The only thing I ever wanted to be was smart. College was the place to accomplish that. Learning how to learn was the pathway to unlimited capability. To borrow a phrase from a colleague, what is the sine qua non of college? Is it not thinking? Is it not learning how to analyze, how to take existing information and create new?
With the unapologetic embracing of AI as a legitimate learning tool, we are eviscerating the notion that college is actually higher education.
The argument made ad nauseam by supporters to embed AI into higher education—”They need to know how to use it when they leave here”—is antithetical to the point of college.
I keep reading that employers want the skills of reading and writing, critical thinking, and creative problem solving. Our response is to train them to use a search engine. AI is a technology with no expertise requirement. I am not suggesting there is no place for AI post-college. In fact, I am saying it is the only place for it.
You need to have those critical skills well developed to even understand how to use AI as a tool. The top of the class who leave with academic skills will flourish, but ultimately, the use of AI in this critical period will create a small-thinking class and a very large unthinking class—all with the same degree.
Why do I have such disappointment in the current higher education zeitgeist regarding AI? Maybe I just wanted to see this question somewhere in the last few years as the pitch for AI training echoed through the webinar ecosystem: “Why do students struggle to formulate questions of depth or specificity?”
If we focused on teaching students how to break down writing, interpret meaning, recognize varied perspectives, and find creative understanding, then they would be able to write a meaningful prompt. Worse yet is the advocation to let AI write the papers for them and allow students to act as editors.
Rhetorically, if students cannot even grasp the concept of writing a meaningful phrase or sentence to use ChatGPT, what reality has them qualified to be editors who can meaningfully critique writing? The concept literally makes no sense.
Artificial intelligence is not even trying to hide its shortcomings—it’s in the name. Digital learning is woefully inferior as a learning tool. AI is digital learning on steroids. College should be about developing the brain, not working against its basic physiology. Using AI to summarize an article, create an outline for a paper, write code, or generate paper topics offloads work from your brain.
At a fundamental level, learning and memory involve the creation of neural pathways across different areas of the brain. The neural networks of learning rely on usage. The brain is efficient; it prunes the networks you do not use. It is pretty hopeful to think the brain will keep dormant networks. It is further wishful to think those more complex processing skills will just come back—not everything is just tying a shoe.
Research has reliably shown that reading off of a screen yields less comprehension and understanding compared to a physical book, and the gold standard for note-taking remains pen and paper over typing, over reading a transcript. The incorporation of the tactile—engaging multiple parts of the brain—is the widespread theory on the power of the old. AI and a digital future throw all of that away. Again, where is the questioning from higher education?
A terribly overused cliché in academia is “data-based decision making.” The aforementioned data on learning and neural mechanisms is not new. Neither are the ever-declining results of reading and math testing that place the United States at the bottom of the industrialized world. That is the data.
What decisions are we making from it? Adopt a technology that does not and cannot expand students intellectually? AI is still, by objective measures, not a good product. It is still plagued by hallucinations, incorporating data from discredited sources—and many articles read for this editorial used the phrase “often inaccurate.”
A study published in February showed ChatGPT misdiagnosed more than half of the case studies examined. Still, AI is trumpeted by industry insiders and acolytes alike as superior. Data-informed indeed.
We need to embrace the primacy of teaching, of educating. We are the vital component, not an algorithm. If people could teach themselves … I would think the lure of keeping a job would have primacy over getting home a few minutes earlier.
In closing, why are the most educated people not displaying critical thinking? We read, “It is here, we cannot stop it. We should learn to work with it.” That is a weak capitulation. We know it is bad for students; it is bad for educators. In 1964, Marshall McLuhan stated that “the medium is the message.” It remains true with every technological advance. Today the AI message is “do not worry, I can think for you.” Good luck to us.
