The End of Education as We Know It? (Part 1 of 2)

The End of Education as We Know It? (Part 1 of 2)
AI's impact on education is reminiscent of the Trojan horse from Greek mythology, because there are hidden dangers underneath its appealing veneer. (Kaspars Grinvalds/Shutterstock)
3/28/2023
Updated:
3/28/2023

It’s times like these that try men’s and women’s souls—in education, particularly. And the times just got a whole lot more trying.

Two-plus years of unparalleled COVID-19 pandemic disruptions, rancorous rows over who gets to decide curricula and the placement of critical race theory, and increasing student incivility have been enough to send many a veteran teacher packing for early retirement. Those newer to the profession are swimming in rougher waters than previous generations charted or perhaps could have imagined. There’s now a national shortage of teachers and support staff to show for it.

On the bright side, at the classroom level, things were just starting to gain some semblance of normality—or so most teachers I’ve spoken with were reporting at the start of the year. Sure, many a student was showing the lingering effects of pandemic-era purgatory, having fallen behind in basics such as reading and writing after months squirreled away at home, learning online. But things were back in full swing—in person, thank goodness.

And then the kraken from “Clash of the Titans“ was released.

Or perhaps, if we’re trafficking in ancient Greek memes, we should say the Trojan horse. Because in this case, the monster that’s been unleashed upon educators (and parents, too, take note!) hardly comes with 10-inch teeth and scaly serpentine skin.

It’s called ChatGPT, and if anything, it’s being heralded with something sounding more like a hallelujah chorus in some circles, at least in Silicon Valley. The mere mention of the name among educational administrators and staff, by contrast, evokes dread and despair. It seems to be about the one thing everyone in education can agree upon at the moment.

What Is ChatGPT?

Books on Amazon herald the revolutionary power of the tool, with titles such as “The ChatGPT Millionaire,” which promises in its subtitle that “Making Money Online has never been this EASY.” Other works vouch that it can write nonfiction for you, that it could teach you to “crush” job interviews, and even that it holds “The Key to the New Future of Medicine.”

Its own creators don’t package it with so much hyperbole (they’re computer engineers, after all). But they do list a pretty impressive bevy of powers the new online AI tool offers. Among them are correcting grammar, translating text, summarizing notes, generating creative copy, solving math problems, debugging code, and even writing speeches and letters.

All this shouldn’t be so surprising, given the blistering pace that Artificial Intelligence (AI) has been progressing at in recent years. It was but a little more than two decades ago that an AI computer, Deep Blue, defeated the world’s reigning chess champion. Just seven years ago, it was the world Go champion’s turn to leave the arena in defeat.

But what has teachers staying up at night in cold sweats is what ChatGPT can do for students. And do eerily well. Things, mind you, that nobody should be doing.

Such as cheating.

Think of ChatGPT as cheating 2.0. The barrier to entry just got lowered a whole lot for would-be delinquents.

No longer does the wayward 9th grader have to beg his older brother for that dusty copy of his English essay analyzing “The Grapes of Wrath.” Nor does the only child cheater have to dole out big bucks for online essay services and wait—heaven forbid—a full day or two for the illicit rewards.

Now it’s down to less than a minute. And it’s free.

The only limit is the kid’s creativity.

“Write me a five-paragraph essay comparing Macbeth and Hamlet,” he asks, and moments later it’s populating the screen—with perfect grammar, a possible quote or two, and alarmingly good logic.

What, no citations? You can ask for those too.

Whoops, the paper looks a little too good to be true? “Rewrite the essay below at an 8th-grade level,” you prod. And it will. Startlingly well, in fact.

It’s not perfect, however, as any teacher who’s given it a spin will attest to. Sometimes its references are off, its facts a little garbled, and its points not quite valid. But it’s hard to tell, and most likely any student resorting to using it isn’t about to pick up on the output’s shortcomings. Or care, when the assignment is due in a matter of hours and they haven’t read the 400-page book.

The scariest thing is that it’s not very traceable (although that may change, according to its makers, OpenAI). Each prompt that’s given to the uber chat machine generates output that’s truly unique, in real time. No two essays it whips up are going to be the same. Nor will they have ever existed anywhere online or in print before. It’s not savvy plagiarism. It’s genuinely new.

Only it’s not yours. It isn’t the product of your own mind and effort. And that’s the whole point of school assignments in the first place.

All of which brings us to the matter of consequences.

Changing Education

Unlike the fictional kraken of “Clash of the Titans” fame, this beast wasn’t unleashed with malevolent intent. (It’s about revenue, of course.) But much like the creature of lore, its every move sends waves big and fierce enough to crush the unsuspecting citizens at its shores.

In practical terms, I think ChatGPT is most likely going to play out in several ways on the educational front.

First, its effects are apt to be inequitable. One can readily imagine the achievement gap only worsening in its throes.

This will partly come down to which schools have the bandwidth for combating essay-writing, math-solving AI tools of its ilk. Yes, trained teachers can spot a darned good deep-fake—if circumstances allow for it. You first have to know each of your students well enough to know their baseline performance, their voice, their command of the language, and so on to know when something is written in another hand. That’s exponentially harder to do when you’re teaching in a public school district with swelling class sizes and a teacher shortage. It’s one thing to know each of your 40 students well, but it’s another to say the same for 125 or more. With a stack of 100-plus essays going into the weekend to grade, simple efficiencies dictate that you can’t treat each with the same degree of nuance and attention that the teacher in the same subject at, say, an elite boarding school might with a stack of papers one-third the size.

Similarly, teacher abilities aside, some schools are much more likely to have the resources at their disposal to purchase the kind of plagiarism-detecting software needed to spot a ChatGPT-type fake.

(For anyone pining for the Looney Tunes theatrics of their youth, you can now tune in to the cat-and-mouse game that’s emerging in the AI arena—with AI software tools popping up left and right to combat the fake essays being authored by the sibling AI software across town. Oh, the irony of it all.)

We might also imagine the impact playing out with great discrepancy across different demographics of households and communities. A good deal of the onus to thwart naughty student doings is going to fall on parents, and for single parents or those working double jobs who lack the extra time and energy to carefully watch a child’s online adventures during homework time, bad things are prone to happen.

By extension, I might envision character education playing a bigger role in schools with all this (if there’s any silver lining to this artificial cloud!). Never has it been more imperative for students to cultivate the kind of self-awareness, discipline, ethical judgment, and self-control that were once such a big part of education and a proper upbringing. Think of it as ChatGPT’s implied moral imperative, if you will.

Those who ignore it do so at their own peril. The cost of cutting corners on essays, research assignments, and math tasks done in the dark will be steep and painful come in-class exam time. But by end-of-term, the damage may already be done. That time on task can never be recouped. Hence, a widening achievement gap is likely.

We might also anticipate—contrary to what many might expect—that it may well be college students who suffer the most from the quick-fix tool. Sure, they’re more mature than their younger adolescent selves. But these students will be the ones with the most unsupervised time, as the bulk of college-level work (essays, research, and so forth) is done on one’s own. Savvy teachers and administrations at the elementary and secondary levels might be able to pivot quickly and opt for more of a “flipped classroom” approach, where the bulk of work is done in the classroom. Many teachers who weathered the pandemic and mastered the switch to online or hybrid teaching will be especially poised for this.

At the college level, by contrast, you’re on your own. That 10-page paper on Newtonian physics ain’t going to happen on your professor’s paid time in the lecture hall.

Lastly, we might also expect that certain subjects are going to be hit harder by all of this (and have to be more creative in their coping mechanisms). For example, the world language teacher has much less to quake over than, say, the American Lit instructor; the former’s students will be doing oodles of performative work, and formative assessments will be taking place daily in the classroom. AI can’t read aloud the lines of your French dialogue in your voice at your level.

Or at least not yet ...

Editor’s note: This is the first of two pieces exploring ChatGPT and AI’s impact on the educational landscape. The next installment will explore what teachers, parents, and schools can do to mitigate the deleterious potentials of such tools.
Matthew John is a veteran teacher and writer who is passionate about history, culture, and good literature. He lives in New York.
Related Topics