Instead of Integrating AI Into Your Classroom, Do This
Seventeen ideas about teaching for the fall semester
Thank you for being here. As always, these essays are free and publicly available without a paywall. If you can, please consider supporting my writing by becoming a patron via a paid subscription.
My tenure packet is due at the end of this week, so I thought I'd go a bit lighter today. Well...twenty five hundred words later and I guess I've failed. If I had more time I would have written you a shorter letter. I hope that you find something inspiring, challenging, or useful here!
![](https://substackcdn.com/image/fetch/w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fe391559a-bb17-43a6-b433-f2dd8a5fa770_2000x1400.png)
We're now squarely in August and the first day of classes on the 26th of the month is quickly approaching for me at Harvey Mudd. Those of you in and around classrooms are likely starting soon if you haven't started already.
Next week I'm excited to facilitate two workshops on teaching in the age of AI for my colleagues at Harvey Mudd. My main thesis might surprise you.
I've been thinking about how best to engage AI from a pedagogical perspective for a long time now. In writing and sharing my thoughts, it's been interesting for me to reflect on how my own schooling experiences have shaped my views on how school should work and what education is for. Throughout my life, I've been in situations that have been decidedly small in size and relational in nature. From my homeschooled elementary and middle school days to private high school to a small undergraduate technical university to a small—at least by population—institute of technology for grad school, I've never attended a school with more than one or two thousand students.
What I've learned more than anything from these experiences is the power of personal relationships in an educational environment. I am out of fingers counting the teachers and professors who have had a significant influence on my formation.
I'm sure that there are aspects of my experience that don't translate well to all schools and situations. To be educated in places where you are a name, not a number—and if you were a number you'd be one of twenty not one of two hundred—makes a big difference. But both my experiences and my short career thus far teaching has convinced me that truly transformational teaching cannot exist except in the context of personal relationships.
In many ways, this foundational belief in the power of the relationship between teacher and student is the force that has stabilized my own approach to the disruptive forces of generative AI that we've seen unleashed in the last few years. For every truly novel pedagogy and educational approach that generative AI enables, there are ten poorly-motivated and ill-justified existing approaches to teaching that it makes untenable.
With this in mind, today I want to offer a peek inside my brain as we approach classes this fall. This list is for me, as much as it is for you: reminders of the practices I want to embody to live out my values as an educator. As we think about how to engage AI in our courses this fall, we ought to spend at least the same amount of time thinking about all the other, more foundational aspects of what happens in the classroom.
You may find that some of these don't work in your context, and that's ok. I realize that there are a wide variety of contexts and situations. And yet, I would challenge you to consider how you might creatively integrate some of these ideas into your own practice. Perhaps the greatest impact that generative AI will have on your pedagogy this fall is helping nudge you toward doing the things that we already know are effective.
The best way to prepare to engage AI this fall doesn’t have anything to do with AI itself. Rather, it’s about cultivating classroom communities of trust. Get the classroom culture right and you’ll be in good shape to engage AI. But if you fail to get the foundation set, you’ll at best be building castles in the sand.
If something here jumps out to you or my list inspires one of your own, please do me a favor and leave it in the comments below. Here we go.
1. Learn students’ names.
Addressing your students by their names is hands down one of the easiest things you can do as an instructor to cultivate a fertile classroom community. It takes some effort but is well worth it. I always use an Anki flashcard deck to help me learn with spaced repetition and would recommend you try it if you’re not familiar. If you can, spend some time learning your students’ names before the semester begins as starting class on the first day with names is huge.
2. Foster metacognitive habits.
One of the most important and yet often-ignored aspects of learning is the power of metacognition and reflection. The course content is itself important, but we often fail to incorporate many opportunities for students to take a beat to reflect on what they have learned and how they have grown.
One way to do this is with a learning journal which students can return to throughout the semester to document their progress and how they are feeling about their growth. This semester I’m going to experiment with having students write down their personal motivations and learning goals on the first day of class and periodically return to make small entries reflecting on how it’s going throughout the semester.
3. Teach with transparency.
Transparent teaching has been a core tenant of my philosophy of teaching for a long time now. The basic idea is to help students understand how they learn by being explicit not only about what we're doing but why. This is another relatively low-lift tuneup to your existing curriculum: just add a bit of context either in writing or verbally in class about the design of the course and the rationale behind its components. I do this with a section in my syllabus.
4. Communicate explicit learning objectives.
Another easy upgrade to your curriculum is listing learning goals explicitly on each assignment or lecture. I do this with a slide near the beginning of each lecture or a list at the top of every assignment. I don't want students to be guessing about what they're supposed to be getting out of the work I'm asking them to do. Here’s an example of a list of learning objectives at the top of one of my assignments.
5. Make communication policies clear.
I adopted this one in response to some advice from fellow Substacker and Professor
. In this post from last year, he articulates why a communication policy is a helpful part of any syllabus to set expectations and boundaries.Without a doubt, one of the simplest and sanest things you can do to get some control over your work and some freedom with how you choose to do it, is to set boundaries around your communications. This is what I mean by a communications policy: A simple set of rules for yourself and the people with whom you work, for how, when, and how often you will engage in communication. Crafting this policy and trying it out now, in July, and then deploying it in full when the semester starts is one of the smartest and simplest things you can do to really get ready for Fall.
6. Create frameworks for feedback.
I read Radical Candor by Kim Scott last summer and learned a lot from it. While it's geared toward industry teams and management, I found a lot of the advice fit well in a classroom context as well. Now I have a slide on the first day of class in each of my courses to give my students a crash course in radical candor and to explain how I want them to challenge directly and care personally for me and one another.
We can probably all identify times when we've experienced ruinous empathy or obnoxious aggression in the classroom. Having a framework to name those pathologies can help us to avoid them in our classrooms.
7. Double down on active learning.
Every year as I tweak my courses I am on the lookout for ways to incorporate more active learning. Getting students out of a passive listening posture into one that is active and engaged is almost always a huge win. If students could watch recordings of your class sessions without losing much, then you've likely got lots of room to build in more in-class activities.
The good news is that this is fun. I try to incorporate a wide range of activities from quizzes to bingo to in-class design or programming exercises. An LLM can be your friend here! In preparation for this fall, I fed my syllabus to Claude and asked for some ideas for active learning activities. There were some seeds of good ideas there that with a bit of finesse will shape up into nice activities.
8. Encourage experimentation.
The power of the prototyping mindset is in iterative improvement after failure. One way to de-risk failure is to put less pressure on it. Help students to think of their work as an experiment where they go in with a hypothesis and a plan but don't hold too tightly to the outcome. Learning to embrace process over product is key. What matters most is the iterative process of failure, reflection, adjustment, and trying again.
9. Cultivate community.
Classrooms have great potential for community and connections between students but we often leave so much of that potential on the table. Metcalfe’s Law applies to the classroom too. Most of the time, we end up focusing on the one-on-one relationship between the instructor and the individual student or the class as a whole. But what if we did more to leverage the connections between students?
This fall I'm going to be more intentional about mixing students up in groups and getting them to work together, especially at the beginning of the semester. The classroom can be a great place to seed relationships between students to more fully realize the potential for forming relationships to help students feel that they belong not just as individuals but as part of a community.
10. Connect individually with each student.
I realize that this may not be feasible for all classes, especially for bigger ones. But nonetheless, I will encourage you to be creative. Even if it's not much, find ways to spend individual time with each student. This might look like being intentional about how you roam the room during in-class activities to connect with all students and not just those who ask for help. Take the initiative to stick your nose in your students' work to ask how they're doing. Many times, even if they're not asking a question, they've got one. Just the small act of moving towards students and building rapport with them can help to break down barriers when they need help or are feeling unsure about coming to your office hours.
11. Build shared responsibility for learning.
Learning in a classroom is a shared enterprise between instructors and students. Each party must bring something to the table for it to be successful. The first step is to make this explicit for students. What are you bringing? What do you expect your students to bring? What do they feel they should bring?
Especially at the outset of a course, this sort of framing can be a helpful segue into discussions around classroom culture and norms as well as a way to collaboratively construct policies on technology, computers, phones, and attendance. The more you can do to build these things with your students instead of for them, the better off (and happier) everyone will be.
12. Get alongside students.
A corollary to the previous point is that the more you can do to get on the same level as your students the better. I find that it's more fun for me and more effective for students if we can see learning as a shared enterprise. As an instructor, I'm not here to lay out the gauntlet you need to run to clear the bar and get a certain grade in the class. I'm here to help you build valuable skills (see points about learning goals above) and to give you useful and honest on your work so that you can grow and improve. We could learn a lot from thinking more about the coaching aspects of our roles as teachers.
13. Model vulnerability.
If you want your students to be honest and vulnerable with you, you had better model it for them. Given the power structures at play—we do make the rules and give the grades after all—it is a lot to ask students to be open about their challenges with you and to be honest about the places where they've fallen short.
So, be the first to admit when you fall short. When we fail to get a lecture posted on time or are late getting our students feedback, own up to it and apologize. If we want our students to be honest with us, we had better be honest with them.
14. Reframe from "have to" to "get to".
Student agency exists, even in a required course. It's obviously easier for students to feel their agency in a course that they've chosen as an elective, but the truth is that they've chosen, one way or another, to be a part of every class that they're taking.
One of my favorite reframes is going from "I have to" to "I get to". If you're struggling with the things ahead of you, consider the alternatives. Remember the reasons that you're there.
15. Trust your students.
What would it look like to trust your students until they give you a reason not to? There is nothing quite like an air of suspicion to create tension in the classroom. I’m not arguing for a pollyannaish approach, but rather that we give the benefit of the doubt and err on the side of generosity toward our students.
16. Offer opportunities for failure and retries.
If you believe that learning happens in the iterative process of trying, failing, reflecting, and trying again, why not design your assessment scheme to incentivize it? One of my biggest gripes with traditional grading is that it is often one-and-done. Students do some work, lob it over the fence for feedback, and the instructor marks it up and lobs it back. Then the students look at the number on the top and perhaps briefly scan any written comments and that is that. What an efficient use of all the time and effort on feedback. No wonder we’re sanguine about outsourcing it to AI. How much better to design a system for students to have the opportunity to revise and resubmit their work so that they are incentivized and encouraged to engage with the feedback?
One of my favorite aspects of alternative grading is that it opens up lots of options for this type of revision. In the specifications grading scheme I built last year, I let students revise and resubmit their work without penalty. It worked great and students learned more and presented better work than I'd seen in all my previous offerings of the course. Specifications grading might not be the right choice for you, but if you're curious, I'd encourage you to explore the excellent
Substack for some inspiration.17. Embrace friction.
Here's the thing: if there is one thing that we'll need to emphasize more than ever about teaching and learning in the age of AI it is that learning is hard work. Anything worthwhile is. As we go this fall, let's remind students that we're in this journey of learning and growth together and encourage each other to dig in and embrace the worthwhile challenge of learning new things and developing new skills.
Reading Recommendations
Another must-read from
, this time on the latest tech release to have me raising my eyebrows, the friend pendant. At least they have a fancy domain name (that apparently cost a cool $1.8 million).with wise words on how we should consider engaging AI in our writing processes with a particular eye on first drafts.Allow me, then, to close with a simple exhortation: we need people in our lives, not the simulation of people.
over at writing about why having genAI write your letter to your Olympic hero is exactly the type of thing that we’re worried about. The fact that the thesis behind this ad made it to primetime really makes you wonder. John pulls a quote from Matt Steib’s piece on the ad in NY Magazine.Having AI generate a draft for you might not give you the same experience because an AI-generated draft might be something’s child’s draft, but not your own. The creative risks aren’t the same. Yes, you get a fast start from going to a blank page to a rough draft but having a machine do that for you feels like leaping from the crowd in front of the Boston Marathon and sprinting the last hundred yards ahead of everyone else.
Like many things about AI itself, this is something seemingly nobody wants. People were quite upset with the ad, which kept playing during prime time. “I flatly reject the future that Google is advertising,” wrote Syracuse media professor Shelly Palmer. “I want to live in a culturally diverse world where billions of individuals use AI to amplify their human skills, not in a world where we are used by AI pretending to be human.” Brand strategist Michael Miraflor wrote that the ad was quite similar to the Apple iPad commercial from May that was widely reviled. “They both give the same feeling that something is very off, a sort of tone-deafness to the valid concerns and fears of the majority,” he wrote, adding that both were developed in-house.
Last week
and his co-author Paul Bruno released a guide to help you avoid common misconceptions about AI and education. While you’re at it, check out Ben’s Substack .Education Hazards of Generative AI provides a basic scientific overview of how large-language models (LLMs) work and connects this knowledge to practical implications for educators. This document is intended as a resource for teachers, principals, school district administrators, parents, students, policymakers, and anyone else thinking about using generative AI for educational purposes.
The Book Nook
A week or two I started Wanting by
and boy is it eye-opening. I stumbled on Luke and his work on Substack and grew curious about this guy René Girard he was always writing about. I’m sure that I’ll have more coherent thoughts to post once I finish the book and digest it a bit, but so far I’m enjoying the way he is unpacking Girard’s insights in mimetic theory and identifying the ways that they play out all around us. Seeing my toddlers’ interactions through the lens of mimetic theory has been quite something.The Professor Is In
Hard to believe that my tenure packet is due this week. The five years since I started at Mudd have truly flown by. It’s fun to have the chance to sit and reflect on what I’ve been learning and think about where I want to direct my energy moving forward.
Leisure Line
When in Connecticut, eat Pepe’s. Managed to sneak in one last trip before we took to the skies to come back west.
Still Life
Spotted this awesome sundial at a playground in Glendale this week. Very cool, even though it was off by an hour for daylight savings.
Is it really an either/or proposition?
In short, I would say stop teaching and start coaching … regardless of whether you integrate AI or not.
For me, AI allows for more coaching, not less. And coaching allows me to teach the right use of AI better.