AI Must Be Taught, Not Banned

This article was written with the assistance of AI.

Amidst all the conversations about Israel, geopolitical issues and war, I wanted to return to a time before the world turned upside down. It used to be, the biggest things we had to worry about were sports (you know the outcome of the Taylorbowl, but as of writing I don’t yet), extracurriculars (insert generic plug for the Seforim Sale here) and of course, our newest conversation piece, AI.

I remember when ChatGPT first came out and took off, a little over a year ago. I spent my time playing around with it (maybe first during seder; who’s to say). I had it write Tolkien-style poems about the Silmarillion and an article about the glories of piracy written by Blackbeard (yes; I found them while I was writing this, yes; they are still incredible), sent excited texts to people showing off this cool invention that I and a few million of my closest friends had discovered, and then moved on. I thought that I had discovered a fun toy. I should have realized better.

One of the first things that my father, a corporate litigator, asked me to do was to ask ChatGPT a very technical question pertaining to a section of the New York State Legal Code. Ten seconds later, I sent him an answer. It was correct.

In fourth grade, we were taught how to write in cursive. I still cannot write in cursive, and it has never impacted me one iota. Almost no time was spent teaching us to touch type and we were often restricted from using computers, so I had to spend a significant amount of time in high school learning to touch type properly. I was reminded of this when, while doing research for a project recently, I came across the question posed to a group of professionals learning about AI: “Does your client want you to use AI in your role?” To appreciate the immense flaw of this question, imagine if that said “a computer,” “the internet” or “Google.” At our own peril do we fail to learn the proper and measured use of the technologies that we have available to us.

And yet ChatGPT’s release, and the subsequent use of it to write papers and answer questions, sparked an effort by schools and other institutions not to attempt to use this new and incredible technology to the benefit of their missions, but to shut it down. Instead of teaching us to use this powerful tool that we now have at our disposal, our educational institution and others like it have told us that to do so is to cheat. 

Now, I’ll point out that I am not suggesting that you feed your assignments in your next class through ChatGPT. I am saying that the AI I am using to help me write this is a tool, like the computer I am writing it on, the program checking my spelling, the lights letting me see at the ungodly hour I’m writing this and the phone I keep checking semi-obsessively throughout the writing process. The use of the tools at our disposal in writing and doing other work should be encouraged.

Through many conversations with teachers, I have gotten the sense — or in some cases the actual statement — that they do not know how AI affects their subject matter. What happens to classes assigning large readings when AI can summarize and analyze PDFs in a few seconds? What happens to writing assignments based on, say, history, when an AI can spit out the correct answer every time while you tune out the entire class? The answer is not — cannot be — that we simply disallow students from using AI, that we stop them from learning about what may be the most important aid they have in their future careers to write, analyze and do whatever people in Syms do faster, better and more efficiently. That is how you produce students who are behind before the race has even begun, who will enter a new job and spend a week doing research on a question while the other new hire runs the question parameters and various aspects of the needed research through ChatGPT and finishes in a tenth the time. Rather, the answer must be that we educate students on how to use AI properly and maximize its potential utility.

In some of my classes, I told my teachers that I would be using ChatGPT to help my writing. I didn’t use it to write for me, since it isn’t yet great at writing essays by itself. I instead used it to think of the right word when I just couldn’t find it (try asking for a list of ten options), write a punchy title for a paper, give me lists of books and articles on a certain niche topic, along with summaries of those that I found more relevant, or organize ideas I had vaguely thrown at a page into an outline I could revise and use to start writing. These uses and more are what I’ve found and used AI for. We can discover more uses together, if we can just let up on the stigma against using new technology and work to educate students who will have to live and work with it. Ultimately, the realm of white-collar employment must come to terms with the following truth: You will either learn to use AI, or you will lose your job to someone who does. I hope YU makes the right choice about which kind of students to produce.

As an example of how capable AI is, I asked ChatGPT to write me a short poem in the style of Tolkien about how incredible the Lord of the Rings movies are (a topic I am passionate about). I attach the result:

In the Shire's green embrace, a tale unfolds,
On silver screens, where epic saga molds.
From Hobbit's door to Mordor's fiery ring,
In Peter's vision, Tolkien's words take wing.

Elven grace and dwarven might entwine,
In Rivendell's glow, where fates align.
Through shadows deep and forests untamed,
Jackson's lens, the story acclaimed.

Gandalf's staff and Frodo's steadfast quest,
A Fellowship forged, put to the test.
Mountains high, and fields of battle vast,
Onscreen magic, a masterpiece cast.

In realms of fantasy, where courage gleams,
The Lord of the Rings, in cinematic dreams.
A journey splendid, with each frame's allure,
Middle-earth's splendor, forever pure.