OpenAI released their newest model, o1-preview, two weeks ago on September 12. According to OpenAI’s report, the o1-preview model now performs at a PhD-level in subjects like physics, chemistry, and biology. Again, advances in technology bring about change in knowledge, both in how we understand and consume it. And Andover is no exception. According to the 2024 State of the Academy (SOTA), 45 percent of respondents have used artificial intelligence for an academic assignment without explicit permission. At the same time, the Course of Study directs students to “assume, unless the instructor explicitly highlights otherwise, that the use of LLMs is prohibited.”
Life at Andover is filled with learning. This, paired with Andover students’ ambition and innate desire to know, to master, and ultimately, to progress and move on, manifests in a sort of impatience that is defined by the objective to finish assignments as quickly as possible. With students taking heavy class loads, the natural desire is to do as much as possible as quickly as possible. But while this mindset may be motivating at times, its heavy emphasis on speed and completion leaves many students vulnerable to the tempting efficiency of artificial intelligence (AI), especially in a high-pressure school environment where we are already spread so thin and always chasing after time. After all, if ChatGPT can spit out perfect proofs for the mathematics theorems on my homework in less than 15 seconds, why should I bother with formulating them on my own? Of course I’ll read it over, make sure I understand it for the most part, but at the end of the day the quicker the better. In this sense, our “ambition” is clouded by our obsession with results and checking things off our to-do lists, and we give up our genuine learning to the magnetic pull of an AI that does everything we don’t want to do or don’t have the time to. On the surface, it may seem like we are gaining time, freedom to invest in other activities. But this perceived benefit comes at the sacrifice of the invaluable experiences of working through an especially challenging physics problem or dissecting a difficult and convoluted history reading, a kind of suffering of the mind that is conducive, if not necessary, to the development of critical thinking skills, the ability to synthesize and retain information, and a consciousness of the self and how we learn as individuals. If we choose to rely on AI without truly understanding the answers ourselves, without working through them, we are depriving ourselves of the gratification and growth of figuring out something on our own, of emerging victorious after a long and hard battle. The issue of academic integrity presents a nuanced consideration when it comes to AI usage. Andover, like many schools across the country, places academic integrity on one of the highest pedestals of student virtue, as it rightly should. Plagiarism is vehemently condemned, and the beginning of every term sees teachers reviewing academic integrity policies with their students. In both formal writing and real classrooms, Andover instills the value of academic integrity into our minds and teaches us to disdain academic dishonesty through rhetoric that praises honest engagement while associating all violations, even inadvertent ones, with an air of punishment. And while it is true that AI usage can entail a severe breach of academic integrity in many cases, it is imperative to note that AI is still a continuously evolving technology. In shoving AI and LLMs under the umbrella of academic integrity violations, Andover’s policy has the effect of vilifying them, fostering a culture that overlooks the myriad of ways AI can also be used to support and enhance learning, too.
In many cases, the question of whether AI does more harm than good depends on whether the assignment is process- or result-based. For tasks like textbook readings and analytical essays, the purpose is not only to take in the information at surface level, but to learn how to process and interpret it in a way that allows us to grasp its meaning and form our own ideas about it. In these situations where the value of the assignment comes from going through the process of reading and writing ourselves, the use of AI only hinders our own learning. But for result-based homework like looking over instructions in order to prepare for a lab, AI can serve as a potential tool for both aiding comprehension and reducing the time needed to complete the task, provided that we are not leaning on it as a crutch to understand the content.
Often, the appeal of AI to students lies not necessarily in its expertise, but in its accessibility and memory. When we are struggling with a specific chemistry topic or need help memorizing human anatomy for our next bio test, we cannot expect our teachers to be available at all times of the day, ready to spend as much time assisting us as we need. As a study buddy, AI has a convenient on-demand quality that allows us to call on it whenever, and its ability to remember the context of our previous questions increases its effectiveness compared to simply searching online. It can also perform functions such as formulating practice questions or making flashcards that may be tedious or difficult to do on our own in real life. As long as we are resolutely cognizant of our use of AI and ensure that it does not cross the line from aiding our studies to completing our studies for us, its vast array of capabilities does highlight promising benefits to our learning.
Changes in artificial intelligence are irreversible. With each new model, each iterative improvement, a new paradigm of creation, learning, and development is unlocked. It can harm us and aid us all at once. Regardless, it will be an unmistakable part of our lives.
