“This invention will produce forgetfulness in the minds of those who learn to use it, because they will not practice their memory…They will imagine that they have come to know much, while for the most part they will know nothing.”
That sounds like sharp criticism of AI, but those were actually the reported words of Socrates about 24-hundred years ago, and the invention he worried would ruin young minds was writing. Fortunately – and ironically - one of his students, Plato did write those words down.
Now, no one wants to get rid of writing, but Socrates have a good point. Every new advance in information technology has raised worries about what we now call cognitive offloading. We don’t use paper maps in the car anymore. We just do whatever Waze tells us to do. Cell phones have always had a Phonebook app, so most of us don’t remember anyone’s number. And when our phones got “smart” we gained the magical power to get the answer to any question immediately.
Want to know about ancient Mayan pyramids? There is, I think, very close to zero chance you will go to the library and check out a bunch of books. Google provides 11.8 million hits on that search term in .19 seconds.
Still, this time with generative AI it feels different.
After Brown University professor Roberto Serrano gave his first take-home exam of the Spring semester, he knew something had gone wrong. The average score for the class came back at 96 out of 100, with forty perfect scores. So, Serrano told the class the final exam would be in-person and closed-book.
A flood of students dropped out, including 22 who had received those perfect scores. The average score in the final dropped from 96 to just 48. So, Serrano and his colleagues everywhere are reassessing not only how they teach, but how they assess what students have actually learned.
Some are taking a hard line.
The University of Chicago banned the use of AI in all of its core social science courses. Its law school had already banned laptops, tablets, and wearable electronics from first year classes to encourage pen-and-paper critical thinking. Duke University has even banned applicants from using AI to write admissions essays.
However, more than $4 trillion in AI investment in just the last two years suggests it’s here to stay. That's why some others argue it might make more sense to treat AI like the proverbial horse and ride it in the direction it’s going.
In his piece in Juristnews three months ago, Maryland Carey Law professor Donald J. Tobin, JD, wrote this about restrictive policies at certain other law schools:
“The deepest flaw in these restrictive policies is that they misidentify where the cognitive work now lies. Increasingly, the lawyer’s task is not to produce a first draft from a blank page but to direct, interrogate, and evaluate machine-generated work product.”
And if there are to be rules about the use of AI, who should make them, the schools or individual professors? Broad policies may be easier to articulate but might not fit every subject let alone every class. On the other hand, students may be confused to find different rules obtain in each class.
And what do students think about AI in the classroom? In a University of British Columbia survey, students’ concerns closely tracked with their teachers. About three quarters felt unrestricted AI use harms learning. An AAC&U study also revealed concerns about decreased attention spans and of course cheating.
So, the questions for academics seem clear: If AI can write essays, solve problems, and synthesize research, how do we preserve the kind of deep disciplinary understanding the next generation will need to best serve their patients and clients? And, how do we make sure students and faculty have the proper baseline facility with AI tools for their studies and their professions?
On the October 28, 2026 edition of Virtual Face to Face, guest host Roger J. Ward, EdD, JD, MSL, MPA, tackled those questions with the help of a panel of University of Maryland, Baltimore (UMB) faculty.
Isabell May, PhD, heads the UMB Writing Center, has led UMB discussions on generative AI numerous times, and also authored UMB's recent Writing, Learning, and AI: Emerging Conversations at UMB.
The University of Maryland School of Pharmacy's Scott Riley, PhD, co-hosts the Faculty Center for Teaching and Learning’s podcast, Moving the Needle. From September 2025 through May of this year the podcast ran a series titled, AI Unscripted, addressing many facets of AI use. Riley is also scheduled to lead an online Teaching Community session in January called Confidence by Design: Transparent Teaching with AI.
And University of Maryland Francis King Carey School of Law professor Donald Tobin, JD, added to his long list of books and publications on tax law and constitutional issues this summer with a sharp critique of restrictive law school AI policies in Juristnews titled: Law School AI Bans Teach Exactly the Wrong Lesson. A year earlier, Tobin joined other colleagues from Maryland Carey Law behind lead author Andrew Blair-Stanek on a series of papers analyzing several ChatGPT models’ performance on law school exams, titled: “AI Gets Its First Law School A+s.”
Watch the one hour discussion by accessing the link at the top of this page.
Excellence in Teaching
