Sitting at her desk with homework piled up, Laura* (11) found herself struggling to finish her questions. The allure of plugging the questions into ChatGPT to get a simple answer seemed too tempting to avoid. Even when she tries to limit taking shortcuts, Laura said she ends up using AI as an easier way to finish her work.
“When a new semester starts, I’m like, ‘I’m not gonna use ChatGPT anymore,’” Laura said. “But then, in the middle of the night, when you’re just tired and you have a lot of work, you just end up using AI because it’s easy.”
It’s a scenario high-schoolers are all-too familiar with, and one facilitated by the growth of generative AI, especially large language models (LLMs). The use of AI is one of the many shortcuts that students often take. When faced with difficult problems, long projects, or arduous processes, students tend to look for an easier way out in many aspects of their lives.
Because she is faced with a heavier workload this year than ever before, Laura said it’s easier to have AI skim notes, lessons, and topics and provide easy answers than to go through the process of learning all the material. However, she also noted that her frequent reliance on AI has harmed her ability to learn material on her own.
“AI teaches me, but it’s less [work than a class], because I don’t have to search for the answers, it’s just giving them to me,” Laura said. “For classes like Human Body Systems, I can [put] the article [into] ChatGPT, and I don’t have to read it. But it’s bad for those classes. I feel like it’s [hurting] me because I use it for everything.”
Jeffrey Greene, Ph.D, associate dean for research and faculty development at the University of North Carolina School of Education, said that taking shortcuts, especially in learning processes, reduces an individual’s ability to learn, retain, and master fundamental skills. AI, in particular, allows students to find answers without doing any of the work, which results in a lack of understanding of the material.
“If you’re always depending upon something else, whether it’s generative AI or anything else to do the work, eventually, when you’re asked to do the work, you’re not gonna be able to do it yourself,” Greene said. “You are going to run into a kind of a wall at some point in your career. We see that even in education. The example that comes to mind [is] when I was in school, we learned algorithmic math. We just learned little tricks and formulas to do the math, and that’s fine until you actually have to understand the math to do geometry. Then all those tricks don’t work, and if you don’t actually understand the basics of math, geometry is impossible.”
Laura said she has already seen the effects of frequently using digital shortcuts.
“I realized I lost my ability to think without [AI] because whenever [there’s] a question when the computers are away and I go to do the work in class, I just can’t do it,” she said. “I just sit there and then wait till I get my computer so I can do it.”
Aside from academics, shortcuts are also prevalent when learning new skills in activities or sports. Swar Sinha (9) said she struggled with taking a shortcut while she was learning how to serve a tennis ball.
“You’re supposed to use the continental grip for your serves,” Sinha said. “A lot of times for beginners, because it’s really uncomfortable, they tend to use their forehand grip, which isn’t the way you’re supposed to do it, but it’s much easier. A lot of people use that way instead, but it just ends up hurting your serve over time. If you’re used to doing the wrong grip, then you’re just never going to [improve]. I think I’ve been playing for a couple of years now, and I think that if I had not done the wrong grip in the first year that I was playing, I would have seen much more improvement in my serve. My serves didn’t improve, so I had to re-learn my serve with the right grip, and that wasted a lot of time.”
According to Greene, these negative effects, resulting from taking shortcuts in education, also carry over into future careers.
“If you’re always dependent upon something to do everything for you, at some point in your job or in college or somewhere else, you’re not gonna know what to do, and you’re not gonna be able to ask AI to do it for you,” Greene said. “You’re gonna be stuck, and that’s not a good place to be. Relying on generative AI to do the work for you is indeed a shortcut. It’s a quick way to get a very short gain for a [short time], but eventually that will lead to a very big loss.”
With the pressures of AP classes, Laura said that she sometimes lacks focus and motivation in her classes and ends up using AI to take an easy way out of her work.
“Sophomore year is when I started using it a lot, because I got to my sophomore year, and it was all busywork classes,” Laura said. “And I used to think, ‘oh my gosh, you don’t need to use ChatGPT,’ but now I use it for everything.”
According to Greene, one of the leading reasons that motivates students to take shortcuts is the increased pressure for kids to go to a prestigious college. Especially with more competition in college applications, students feel more of a need to receive all A’s.
“Frankly, it’s kind of a lot of pressure from society,” Greene said. “I know that nowadays a lot of students feel like to get into college or get a good job or have the career they want, they have to do [these] things: ‘I’ve got to have great grades, [have] a million extracurricular activities, I’ve got to start a charity or something.’ And when they feel that way, I understand why sometimes, it’s 11 o’clock at night and you’re trying to finish some paper and go to sleep and you’re like, ‘I just need the answer.’ I think as a society, we have to think about maybe changing some of those pressures.”
Other times, Greene said, students simply lack the motivation to put in the effort that the learning process requires.
“Sometimes students just don’t see the value in what they’re being asked to learn, and I think that’s a place where personalization can really help,” Greene said. “They kind of think, ‘Why do I need to understand it?’”
Greene said these difficulties are often referred to as “desirable difficulties.” Desirable difficulties often require a considerable amount of effort but result in better long-term retention of a subject. Furthermore, in order to fully understand and master a skill, students need to engage in deliberate practice.
“There’s a lot of good research that shows that effort matters in learning,” Greene said. “You could, in theory, just memorize everything or cram before the test. But you won’t understand it, and you won’t remember it long term. So there are these desirable difficulties like spreading out your learning, spacing out your learning, or forcing yourself to explain what you’re learning to yourself or someone else. That makes the learning more difficult, it takes more time, it takes more effort, but all of that helps you understand it better and makes you more likely to remember it for longer.”
Furthermore, when students rely on AI to shortcut the process of learning and understanding, they aren’t engaging in critical thinking themselves. Over time, this can decrease their ability to reason, reflect, and analyze. According to the study “AI Tools in Society: Impacts on Cognitive Offloading and the Future of Critical Thinking” by Michael Gerlich, “younger participants (17–25) exhibited higher AI tool usage and cognitive offloading, but lower critical thinking scores.” These findings show that students who tend to use AI shortcuts for learning find themselves struggling to think critically.
Another negative effect of relying on technology and AI to take the easy way out of things is that the information produced by AI isn’t always necessarily true.
“Generative AI is really fascinating and does a lot of fairly amazing things, but it also hallucinates,” Greene said. “You’re not sure you can always trust it, and so that led me to take an approach to using generative AI that’s more about understanding the content first or knowing as much as you can about what you’re trying to do with generative AI, so you’re not dependent on it, but rather, it is augmenting what you do.”
According to Greene, one way this issue can be combated in education is through the help of educators and teachers.
“I think teachers can help students make a connection with [subjects] that otherwise maybe they haven’t connected with, and that can give them enough motivation to not take a shortcut,” Greene said. “No student’s going to love everything they’re ever asked to learn, but usually we can find ways that students can go, ‘okay, this isn’t my favorite thing, but I understand why it’s important, and I’m going to do the work,’ and that’s what teachers can do.”
When it comes to the utilization of AI, Greene said he believes its benefits depend largely on the user’s expertise and self-awareness. AI can be beneficial in learning when used as a supplemental tool rather than the main source of answers.
“I think the more you know and the more you can do yourself, probably the more helpful generative AI can be,” Greene said. “So for someone who’s really good at something, AI can help them do some of the kind of lower-level stuff that takes a lot of time. AI produces output, and they know to check it and make sure it’s ok. I think there [are] some really promising models, but when you’re learning, it’s not the time to depend upon generative AI. You don’t want to offload learning to AI.”
