- Cris
- Jan 23, 2023
- 18 min read
We figured that we'd go big by launching our MTI Blog with a bold declaration from one of our founders, Dr. Cristofer Slotoroff. What follows is a detailed commentary on the changes ChatGPT can and will bring to public education.
Six years into my teaching career, I could see its inevitable conclusion. It was not the students, it was not their parents, it was not my administrators. It wasn’t even the pandemic. It was the stagnated salary, it was the diminished benefits, it was the increasingly anemic pension, it was the existential expectation that I do more with less. It took me another six years to engineer my escape, and it is only now—a bit shy of a year since leaving—that I realize the end of my teaching career was actually about far more than the money.
Last May, after 12 years teaching high school English, I traded my beloved AP Language & Composition courses for a position as a school district administrator. I had never intended to leave the classroom, and while I am sure that many teachers jump to administrative positions with altruistic ambitions, I candidly acquired such virtue only once I’d started at my new job. I recount all of this to, again, make it clear that my students had nothing to do with the choice to go. On the contrary, my students are what kept me in the classroom teaching for so long.
As the educational cliche still so often proves true: over the course of those 12 years, I am genuinely sure that I learned more from my students than they learned from me. When I zoom out and consider why I left teaching, it was more than money. It was because I learned from my students what it takes to be successful in America right now, and it was because I learned from my students the extent to which public education is not preparing them for their postsecondary goals. When one of my academically sound-if-unspectacular students showed me how he’d built his sneaker obsession into an online-only empire of a store selling rare kicks to celebrities, I learned. When one of my most brilliant students rejected her admission to an outrageously prestigious university on the basis of its inevitably commensurate student loan debt, I learned. When one of my introverted, almost altogether silent students showed me how he had become a worldwide innovator in the art of origami, I learned.
The lesson is the same one that other educators are coming to grips with right now: there is a profound misalignment between public education and the postsecondary realities for which the former allegedly prepares the latter. Today, many educators are learning this the hard way, via ChatGPT.
Yes, ChatGPT, the Chat Bot that uses predictive AI to address just about whatever users would like it to address, has rocked public education to its core. New York City Public Schools have already banned it. In my capacity as a school administrator, I have already issued district-wide guidance on the topic on the heels of the grave concerns expressed to me by teachers.
My official message to our district’s teachers was simple: Something like, “Yes, plagiarism is a near-term concern. No, the sky isn’t falling. Let’s start the conversation; make sure your students are in on that.” The faculty reception was, at best, lukewarm. This, I expected.
I’d like to think that I see AI for what it is: a ubiquitous technology that public education curricula has largely ignored, a technology that has governed just about every other aspect of our students’ lives for many years now. ChatGPT can do their homework, but YouTube has already told them what’s worth learning, TikTok has told them what is worth doing, Instagram has told them what is worth publicly cataloging/advertising/purchasing, and Amazon has told them how and when what interesting things worth learning, documenting, wearing, and doing will be delivered.
I write these sentiments not as a curmudgeon, but as a realist and as a pragmatist. This is not a lament; this is not shocking. This…is.
YouTube, TikTok, Instagram, Amazon, and just about every other tech-based company offering content, a platform, or both, have been collectively shaping students’ (and everyone else’s) fundamental view of reality for long enough that our schools should have adapted years ago. It is this dynamic that, truthfully, makes me glad ChatGPT has finally made AI accessible to the extent that schools must confront it. Schools now must not only confront AI itself, but confront what they teach, why they teach it, and how they will teach it. This is long overdue.
When we, educators, study it through a well-calibrated lens, ChatGPT is the catalyst for a necessary shift toward an education that—get this—actually prepares students for the future that awaits them.
I feel obligated to address my assertion’s underlying premise: the extent to which contemporary public education does not prepare its students for what comes next. “After all,” some detractors might suggest, “even though most students will neither be mathematicians, nor biologists, nor historians, nor university humanities professors, American students have been memorizing their times tables, learning the periodic table, reciting the US Presidents, and reading The Great Gatsby for some time. What is so fundamentally different about right now?”
Here is what is so fundamentally different: Speed. Relevance.
There are, and there will be few if any conceivable situations in today’s students’ lives during which they will not be instantly (and soon, predictively) able to conjure the information they are being taught. The conjuring process—today, perhaps by Siri or Alexa; tomorrow, by whom, I do not know—is easy, straightforward, and nearly immediate.
Contrast this straightforwardness with the canonized scholastic processes: the teacher has the information, the information is articulated to students, the students absorb the information through traditional means (e.g., lecture, annotation, presentation, reading, formative assessment), and the student demonstrates a sufficiently competent understanding of the information via summative assessment. This second process is clean, tried and true, classic. It is also slow, generalized, and not individually relevant or responsive to the student.
It is not merely that the speed of the first endeavor vastly eclipses the second, it is that this speed matches the rapidity of contemporary students’ lives. When students’ universes were confined to their school buildings, their sports teams, their clubs, their troupes, their ensembles, their houses of worship, and the friends, books, films, and music they could locally gather, life moved more slowly and deliberately. School, as a construct, could transpire accordingly. Also, students and their schools cannot be divorced from pre-internet life’s formerly—by modern standards—sluggish pace. Today’s curated, algorithm-dictated and infinitely personalized, endlessly refreshable media feeds were once The Nightly News, the local paper, and whatever So-and-So said on the phone last night.
Today, going to school is—too often—students’ daily come-down. It looks and feels and operates much as it did in the 20th century. We too often teach and learn as we once did, and these practices come at the expense of teaching and learning, approximating and adjusting, for what knowledge and skills teachers and students need right now.
Unfortunately, the pace of today’s educational process is at odds with students’ ability to learn outside of it. At school, maybe Wikipedia is blocked on school wifi or, more likely, forbidden by teachers lambasting its use for school assignments? This discounts what discerning teachers and students already know: Wikipedia is, though eternally imperfect in many ways, nonetheless the fastest, most expansive, most responsive, and most inclusive archive of human knowledge ever assembled. In these discussions, I think back to when I was a young kid, when the World Book Encyclopedia sales people dropped off new editions to my house every few years, complete with their gold-leaf binding. I would sit on the floor of our family’s office reading whatever I could to satiate my curiosity, and this was possible only because I was fortunate enough to be born into a family bearing both the means and the desire to have these books in our house. What luck. What privilege. The Encyclopedia Britannica is bound—literally and figuratively—to a moment in time, to editors’ decisions, to a level of nearly infinite exclusion doomed to forever dwarf its expansive, yearly inclusions. Wikipedia only becomes more nuanced, more indicative of more context. Oh yeah, and it’s both free and portable for everyone.
At school, the advanced digital devices and services hiding in students’ pockets, (often school-issued and) perched on students’ desks, are relegated to base tasks that could very well have been performed with a paper and pencil, anyhow. Need to know the symbol for silver on the periodic table? Need to know the quadratic formula? “Don’t google it,” the teacher says. Need a critical analysis of Beowulf? “Don’t look on Sparknotes,” the teacher says. Need to recount the context surrounding the first shots at Fort Sumter? “Don’t go on Wikipedia,” the teacher says. Both students and their teachers know the information is there, though, and the information is dispersed concisely, quickly, and clearly. There will be accompanying links; there will be accompanying videos. Too bad; it’s off-limits.
Students are still, en masse, being asked to take notes on teacher-prepared slide decks. Students are still, en masse, being asked to provide a summary of the chapter that they read. Students are still, en masse, being asked to annotate the novel.
Why?
This is where relevance enters the discussion. If speed, and correspondingly, access make today’s educational process too slow, why has the content and knowledge lost its relevance? Enter, again, ChatGPT: the great equalizer.
It is not as though rote knowledge and skill mastery are, in and of themselves, irrelevant. Hardly. It is that a much better educational question is this:
How important, how useful is the content or skill in question, to the individual student? Alternatively, at the very least, how important is it to these students in this class? ChatGPT suggests that the value and utility of students’ application of their own ingenuity far surpasses that of their capacity to accumulate and retain the knowledge and skills that used to serve as keys unlocking the great postsecondary doors. Well, ChatGPT has changed the locks.
Picture it this way: ChatGPT is, literally, a human synthesizer. As the piano to the synthesizer, ChatGPT (and the AI-based technologies like it) does not devalue its antecedent; ChatGPT offers a reframing and reappraisal for what knowledge and skills are valuable for success.
A brief, illustrative digression: When musical synthesizers (and, particularly, digital synthesizers) reached a level of quality that faithfully replicated the Baldwins and Steinways of the world, the piano did not become worthless. The piano became, like lots of other technological advancements, more of a niche item: a beautiful instance of splendor, yes, but also a suddenly gigantic, difficult-to-transport, expensive-to-maintain, singularly focused instrument nonetheless capable of individual beauty and expression. Synthesizers, though often cheap, are endlessly versatile. On my desk right now is a cheap synth with a nearly full manual that plugs into my computer’s USB port. It weighs less than a couple of pounds, it makes no sound on its own, but when it is plugged in—to just about any computer running Logic Pro, ProTools, et al.—it faithfully (and with striking responsivity to the velocity at which I strike the keys,) recreates from my speakers everything from yesteryear’s keyboards (e.g., Fender Rhodes pianos, Wurlitzer pianos, Hammond B3 electric organs), to massive cathedral organs and the grand Steinways, to entire sections of strings and horns, to customizable electronic and acoustic guitars, drums, basses, and anything else I can imagine.
The point here is that while the piano did not change, the piano’s parameters changed: a vast expanse of unmined audio territory revealed itself beyond the piano’s confines. Truly innovative piano playing remains truly innovative. Pianos remain awesome. Still, it is worth asking as time passes: how many people are going to have access to pianos as opposed to cheap-and-convincing synths? Well, eventually, I suppose this number will grow to be similar to the number of those currently capable of accessing a spinet or a harpsichord. Virtuosity takes on many forms, and I will concede that, to this day, no two piano players are alike. Virtuosity remains impressive. Virtuosity, as a recognition of mastery, however, is not unprecedented. Ingenuity is by definition unprecedented. Should we educate students toward mastery or ingenuity? Lately, I see a lot more value in the latter. We do not forbid young musicians from using synthesizers.
Now that a ChatBot can synthesize human–like output according to user demands, and now that it can do so like my desktop keyboard synthesizes a Steinway, should we stop thinking, reading, and writing? No. We retained the piano, we retained those skills. We used the piano to build a synthesizer. We teach aspiring musicians to play the piano and (or) the synthesizer, and we see what they can do with either one. We use AI, like ChatGPT, to better educate, to better create, to better solve problems. Mastery is impressive. Innovation is useful.
Essentially, now in school we must ask: What can the student do? What can the student fashion, contribute, synthesize, craft, create, aggregate? Bloom knew the value of creation a century ago, and most teachers know synthesis to be the greatest demonstration of understanding at this moment. Unfortunately, plenty of teachers who subscribe to this notion are, more often than not, shackled by institutional mandates designed to minimize or reject it. Of course, to manufacture an original work, one needs underlying skills. Yes. This remains true and will do so in perpetuity. The point is how seldom schools ask students for sincerity; this virtue is asked of students in marginal contexts, in extra credit work or independent studies. We need to, at the institutional level, start asking students to solve problems with ChatGPT. This is a kind of information literacy that is urgent, responsive, and relevant. It is, by its very nature, individually tailored to the user’s input.
Incidentally, this kind of education is also way more interesting to students.
“Doing interesting stuff” is the secret sauce. This is (and has always been) the simple educational solution to both ChatGPT’s and previous technologies’ educational incursions. This is (and has always been) what companies’ CEOs actually mean (as opposed to the ineloquent laundry lists they usually espouse, like “MS Office Skills” or “Working on a Team,”) when they say that they want students who are ready for the workforce. This is what colleges ought to mean when they depict the sort of students fit for rigorous study at their institutions.
Curricula, assignments, and teachers can simply be more interesting and more relevant.
Let me offer a small, more concrete example of the kind of shift that schools ought to make:
When I was an English teacher, I was no revolutionary, but I did have a good sense of what was interesting, what got students’ attention, what got them talking, and what motivated them to engage the learning process. It was not long into my teaching career that I stopped asking “What color shirt was [insert novel protagonist] wearing on page 84?” questions. When I realized all of the answers were online, I did not pivot to asking more obscure recall-based questions. I started asking (more specific versions of) questions like, “given the main character’s experiences in the chapter, detail the extent to which you can rationalize these experiences using evidence from your own life. Then, decide how you would react based on equivalent circumstances.”
Is it possible for students to “cheat” on this question using ChatGPT?
Yes, but only under three conditions:
1) The student has access to ChatGPT.
2) The teacher considers a student’s use of ChatGPT as cheating.
3) Most importantly: The student feels that using ChatGPT would be a more rewarding experience than writing a sincere answer to the question.
Let’s break these conditions down.
1) Condition One is certainly manageable in the classroom. Outside of the classroom, however, it is not the Wild West. I am convinced that students are, consciously or not, inside or outside of their classrooms, making deeply informed decisions about whether or not to use ChatGPT on wholly teacher-manageable indices. Foremost among said indices is this: how much the student values the class.
Teachers should know that a student’s decision to use or not use ChatGPT is, 99 out of 100 times, not at all predicated on malevolent “cheating” or “getting one over on the teacher.” It’s simply a matter of economics, decided and dealt exclusively in the two currencies to which students have some measure of agency: their time and their grades. When students feel like their assignments are not worth their time, they get ChatGPT to do their assignments. When students feel like ChatGPT gives students a better shot at a good grade than their own efforts, they get ChatGPT to do their assignments. Solution: Design assignments that respect and respond to students’ assessments of their availability, and make the grading process equally thoughtful, respectful, and responsive to students’ input.
2) Condition Two is a matter of philosophy. If the teacher is trying to educate a student to succeed in the world beyond K–12 public education, more than a little bit of the teacher’s version of postsecondary “success” has to do with “submitting satisfactory work.” Students who are chiefly concerned with “submitting satisfactory work” (as opposed to, say, “students who are genuinely stimulated, sincerely curious about, and truly interested in the nature of their work”) are finding the most direct route to the goal. Is this wrong? I do not think so, so long as both the teacher and student are aware of the assignment’s parameters.
If the teacher does not consider a student’s use of ChatGPT as cheating, and the student uses ChatGPT to answer the question I outlined above, then the teacher has a host of new, more individualized, and—I would argue—more educational avenues for guiding further learning. In fact, the student’s use of ChatGPT is teaching and learning in action. Here are a few:
a) Perhaps the student was burdened with other work and could not be bothered to devote ample time to the question?
This is an opportunity for the teacher to have a dialogue with colleagues regarding assignments’ scheduling. Is there a cross-curricular deluge of tasks overwhelming your students?
b) Perhaps the student did not feel that the weight of the assignment was worth the effort the teacher is asking of the student?
This does not suggest anything deficient in the student. This strikes at the fairness of the assignment, and it even suggests that the student is a discerning time manager. Sure, the traditionally virtuous action is to answer the question faithfully and on one’s own, but the question of its effort remains, to me at least, essential. This isn’t unlike most human behavior: if one is hungry, and one owns a car, and one is two miles away from a beloved restaurant, despite one’s awareness of jogging’s health benefits, one will likely elect to drive to the restaurant. Jogging to the restaurant, despite its clear advantages, is not perceived to be commensurate with its requisite effort. Why should discerning students be castigated for deploying incisive logic?
c) Perhaps the student used ChatGPT merely to get a sense of how assignments of this nature ought to be conveyed in terms of tone and syntax?
This sounds an awful lot like research or due diligence to me. It is a fantastic impulse to get a sense of a task’s norms before drafting, is it not? If what ChatGPT compiles is similar to what a student intended, then the student knows what “normal” is. What the student does with that knowledge, however, is the learning opportunity.
d) Perhaps the student has neither accumulated experiences in number nor in value to the extent that the student has an ample basis for faithfully answering the question?
What a great premise for exploring the value of one’s experiences: are our experiences notable because they are unique or because they are universal? What a great premise for teaching students to demonstrate the meaningfulness, applicability, or character of their experiences in relation to those of fictional characters. What a great premise for exploring fictionalized experiences versus those in real life.
The permutations and the opportunities here are endless. ChatGPT can launch these conversations, teachable moments, tasks, projects, and indeed curricula if we treat it as one of many tools reflective of those that students must master to attain a measure of success in the ensuing decades.
3) Now, finally, we move to Condition Three, which is the real crux of the issue.
In education, there is perhaps no truth more obvious and ignored than this one: people tend to learn better when the learning process’s usefulness and its enjoyability are self-evident. I took Latin classes in high school and I took Spanish classes in middle school, high school, and college, and I am not proficient in either language. I taught myself to speak and read pretty decent French a few years ago, however, when I had:
i) an enjoyable means of doing so (by way of Duolingo, which continues to innovate, impress, and be offered free-of-charge-or-close-to-it to educational institutions, while it also manages to go unadopted by scores of school districts), and
ii) an actual reason for doing so (reading The Stranger, a book that I like, in its mother tongue).
For some reason, we collectively forgot, or worse, neglected—amid perpetual educational woes fixated on the standards, skills, benchmarks, assessment of learning, professional development, continuing education, the certifications, and the graduate programs—to ask:
Is this fun? Is there a direct payoff?
ChatGPT can inject a great deal of fun into education, but so can lots of other innovations. We can still teach novels, of course, but we need to point them in the direction of something tangible. Yes, concepts like “delayed gratification” and “studying for study’s sake” are virtuous acquisitions for school-aged children, but at what cost?
Schools need ChatGPT. Schools need a robust curriculum that teaches students how AI works, the extent to which it is embedded in so much of what they consume, so that students can properly assess, deploy, and innovate with ample awareness of the technology’s value and its ubiquity.
Schools need ChatGPT so that students can take advantage of it. Schools need to move faster, they need to be more relevant, and they need to center—rather than feature—ingenuity and project-based learning in their curricula.
Finally, schools need to go a damn long way to be a lot more enjoyable.
If schools are not, they will fully realize what many of them are already becoming: a mere interruption to the rapid learning that students are doing outside of class, on their phones and computers, every day of every week. You think we can’t control ChatGPT? You’re right. Just like YouTube, Instagram, TikTok, and every other AI-laden platform on which students generate and soak up information by the gigabyte, we cannot control ChatGPT.
We can, however, control what we decide is important for our students to learn in school. ChatGPT is only a problem if we consider it so. Remember: students do not consider it a problem. Adults do.
Each of these conditions and their details speaks not to students’ deficiencies, but to their assets. If they speak to anyone’s deficiencies, they speak to those of educators like me.
Whatever we, educators, do about ChatGPT, I hope that we do not take ChatGPT or any other tech-based moral panic (e.g., AOL Instant Messenger, MySpace, Snapchat) as an indictment of the modern student. Take a moment and remember that students did not ask for the reality distributed to them by their (now, often, as-or-more tech-indulgent) adult predecessors. Many a concerned thinker has typed many an alarmist think piece (A Nation At Risk, anyone?) lambasting the contemporary state of American public education versus some vague and romanticized semblance of Back-in-My-Day. Chief among collective concerns is the perception that today’s students are somehow less capable, somehow less equipped for “The Real World™” that awaits them. ChatGPT is only the latest addition to this hogwash.
Look, I was there. I am there. I am in these kids’ classrooms. I am approving their curricula and talking to their teachers. Today’s students will not only be fine in an educational paradigm inclusive of ChatGPT, I am sure that they’ll do a whole lot better than we did, and that is because, generationally speaking in the broadest sense, today’s students have it a lot harder than we did. Yes, I mean this.
When I write “we,” I mean millennials like me (I’m 35), and I especially mean Gen Xers, and I even more especially and most accurately mean Baby Boomers. This is not a popular take, but it remains my conviction. Hear me out.
Over the course of my career, I was most often impressed with my students’ resourcefulness, their resilience, their creativity, their entrepreneurial senses, and their worldliness. What too many education pundits fail to see is that such senses coalesced under unprecedented existential duress. I remain impressed by the students I continually meet because these existential prospects—taken, en masse, to be a given—look to me, as an older millennial, akin to staring down a barrel.
When I graduated high school in 2005, things looked pretty rosy: the forecast of American life was, generally, rather positive. “Going to college” was, though increasingly astronomical in cost, widely perceived as an irrefutably good move. For most of my pre-Columbine childhood, school was a safe place. For most of my pre-9/11 childhood, the prospect of terroristic disaster was nonexistent. A house was expensive, but it paid off over time. The stock market performed rationally, on a gently upward trajectory, based on canonized and demonstrable principles. Paper money was the principal representation of value and currency. There were essentially no collective memories of an internationally disruptive pandemic. Though most politicians were, as ever, loathsome, decorum at least required the façade of competency.
And now?
For most high school students, here are the rules of the game they enter: without a substantial scholarship, a parent’s help, or a lifetime of predatory-interest-laden and federally (or, worse, privately) issued student loan debt, college is unquestionably unaffordable for just about everyone who isn’t unfathomably loaded. One’s education will cost as much or more than home ownership, a formerly unquestionable rite of passage toward self-sustenance fast becoming scarcer and equally unaffordable. One’s career need not be based on any passion or sincere interest, it must pay off one’s education…over a lifetime, not a few years. One’s employer must provide health insurance, and—even with one’s health insurance in good standing—one still runs the risk of getting hurt one day and incurring a surprise six-figure medical bill. One will not be able to take a vacation, one’s occupational retirement provisions—if these are even included in one’s job description—will be increasingly meager, and if one’s job is not inevitably automated, it can be cut without notice, cause, or context…as the company rewards its top brass.
Look, I know that there are exceptions to all of this. I am talking about widespread perceptions among students, gleaned over the course of over a decade teaching ambitious high school kids. It’s enough to make one’s skin crawl, and yet today’s teens remain—with massive anxiety, trepidation, and fear in tow—somehow, undaunted. They are courageous; they press on with the same cold hearted pragmatism that society foisted upon them. Like I said: students are learning.
To most of the students I meet, these conditions are unfortunate yet immovable. “Yes,” society has told them, “you might be a mass shooting casualty while attending a concert. No, no one is going to do anything about that.” “Yes, you will likely be in debt forever. No, no one is going to do anything about that.” “Yes, you will do more work for less money around the clock than your parents and grandparents did. No, no one is going to change that.” “Yes, you will likely have to work until you’re in your sixties, and yes, it might get taken from you for no reason. No, no one is going to do anything about that.” “Yes, when global health emergencies arise, these will be needlessly exploited, disinformation-laden, and mined for advantages. Scores of people will needlessly die. People will suddenly pretend to care about your education in the process. There will be no consequences for bad actors. No, no one will help you.” There will be no New Deal. There will be Deal or No Deal. Deal or Be Dealt.
Grim? Sure. Like ChatGPT, this is what today’s students were given. Astoundingly, this collective distrust in many of the institutions previous generations of Americans traditionally held dear—government, school, banks, and now, education—spanned demographics, backgrounds, and political affiliations. Actions have consequences. I have no solutions for many of these institutional breakdowns, but I do have one for education.
Educators: Let’s take ChatGPT as an opportunity. It is an opportunity to avoid giving students one last reason—i.e., my school won’t let me use this instantly accessible tool to instantly answer just about whatever question is asked of me—to write off school as just another irrelevant series of events interrupting their inevitable struggles to survive. Let’s take ChatGPT as a new avenue to address these struggles, to solve these struggles. Let’s use ChatGPT to cultivate and essentialize students’ ingenuity. I have faith in today’s students. When educators ignore, restrict, and especially when they block access to ChatGPT, they are telling their students the opposite. They are eroding an already diminished trust, and what comes from students’ collective distrust in public education is too frightening to imagine.
01-23-2022