parameters often demand an overwhelming amount of electricity. Researchers have estimated that asking ChatGPT a question uses about five times more electricity than a simple web search. These increased amounts of electricity can also lead to increased carbon dioxide emissions. In addition to this, large amounts of water are used to cool the data centers that house the computing infrastructure that runs and trains AI models. It has been estimated that, for each kilowatt hour of energy a data center consumes, it would need two liters of water for cooling. As AI grows, these issues are only going to get worse
Despite my personal biases, there are real facts shown above that convey how detrimental AI use can be. There are harmful environmental implications that will only be exacerbated as these engines continue to be used. Additionally, as demonstrated in the University of Pennsylvania study, the use of AI for academics might lead you to preform worse in class as you synthesize less information. I think this is an important fact to remember given that 52% of ChatGPT answers contain misinformation.
The bottom line for me is to be wary, think critically, and take the time to learn about how the things you use can not only affect you, but the world around you.