Banning AI in the classroom would be a generational mistake
American higher education is in a tug-of-war over the merits and potential abuses of AI. Fearing the self-learning artificial intelligence algorithm will render meaningless the traditional teacher-student relationship, some want to ban the technology from the classroom altogether.
That would be a generational mistake.
Each generation of students, it seems, is faced with some newly discovered technology that threatens to destroy the education system as we know it. At one point it was the handheld Texas Instruments calculator, a wonder of 1970s digitization, that many mathematics instructors initially forbade students from using. Mathematics was considered too important for personal development to hand off to a machine. It was pencil and eraser or nothing.
Of course, it turns out calculators didn’t upend math education as many feared. Educators adapted the new technology into the curriculum and pivoted to teaching broader concepts than simple addition and subtraction.
Similarly, AI holds the promise of simplifying the learning process. Granted AI is a far more powerful piece of technology than a passive desktop calculator. But history tells us it would be equally naïve and unforgiveable to shut AI out of the classroom out of fear of its disruptive potential.
Colleges have an obligation to prepare students for a world in which AI is as much a part of everyday life as the smartphone — or the calculator. Rather than running scared out of fear of its misuse, they should embrace its potential optimistically. AI is bound to make some post-college jobs obsolete, but those that remain will be the ones that require capabilities that are uniquely human. Developing and enhancing those capabilities will be key.
Despite rapid advancements, generative AI in its current form is far from a polished product. It does not (yet) have the ability to discern fact from fiction in all instances. The models don’t react well in different contexts. They often lack common sense, and sometimes respond in ways that are awkward and outside the norm.
And no matter how far it advances, AI will never fully supplant the classroom experience, or generate the kind of creative understanding that is at the heart of a well-rounded education. At the University of San Francisco, for instance, undergrads are taught the rules of rhetoric, a cornerstone of Jesuit education for 450 years. Students learn to construct a thesis, craft an argument and then write an essay or a lab report that backs up their ideas. Underlying this effort is the responsibility to say something that is true, good and perhaps even beautiful.
Students, in other words, are taught to think beyond mere concepts and apply what they learned to real-world experiences in a way that is unique, creative — and human.
Inviting AI into the classroom will require a level of trust that students won’t abuse its conveniences — another reason to maintain strong teacher-student relationships. It will also likely mean a return to past assessment practices. Back in the mid 1980’s, it wasn’t uncommon for professors in certain courses to give oral exams to assess not only what students had learned but also their ability to think on their feet.
Similarly, AI could perpetuate a return to active classrooms, where students are graded on their ability to demonstrate oral mastery of the subject and the rules of interpretation. In-class writing assignments, using pen and paper, will be another means of assessing learning; some faculty never stopped giving bluebook in-class finals.
None of these older forms of assessment preclude the use of AI as a useful educational device.
With or without AI, the purpose of education remains the same — to acquire knowledge, develop the ability to reason, make refined judgments, imagine new realities and become an ethical adult. Ultimately, AI can be a helpful servant or a hurtful master. Our choice.
Paul J. Fitzgerald, a Jesuit priest, is president of the University of San Francisco.