No, Teachers Are Not in a Panic About ChatGPT
But they can use help rewarding students’ authentic work
"Donkeys live a long time,” Benjamin tells the other animals in Animal Farm, suggesting he’s seen it all. Teach long enough, and you can feel like Benjamin: the latest pedagogical fad is all-too-familiar, the newest tech seems just like a different way of performing an old trick.
Much of the media coverage of an AI prototype that was launched in November focuses on a new wave of plagiarism it likely presents for educators. While the capabilities of ChatGPT may be astounding, the fundamental problem — academic dishonesty — is older than any teacher.
Analog forms of cheating dominated my early teaching years, amid another wave of media attention on the problem in K-12 schools and colleges. A cheat sheet pasted onto a water bottle in 2003 caused a scandal in my district that made it into The New York Times. Worries about academic dishonesty escalated with the release of polls showing cheating had become common.
In my own high school days, paperback CliffsNotes, bound in caution-tape black and yellow, had already been available for decades. The one time I anxiously sought them out, I blamed Shakespeare, then snuck downtown to Woolworths to spin the metal display rack of “notes” to see if I’d get lucky and find the one for the play that was giving me a headache.
The development of the internet and the proliferation of smartphones, tablets, and laptops in schools and homes, of course, made all kinds of cheating easier and thus more tempting. CliffsNotes went online and other, often free electronic “study guides” such as SparkNotes, launched in 1999, were banes of my existence as an English teacher. Over my career, more students began replacing reading and thinking about books with skimming the summaries and analysis on these sites.
There were worse and better ways to handle this challenge to literature instruction. Some schools abandoned full-length texts for a focus on building skills through analysis of passages and excerpts. Teachers who never gave reading quizzes started doing so (guilty as charged!); others revamped theirs, making them harder to pass if a student had read only the study guide. Solutions like these, though, tend to punish all students.
Providing students more time to read in class, including less common or more contemporary texts on syllabi, assigning fewer books but delving into each more deeply, giving students more choice in the books they study are just a few of the sounder ways literature teachers have sought to disincentivize cheating by incentivizing reading.
When it comes to writing, once again, new technologies have presented new temptations while traditional ones endure. Students still plagiarize others’ work; parents still write papers for them; and test-takers’ eyes still wander. Just a few years ago in my last high school, a group of students entered the English and History department offices after hours, searched for, found, photographed, and distributed pics of midterm exams, echoing a 1970s film set on a 1950s college campus — sans the hilarity.
Meanwhile, the pedagogical advantages of tech are often a double-edged sword. Shareable electronic documents made it easier for me to provide quicker, fuller feedback on essay drafts but also for tutors to meddle from afar — a colleague and I once opened our student’s document to provide feedback on an outline only to catch a tutor in the act of writing it while the student sat in math class.
Too often, it seems, the focus is on either the pedagogical opportunities or the challenges of new technologies. So we see some bans on ChatGPT while others herald it as a newly invaluable teaching tool. The media has largely been presenting the AI chatbot as an unprecedented threat, one that has teachers in a panic and that will fundamentally change the way writing is taught.
Meanwhile, many of these pieces propose solutions that are often already in widespread use by educators as methods of preventing cheating, such as requiring that essays be written by hand and/or in class. It’s inaccurate and insulting to suggest teachers aren’t already battling plagiarism daily and that they simply haven’t been bothered until now to create assignments that require authenticity and the deep thought only humans can produce.
The emergence of ChatGPT should be focusing us on the evergreen challenge of academic dishonesty. Journalists and pundits can help by shifting the conversation toward the underlying causes and effects, problems and solutions of cheating and plagiarism. Regardless of mode or manner, cheating deprives students of learning skills and building knowledge; it tilts the educational playing field, contributing among other things to grade inflation; it sucks up untold teacher time and energy as they try to prevent it or provide accountability; it contributes to an icky culture of cynical expediency and empty credentialing.
School administration and parents can help teachers deal with ChatGPT, not by hyper-focusing on ChatGPT but by recommitting to maintaining both academic integrity and improving instruction, regardless of the technological landscape.
Because students are more likely to cheat when they don’t care about the work they are asked to do, schools need to provide teachers with the freedom, the time and the training to develop more of the kinds of activities, assignments, and assessments that engage students and provide for deep learning.
As John Warner has argued (and I highly recommend you read all he’s written on the topic), K-12 students are too often disincentivized to write because K-12 teachers are too often disincentivized to teach writing in ways other than those that might be seen as traditional or rigorous but can be cart-before-the-horse assessments designed to be easily scored or graded. Thank businessmen-with-outsized-influence-in-education such as College Board’s David Coleman for helping to divert professional development away from the teaching of writing with voice and from the genres of personal writing and opinion writing, genres my students tended to be enthusiastic about because they were able to tap into their unique human experience.
Schools can also provide timely, realistic professional development around new technologies rather than rely on techno-fixes like disabling access to websites on campus or plagiarism detectors that respond inadequately to technology with yet another technology.
Some need to move beyond publishing honor codes to create or bolster meaningful districtwide, age-appropriate programs around academic integrity. When I taught middle school, teachers worked across disciplines with students to help them understand why cheating mattered and how it wasn’t a victimless offense; we worked with them on case studies to help them understand what was and wasn’t acceptable behavior in collaborative work, homework, test taking, and essay writing. This had multiple benefits, including increased consistency among teachers and reduced stress among honest students.
Schools can provide and follow streamlined processes for dealing with cases of suspected cheating and plagiarism that are fair to students but hold them accountable when they are found to have cheated. Cumbersome processes and administrators who cave to intimidating parents can lead teachers to avoid reporting cases of suspected cheating, encouraging more students to take the risk.
Parents can talk to their children about expectations for academic honesty, making explicitly clear that they value learning more than grades, honesty more than hollow achievements. They can examine the implicit messages they might be sending their children about what it is about their education that is most important. They can be clear with tutors about their desire for their child to learn, not have their thinking and learning replaced and subverted. They can accept and reinforce reasonable consequences for a child who has been shown to have cheated, understanding that while most students who cheat do so rarely, for some it becomes a habit that once engrained is hard to break.
On message boards, in Facebook groups, and on blogs, I see few teachers apoplectic about AI. Instead I see many worried about its potential impact but deeply interested in understanding what the software can and can’t do. That most view it as another curveball that’s been tossed their way underpins my confidence that they will find ways to limit its abuses. With the help of other stakeholders, they can also respond in ways that improve teaching and learning.