Despite hitting the public market less than three years ago, AI has already become an incredible staple in the average student’s life. According to a recent Inside Higher Ed survey of over 1,000 US college students, approximately 85% reported that they used generative AI for schoolwork in the past year. According to another international survey of students at the bachelors, masters and doctoral levels, 54% rely on AI frequently, referring to it for help on a daily or weekly basis. It’s being shouted from the rooftops by news outlets everywhere that AI has “unraveled” academia and despite the sensationalism, they’re not entirely wrong.
It’s clear that choosing to use AI is the path of least resistance right now, with incomprehensible amounts of money being poured into newer versions of ChatGPT that are increasingly effective at making school easy at the surface level. Still, what’s easy is not always right and what is popular is not always good. This is the hill which I’ll die on today: to the 85% of students who use AI in classes, there is something important lost in the ease of AI, and we as students need to take it out of our schoolwork.
The aforementioned Inside Higher Ed survey helps to contextualize the situation of AI in the university. To begin, there are important considerations in how students are using AI, particularly in completing written work at each stage of the writing process. 55% of respondents reported having AI help with brainstorming ideas, 38% used it to generate summaries of material, 31% prompted it to outline their papers, 25% just allowed it to complete assignments for them and 19% simply had AI write entire essays for them. In all these use cases, AI is used as a substitute for a very specific and crucial part of the different stages of assignments – the part where you think about them.
Whenever you outline a paper, you are actively practicing logical reasoning to connect different facts and ideas in a way that creates an argument. When you brainstorm ideas for an artistic or academic project, you are practicing creative thinking to come up with something new by making connections about things you already know. The list goes on, but the point remains the same: when you have AI to do these things, you stop thinking.
Now, if this high a percentage of students are using AI on a daily or weekly basis, how much less critically or creatively are they thinking compared to those who don’t use AI? What are the consequences of that? Well, MIT recently published a groundbreaking study which showed that people are actually just becoming dumber as a result of AI use. To be more specific, researchers at the MIT Media Lab compared the brain functions of groups of subjects after writing basic SAT level essays, with the difference being the resources available to certain groups in the writing process. One group had no access to outside resources, while another could use a traditional Google search and the third had full access to ChatGPT. The findings were incredibly evident: the group that used ChatGPT had significantly lower brain connectivity and activity on neural, linguistic and behavioral levels than both the other groups. They got progressively lazier as the project went on and often couldn’t even remember what they had said in their essays. Let it be noted that these results came from several sessions over the course of four months, which pales in comparison to the amount of usage that, as stated above, most students are using AI in their own schoolwork.
The ramifications of this study are clear: AI is completely antithetical to the goals of an education and the learning process. The more we allow it to creep into our lives, the more we are going to enable the systemic creation of future generations that simply lack the ability to think for themselves. It should be fairly clear without much explanation that that is very bad.
Yet, it’s also unfair to completely ridicule those who do use AI for help, especially since it has become a seemingly large majority. There is still truth to the fact that people want this for a genuine reason, and that underlying reason must be addressed. Inside Higher Ed showed that the top two reasons students reported for wanting to use AI in school were a pressure to get good grades and a lack of time to commit to schoolwork. It’s not that they were just lazy, but that when given a choice between prioritizing the learning process and getting good results in class, they chose the latter. This is the result of an education system that routinely prioritizes results over process, GPA over actual understanding of content and career preparation instead of the academic pursuit. In this case, the allure of AI, which provides a shortcut to results, is completely understandable. The “unravelling” of the university is in fact completely self-inflicted.
Still, that does not mean that we as students do not have the agency to choose otherwise. Regardless of the context, choosing to go toward AI isn’t beating the system – it’s just falling victim to it. Even if the education system we exist in may not value actual learning as much as it ought to, students can still make the choice to embrace it. For the sake of our own lives, which are surely enriched by the experience of living and learning, as well for the sake of future generations, we must reject the use of AI.