Since ChatGPT became available to the public in November 2022, educators have found themselves caught in perpetual AI uncertainty. Chatbot programs caused a generational disruption to the traditional classroom dynamic, forcing professors to rewrite syllabi that had gone unchanged for years.

It’s a tremendous responsibility for professors to design a balanced policy on artificial intelligence use in their classrooms, with ethical considerations far less straightforward than traditional plagiarism. As AI becomes more familiar, an increasing number of professors are making space for it in their classrooms.

However, if professors want to uphold the tradition of liberal education, they should limit the level of AI in their classroom as much as possible. 

University of Minnesota communications associate professor Susan LoRusso adopted a more flexible stance, where the use of AI is permitted so long as it’s effective and useful.

LoRusso said she generally permits the use of AI as a data aggregator for research projects, even incorporating an assignment designed around creating data tables in Google Gemini for her upper-division courses. Her policy is nuanced and highly case-dependent, but the bottom line is that any AI use must be done openly and cannot replace fundamental components of critical thinking. 

“As long as you disclose, even if you used it in a way that doesn’t fit the parameters of the class, that’s actually not being dishonest,” LoRusso said. 

This less punitive approach allowed her to better understand how her students are using AI, and she learned many of her students lacked a basic understanding of its functionality. In response, she developed guides teaching them to use it more effectually, an effort aligned with a broader push at the University to promote constructive AI use.

Much of this initiative is concentrated in Walter Library’s newly unveiled AI Makerspace, a project designed to offer students more exposure with AI and help the University stay competitive in this evolving technological landscape. It acts as a general resource for everyone, ranging from first-time users to advanced computer science students. 

AI Makerspace manager Joseph Finnegan said the primary function of the Makerspace is creating models that solve specific problems, like streamlining various parts of academia, such as combing databases for information. 

“What we’re finding now is that AI is a great way to pick some of the low-hanging fruit out of the day-to-day work that people do,” Finnegan said. 

Finnegan said the Makerspace regularly hosts a range of events, including presentations about responsible use of AI in academia and workshops that familiarize students with the different AI tools emerging in their prospective job fields. These workshops will likely become integral parts of different departments as they begin adopting their own specific versions of AI education, he added.

Many view strides like this as exciting, taking them to mean that the University will be on the cutting edge of the next digital age.

University associate professor of mechanical engineering Jennifer Alexander, however, said she is not convinced. Since ChatGPT publicly launched, she said her students’ writing has shown marked deterioration and similarity, notably becoming far less argumentative in nature. 

“When students hand in something that’s generated by AI, you can see the blandness, the lack of personality,” Alexander said. “The word choice is so impoverished.” 

Fatigued from nearly three years of uninspired AI-generated work, Alexander has reverted to an entirely analog classroom as of this semester. Justly, she said she holds herself to the same standards as her students and refrains from using AI for grading — believing it does them a disservice. 

“If something is written, it ought to be read,” Alexander said. 

Alexander said she doesn’t universally disavow integrating AI into curriculum — acknowledging its usefulness for fields like accounting — but detests the idea that any aspect of her courses should become mechanical exercises. She believes AI grading severs a critical part of the student-teacher relationship, and undercuts the basic principles of liberal education: self-development and the pursuit of knowledge for its own sake.

“I think our sense of education is moving more towards what employers want, right?” Alexander said. “Most people start there, but later in their careers they often have more opportunity, so why not prepare people for that larger vision right away?”

For students who feel passionate about what they’re studying and also use AI to work more efficiently, it’s worth considering where that cost is being cut. 

In a traditional classroom, AI substitutes some steps of the learning processes, whether within ethical guidelines or not. Like smoking cigarettes, its potential to create deficiencies and dependencies may not be apparent until much later.

In this way, the practice of academics implementing AI into their work is ultimately self-defeating, especially for those in humanities. If you’re an aspiring writer, is it worth letting Grammarly keep track of your commas for you? 

AI is undoubtedly here to stay, and the University should continue developing new uses for the technology through programs like the AI Makerspace. However, AI promotes efficiency and optimization, which are virtues of business, not education. 

In a culture where higher education is increasingly seen as a means to an end, if academic institutions want to act in the interest of self-preservation, they may need to take a harder stance on the encroachment of AI into traditional venues for learning.