To say the demands of teaching have expanded considerably in just the past several years is an understatement. I taught high school English for 34 years before retiring in June. I am currently completing my second semester of teaching introductory composition at the Middletown campus of the Connecticut State Community College system. Whether in high school or at the college level, the role of teachers is approaching a crisis point.
At a time when digital resources as well as digital distractions are ubiquitous, student curiosity for learning – and for the printed word, in particular – has diminished. In short, my students’ overall interest in reading and writing has declined dramatically. The emergence of large language models (LLMs) like ChatGPT a mere three years ago has only exacerbated that indifference.
A typical homework assignment in my college class includes reading a nonfiction article of up to 10 pages and then completing a template I created called a “Reader Response.” Among the required tasks is to identify and define essential terms, to write an original paragraph relating the article to current times, and to write another original paragraph describing a personal connection. I’ve quickly discovered that the writing style of a number of these Reader Responses from students does not match the style of the handwritten work they composed in class on day one.
Gee, what could have happened? In a word, “ChatGPT.”
As an experiment, I copied and pasted my Reader Response template into ChatGPT’s prompt along with the words “Letter from Birmingham Jail,” the classic Martin Luther King composition students read for homework. The two samples ChatGPT produced were shockingly on point. In fact, both of them concluded by offering more: “If you’d like,” commented ChatGPT in its oh-so-friendly tone, “I can help you turn this into a polished essay, add citations, or rewrite it in a more formal academic tone.”
What honest, self-respecting college student would actually take the low road and turn in a plagiarized research paper via ChatGPT? Turns out, quite a few.
According to a survey of college administrators by the American Association of Colleges and Universities and Elon University, “Some 59 percent said cheating has increased since generative AI tools have become widely available, with 21 percent noting a significant increase – and 54 percent do not think their institution’s faculty are effective in recognizing generative Al-created content.”
Unfortunately, cheating is just one of the troubling issues with artificial intelligence in the classroom. The way AI inhibits – nay, kidnaps – human thinking is another. For example, LLMs enable students to bypass the entire process of preparation when writing. There’s no need to brainstorm ideas, research sources, or organize an essay because AI does all of that for the student. And then, of course, it also writes the essay.
Artificial intelligence, simply, takes the onus for learning away from the student. Even using Google searches is more educational than AI.
“When people rely on large language models to summarize information on a topic for them, they tend to develop shallower knowledge about it compared to learning through a standard Google search,” according to a new paper by professors from New Mexico State University and the University of Pennsylvania. “One of the most fundamental principles of skill development is that people learn best when they are actively engaged with the material they are trying to learn.”
“When we learn about a topic through Google search,” adds the paper, “we face much more ‘friction’: We must navigate different web links, read informational sources, and interpret and synthesize them ourselves.”
Perhaps the most worrisome factor with artificial intelligence is how platforms produce “hallucinations,” or information that is misleading, biased, or simply wrong.
“Large language models such as those that underpin OpenAI’s popular ChatGPT platform are prone to confidently spouting factually incorrect statements,” reports the journal Science. “These blips are often attributed to bad input data, but in a preprint posted [in September], a team from OpenAI and the Georgia Institute of Technology proves that even with flawless training data, LLMs can never be all-knowing – in part because some questions are just inherently unanswerable.”
So what’s a teacher to do? Many colleges and schools in Connecticut are embracing AI as a learning tool – including Yale University and New Haven public schools – which see it as a resource that can aid student learning. But even the students are leery.
“I do think I see an alarming trend of students starting to rely on AI,” said first-year Yale student Tiffany Sadiq, “and it’s not producing the minds that we need to build our society up.”
Adds Abdella Aly, a junior at Hill Regional Career High School, “People have used it for the right things, and I think a lot of people use it for the wrong things.”
The situation presents a colossal conundrum for teachers. Artificial Intelligence is not going anywhere, and the many educators who have responded with blithe acceptance are only asking for trouble. Retired high school teacher and education writer Peter Greene does not mince words regarding the danger: “I am daily frustrated by people who have fallen into the notion that AI chatbots are somehow paragons of objectivity that scan the web for the best advice, weigh it logically, and deliver an unadulterated evaluation of the sum total of human knowledge on whatever you have asked. They aren’t, and they don’t.”
As for my own classes, I’m still adjusting. On occasion, I’ll give a pencil-and-paper pop quiz based on the homework reading to keep the students honest. Also, I’ve assigned on-demand, in-class written responses that students must compose by hand.
If it seems like I’m going “old-school,” it’s because I am. Readers of a certain age might remember the classic blue books used for handwritten college exams. Blue books are now making a comeback as one way to stave off student plagiarism. But it’s merely a stopgap at this point. Moreover, the problems artificial intelligence creates go well beyond the world of education – the tip of a societal iceberg, you might say. And we’re all watching from the boat deck of the Titanic.
