FBF 53: Being skeptical about AI in education might be the biggest mistake of your teaching career
See how the thoughtful and ethical integration of AI could be the career-defining move you're missing out on
Trying to hold back the AI revolution is like trying to hold back the ocean with a spoon.
ChatGPT is the single most mind-boggling thing I have seen on the Internet in the past 20 years.
But many educators, stuck in their ways and afraid of any change to the status quo, are pulling the fire alarm and seeking to “ban AI.”
Mainstream education is currently fixated on developing artificial intelligence capable of identifying the writing of other artificial intelligence. These efforts are both understandable and necessary. We clearly need robust assessment methods ensuring students can independently master material—indeed, old-fashioned blue books and oral exams are making a long overdue resurgence.
Nonetheless, the real revolution lies in recognizing how AI, when thoughtfully and ethically integrated, can enhance learning. This notion seems to have not made its way into the stodgy ivory towers.
For every student who uses AI as a tool for intellectual sloth, there are five others eagerly typing in prompts right now to broaden their horizons. Or at least there would be if we taught them how.
I am keenly aware that encouraging the use of ChatGPT in the classroom feels like opening Pandora’s box. But in my opinion Pandora is already out of her box, and our best bet now is to adapt our educational systems. We must craft a larger, more inclusive box, where AI and human intellect enhance each other.
If our schools intend to cultivate the leaders of tomorrow, they need to pivot from banning to embracing artificial intelligence like ChatGPT. And it’s the teachers, the frontline warriors of education, who must lead this charge.
If you are a teacher yet to harness the power of AI in your classroom, I am here to show you how to level up your teaching and achieve better results through a responsible and innovative integration of AI into your pedagogical toolkit.
1) Prompt Writing 101
Albert Einstein once said, “If I had an hour to solve a problem I’d spend 55 minutes thinking about the problem and 5 minutes thinking about solutions.” Charles Kettering remarked that “A problem well stated is a problem half-solved.”
Up to this point, education has been largely concerned with correct answers. In the AI era, the burden has switched to asking the right questions.
AI's effectiveness hinges on human ability to correctly identify problems and ask insightful questions that guide AI towards meaningful and innovative solutions.
Indeed, prompting is arguably the single most important and marketable skill that we can teach today. One effective technique is called RTFC.
After teaching this format, the homework would be two-fold: 1) Identify a problem where this framework would be most effective and 2) Write a comprehensive prompt using RTFC to solve the problem.
The homework would be assessed on the quality of the ChatGPT output. I tried this very exercise when teaching The Catcher in the Rye last year, and the results were remarkable.
2) Empowering ESL Students
When I first discovered ChatGPT, I showed it to the principal of my school, an Algerian Islamic scholar. Although he is very sophisticated, writing in English—his fourth language—had always been a struggle.
For many years, when he needed something written he would send voice notes to a teacher on staff who was a former newspaperman who would whip them into the Queen’s English in no time.
The principal grasped the power of ChatGPT instantly. Ever since, his ability to communicate clearly and efficiently with the parents, staff, and students has soared.
One of the most exciting features of LLMs (Large Language Models) is that they can level the playing field for ELLs (English Language Learners) and other non-native speakers. Many people wrongly assume that non-native speakers are less intelligent because they lack the language to express themselves.
Teaching ELLs how to use ChatGPT to express themselves more clearly not only gives them access to greater self-expression in their second language, it also allows them to become agents in their own learning.
They can ask the program to correct their grammar mistakes and explain why they were wrong. They can then ask the program to generate a quiz to test their understanding of the new grammatical concept. If they are still confused, they can ask it to explain the material in their native language. Truly amazing stuff!
3) Teach them what AI can’t do
Before you assume that I have wholly capitulated to the bot overlords, let me emphasize this: Chat GPT will NEVER be able to THINK like a true human, inshAllah.
A couple of years ago I wrote a series called 30 College Essays in 30 Days. Before the ink had dried on my final essay, Stephen Marche’s Atlantic article with the disheartening title, “The College Essay is Dead,” was published. Based on the title alone, it seemed my month of daily reflection was in vain—soon AI chatbots would be pumping out college essays at a pace that put mine to shame.
The basic premise of the Atlantic piece is that AI has become so advanced that scholarship within the humanities, already under attack by the rush to STEM, may soon become obsolete. If a computer program can write an essay as sophisticated as and better reasoned than that of an undergraduate, then the death knell of humanities has surely been sounded.
While I can accept that computers may well have mastered the 3rd person point of view, I believe we humans can still retreat to the bastion of 1st person storytelling. If anything, the rise of sophisticated AI programs increases the importance of the personal statement because it is one of the few pieces of writing that couldn’t possibly be written by a computer, at least when done right.
And yet so many personal statements are NOT done right. They are Frankenstein-ian mish-mashes of feedback from too many editors; they are airy pontifications from teenagers who think they have it all figured out; or, at their worst, they are fraudulent pieces written by “consultants.” Most of these essays could indeed be written by an AI program, and that is exactly why they don’t work.
Great college essays make your hairs tingle. They make you cry. They make you laugh. They are authentic. They are built around stories that reveal hard-earned truths that have finally been released from the captivity of the subconscious. They bring the reader, for one fleeting moment, into the soul of a high school senior.
While it is demoralizing that AI can be mis-used to help us glide through college, I take heart that we still need our humanity to open the door.
4) A free personal tutor for all
Sal Khan delivered a TED Talk that made me nearly drive off the road.
It was called “How AI could save (not destroy) education.” In it, he argues that if we put the right guardrails, “we're at the cusp of using AI for probably the biggest positive transformation that education has ever seen.”
He explains how the Khan Academy team has built Khanmigo (cringey name, I know), which can serve as a personal tutor for all students on the planet and a teaching assistant for every teacher.
Crucially, and unlike ChatGPT, Khanmigo will not give you the answer if you ask it to. Instead, it will say “I’m your tutor. What do you think is the next step for solving the problem?” When the student makes a mistake, it asks the student to explain their reasoning and is even able to divine the probable misconception in the student's mind.
If a cynical student asks, “Why do I need to learn this?” It Socratically responds, “Well, what do you care about?” Based on the response, it will show how that piece of knowledge connects to a student’s interests.
A good ChatGPT prompt that accomplishes something similar is: “Create a real-life scenario where one would need to use [specific mathematical concept or formula]. Then, walk through the problem-solving process step-by-step, explaining how the concept is applied.”
You can get ChatGPT to become Jay Gatsby or William Shakespeare and then engage in a dialogue. For out of the box thinkers, you can ask it to become the Mississippi River and write poetry about all the ways it is polluted.
The real promise of ChatGPT is that it can solve what Benjamin Bloom called the “2 sigma problem.” Bloom observed that students who received one-to-one tutoring performed two standard deviations (“2 sigma”) better than students who learned through conventional classroom instruction. This means that tutored students perform better than 98% of students in a regular classroom.
The “problem” is how to effectively scale this type of personalized instruction. Through immediate feedback and intervention, mastery-based instruction, and data-driven insights, Khanmigo appears to have cracked the nut of how to supply accessible and affordable personalized instruction at scale.
5) The “floor” of acceptable output has risen dramatically
ChatGPT writes at a B+ high school level…without grade inflation. If I’m totally honest, I would say only about 10% of my students can write better than it. 100% of them, however, could think better if they were encouraged to use it.
In the hands of a curious student eager to understand how the world works, ChatGPT is nothing short of magic. She might write: “Explain the Krebs cycle as if I were a fifth grader. Use analogies and simple language to make it understandable.” The next prompt might be, “Now help me remember the function of the organelles of a cell by attaching each of them to a Taylor Swift lyric.”
When we switch our focus from policing our students’ use of AI tools to encouraging their ethical and responsible use, we can increase our expectations by 80%. “I just don’t get it,” is no longer an excuse in a world in which you have unlimited opportunities to ask follow-up questions.
Just as the calculator significantly expanded the range of mathematical problems students are expected to solve in the math classroom, so too can AI elevate the depth and breadth of learning across the entire curriculum.
If you are like the many teachers I know who have steadfastly refused to explore ChatGPT yourself, I challenge you to lean into discomfort and embrace AI, not as a substitute for your expertise, but as a powerful ally in your mission to educate.
The thoughtful, ethical integration of AI in education isn't just a fleeting trend, it's a transformative shift. Don’t let your skepticism hold you and your students back from exploring the vast horizons of knowledge in ways that were once thought impossible.
If the ideas of this article resonated with you, please hit the❤️ button. If you’d like to explore the single best resource I have found to improve my skill with AI, check out Write with AI by Nicolas Cole & Dickie Bush. It is the only Substack that I pay for, and it is digital gold every time.
Thoughtful & thought provoking. Thanks for these great tips on using AI to enhance the learning experience rather than to replace students skills & curiosity in the process of learning & writing.
Great article, Hamzah! I agree: learning to ask good questions (and beautiful questions!) is a key skill for our students to master. I’d like to see every student become a “Questionologist” (to use Warren Berger’s term). Here’s a great book on how to turn classrooms into places where questions thrive: Beautiful Questions in the Classroom: Transforming Classrooms Into Cultures of Curiosity and Inquiry https://amzn.eu/d/3uVZqeL. And one more excellent resource— The Right Question Institute: https://rightquestion.org/.