This was a fantastic read, and I applaud you for having this conversation with your students. The thing that boils it all down for me was this: "Maybe we need to pause and talk with our students a bit more first. They have a lot to say."
I find that faculty are often dismissive of asking for student input and hearing what they have to say on the matter (for the record, I teach English at a CC). However, when students have buy-in, that's going to lead to better engagement and better learning. This all starts with having conversations such as these.
100% the goal here! In a perfect world, I'd love for these conversations to be happening all over (and not just on this topic) with students being empowered to give feedback and share their experiences and perspectives as a normed part of decision-making in education.
Marcus, I appreciate your commitment to surfacing student perspectives on AI and your transparency about your own uncertainties as an educator.
As a critical friend, I’d like to gently highlight a few biases that shape your analysis to deepen the conversation and to help all of us reflect on how we interpret student voice.
Teacher-Centric Framing: Throughout your piece, student comments are introduced and contextualized through your own lens as a teacher, not so much as a learner. For example, when you share students’ excitement about AI making work “easier,” you quickly pivot to concerns about authentic learning and academic integrity. This frames their curiosity as something to be managed rather than explored on its own terms. Offloading can be destructive, but it can also be useful, for example. It might be helpful to let some student voices stand on their own, even (or especially) when they challenge our professional instincts.
Selective Amplification:
You highlight student worries about AI undermining “real understanding” and “making school too easy,” which dovetails with your skepticism about technological shortcuts. Have you spoken about this with them before? Do they know your strong feelings? If any students expressed unreserved enthusiasm for AI’s creative or accessibility potential, those voices seem less foregrounded. This selective amplification can unintentionally reinforce your own cautious stance, rather than presenting a full spectrum of student thought.
Interpretation Through Adult Values:
When you summarize student responses, you often interpret them in light of educational values like “struggle,” “authenticity,” and “process.” These are core values, don’t get me wrong, but they’re also adult-centric constructs. For instance, a student’s desire to use AI to save time might reflect real-world efficiency skills valued outside school. It might even make sense for certain school assignments. I interviewed a high school junior this week for a piece of research I’m doing who uses ChatGPT every day—e.g. to help her with the verb system in a foreign language class. This is something we might miss if we only see it as a threat to rigor. We can’t know the reality without a free zone to speak.
Your analysis, while thoughtful and heartfelt, is shaped by a bias toward traditional educational priorities and a protective stance toward your own professional identity. This isn’t a flaw. Every educator brings their values to the table but it does mean that student voices are filtered, not raw. I encourage you to consider what it would look like to center student perspectives even more, perhaps by inviting them to co-author or respond directly to your reflections, especially those who disagree.
Thanks for opening up this important conversation!
Appreciate this response and of course happy to have this conversation—that's the ballgame right now, and we need all perspectives in it!
I guess my quick response to this: of course I'm biased! Our opening lesson for research with these sophomores is about the importance of owning and considering our biases rather than pretending they aren't there, so without question there are teacher-biases at play here. (Also sort the goal of this Substack, admittedly: a teacher-centered lens to education conversations that too often leave out the voices of teachers.)
That said, I do want to respond to each of your thoughtful points individually:
#1 - regarding Teacher-Centric Framing: this is the one I'm a bit lost with—as I feel like a post written by a teacher...is going to be teacher-centered in its framing? As in, I believe that you're right, but I also don't feel like I supposed or presented otherwise.
#2 - regarding Selective Amplification: I provided the actual data from the activities rather than just highlighting quotes, so while of course—given the results of the data!—I featured quotes that aligned with it as much as I could, I don't think of that as skewing as much as you might. I don't talk about AI at all in our classroom and honestly try not to share my perspective on most topics unless its pertinent to what we're doing, and students answered anonymously on the gallery walk. In the second activity with juniors, that was definitely framed as them providing feedback on my own choice—but there were three other choices where I received much more negative feedback from students than the one presented, so I guess you'll have to take my word that in our classroom students are very much supported in pushing back and disagreeing with me.
#3 - regarding Interpretation Through Adult Values: While I don't dismiss your anecdotes/examples, I feel like it is sort of an attempt to dismiss what students actually shared (just under 180 students across 6 different sections). I don't presume of course that those ~180 students speak for all others, and I think the best possible path forward would be one where we have MANY more of these types of student-centered conversations and surveys. But just reading your response, it feels like you're approaching it with your own bias, potentially? (Again, I'm biased, too!)
I also think you have to place this piece in the context of what the education debate is right now: one that is without question leaning towards AI enthusiasm in education (often by people whose professional identity is tied to AI at this point) and one that very rarely brings in student voices at all. That was my simple goal here: to pose an example of what bringing student voices in might look like, with the hope that there are many more instances of it going forward.
All this considered, I very much appreciate your thoughtful, well-reasoned response here—and happy to clarify/expand more!
Marcus, I love your point that AI enthusiasm in education right now is often by people whose professional identity is tied to AI. EduTech companies pushing AI are NOT the same as actual teachers discussing AI
Yeah, probably why when I shared this on LinkedIn (where pretty much everyone has some sort of AI-badge in their profile) it was mostly crickets 😂
Beyond this topic, too, I try my best to hold my convictions meaningfully but also loosely, as I know there are many instances in my 13 years in which I've been wrong (looking back now) and, consequently, that suggests that I'm probably wrong in all sorts of ways right now!
I'd rather identify as "remaining curious" than any specific position, even on AI, after all!
The bias that concerns me most is the teacher-centric one, which I’m not sure you can see beyond given your use of the audience you are writing for. Were you not gleeful that your students told you they agree with you? Marcus, it’s obvious your students are not going to act against your mentorship. You are beloved teacher. Ask your kids to tell you about people they know who use AI—their uncle, their friends mother. Ask them what they are hearing from their peers. That’s student centric. Enlist a group of students to write a real survey with validity and reliability and poll the school. I wonder if you would have taken the same tone and stance if the data had come out differently. After 20 years of teaching research methods on social science at graduate levels, it is hard to process the deluge of biases for and against. I decided to comment because I saw on your writing the possibility of examining these biases with a hard critical lens on yourself. You put up a good defense. The argument that the audience you are writing for which shares your bias is justification for presenting bias on this issue is problematic. I had high hopes that I would get something to take to the bank and was a bit disappointed that it wasn’t there. I’m sure those who readily agree with you see my comments as o I dunno stranger or out of bounds. I have the feeling you dont and maybe will take another try at replication. Btw, I think the common advice to talk to your students doesn’t mean survey them (teacher-centric). It means talk to them. I could be wrong here but I don’t think so to quote Randy Newman
We're probably at that point of talking past each other—I think you're looking for this to be something other than what I intended it as, which was a reflection as a teacher on two low-key (definitely not scientific!) conversations and surveys I did across the classes. I'm not trying to pretend otherwise, and I feel like this post is very open about that. (And I would love for there to be more/better-quality surveys along these lines!)
One data point I looked up, though, just to offer context: in those "teacher choice" surveys that I've done this year, I've presented seven different questions to students about choices I've made in our classroom with our readings and writings. Out of all seven, this was the only one in which the average response was over a 6.0/7.0 in agreement with the choice—meaning that my students did indeed push back more critically on every single other choice I made (with thoughtful reasons). As much as I concede some of your points, I don't think it's fair for you to immediately dismiss student voices in a classroom you aren't in just because you disagree with them.
So while I do think we've reached the point where this isn't going to find much reconciliation and I still do very much appreciate your engagement—and I'll keep your points in mind for future writings!
I don’t dismiss them because I want the findings to be different. When I read survey results, I’m prepared to accept them if I can rely on them. As you say, you didn’t intend to communicate reliable results—just to narrate an activity in your class. I found the piece enjoyable to read. You’re a unique and approachable voice in a sea of hostile voices. One thing I wondered about. Is your school middle class or private prep? As I look at surveys across time, I’m seeing a tendency among more affluent student populations to resist AI on the grounds you present, i. e., fear of loss of learning. On the other hand, what little I can find about high-poverty populations differs in that more kids are using AI with many of those still left in high school relying on it. That’s the genesis of my concern about issues of classroom culture. I didn’t know how to ask you
about sociocultural and material culture. I have absolutely nothing to take to the bank anyway. Sorry to trouble you in your space, my friend. I admire and respect what you are doing.
100% fair to ask, and I think one benefit of more-widespread surveys on topics like this would help us distinguish with more clarity how affluence and other contexts might impact what students say.
For my own context: definitely not affluent, about 50% give-or-take eligible for free-or-reduced lunch. About 50/50, too, in terms of white versus non-white. Those distributions look a bit different in my two sets of classes, too, in terms of AP Literature juniors versus English 10 sophomores, which is why I wanted to juxtapose them.
Thank you for this exchange, too, as I reflecting on not just our perspective but how we present it (especially for me with The Broken Copier) is always valuable and worthwhile. 🙏
Thanks so much for your patience! I can be very weird at times. You have the same sensibilities I have re: materially disadvantaged kids. My ten years teaching in k12 were in high poverty schools in California. Then at the university I carried that with me. Now that I’m emeritus I can’t seem to put it behind me. The past four months with the effort to dismantle the federal dept of education…I have to take deep breaths and stay centered. 🙏
I have always admired how you’ve built a classroom where students feel heard and take ownership of their learning. That shows up clearly in this post and in how honestly your students responded. You’ve created something meaningful.
Reading this, I kept thinking about how nuanced this conversation really is. I understand why your students feel uneasy about teachers using AI. Their concerns about fairness, effort, and connection are valid. At the same time, I think there are parts of a teacher’s job that students don’t always see. That is not a knock on them. They just haven’t had to think about lesson design, differentiation, or the many decisions that have to be made before a class even starts.
Students are still developing their skills. They’re in the middle of learning how to think critically, write clearly, and express their own ideas. That’s the work of school. Adults, while always learning too, have already been trained in many of those areas. We’re not doing the same kind of foundational work. When a teacher uses AI to help brainstorm or spark an idea, it’s usually within the context of work they already know how to do. It’s a tool, not a replacement for thought or effort.
It also helps to remember that many professions include support roles that are seen as normal and necessary. For example, paralegals often do much of the background research necessary for attorneys. Would students argue that shouldn't be done? I think thoughtful use of AI, especially behind the scenes, can fit into that same category. It doesn’t do the work for us, but it can help us do the work better.
That said, I fully agree that handing over feedback to AI or using it to replace the teacher-student connection is not the direction we should go. That part of the work is too personal and too important.
This post gave me a lot to reflect on. I’m glad you invited your students into the conversation. It reminds me that the more we involve them, the better our choices will be.
Love this response! I think quite often of our conversation about your work in your district, as I'm actually way more sympathetic to that approach—clear and systemic, with lots of support for teachers to be consistent and prepared—then the ad hoc, spaghetti-against-the-wall attempts I've seen mostly elsewhere. In my hypothetical "perfect world," there would be so much shared learning from these siloes of invention that your district is leading with, and we can then move slowly and collectively towards the best norms for students going forward.
My one reflection (aside from pretty much agreement in all other areas!) is that I think what makes the classroom and education unique is that, in my mind, we ought to be role models of what learning and academic work looks like. I've seen too many students lose faith not just with a classroom but school as a whole when educators don't walk the walk, so I think some sort of consistency (or, at minimum, clear transparency outlining the differences) is needed. It doesn't take many making bad choices to ruin the perception of us all, too, unfortunately.
More importantly, I really value the work you and your district are doing—and of course the ongoing conversation around it. Though I wish I could press a button to slow all of this down so that we could be much more prepared and supported across the board, within the whirlwind that we're in, it's essential to find the good people doing the good work, and without question I consider you one of those to lean on for my own learning! 🙏
I think it's great to bring in student voices and I've also seen student skepticism around AI though that is starting to change. Other than lesson plans and giving feedback on student work, I'm curious what other use cases you've shared with students about how both teachers and students might use AI. Deep research reports? Coding? Image generation? One thing I've seen a lot is the tendency to conflate AI with just writing output when it is so much more than that. Student input on AI use is valuable to the extent that they know and understand the technology (especially the power of more recent models and capabilities) which I find many of them (or their teachers) don't.
Without question, I presented a limited overview of what AI can and has looked like in education—primarily because, despite all the reading I try to do on my own time outside of work, I don't have firsthand experience with it in my own practices or learning experiences. (This is also why I am attentively following teachers who experiment with different structures, tools, and strategies around AI.)
I think right now we are seeing the intersection of a teaching profession and education system more broadly that severely under-resources educators in terms of the time they need to authentically learn about new content and development professionally so that they can confidently bring it into their classroom. Throw in the fact that AI is very much a "moving target" in terms of its capacity and particulars, and it feels like an impossible task to try and keep pace for many. (*raises hand*) This is also why I've long been an advocate for having a "digital literacy" core class for secondary students—which I think would be much more effective than expecting all teachers to adapt it into their own content areas with little to no support.
All that said, I do think there is a tendency to condescend to the students who are skeptical from those who are more enthusiastic about AI. Their skepticism of AI, and in some cases outright rejection, was not simplistic. The conversations I heard were thoughtful and nuanced, and raised very-real ethical dilemmas. So while I believe all the points you raise here are valid, I also think many have decided with a fixed mindset that "AI is the way" and, in that mindset, the only solution to critical student feedback is to criticize the students themselves—which is not something I'm going to do.
You misunderstand me - I completely agree on allowing AI student skeptics to voice their very valid concerns and I have also heard similar nuanced and thoughtful views from my students. My point is simply that frequently the points raised are misleading, wrong, and a year or two behind the curve (i.e., AI doesn't write very well - it does, that's the problem). Students need leadership on this issue. I have used the models - almost all of them, frequently, productively, and for lots and lots of different and interesting tasks. But I know I am an outlier so I sympathize and understand how overwhelmed and under resourced most teachers are. It's a growing divide but sometimes it means people are having two separate conversations.
Appreciate the clarification! and I think I'm at an agree/disagree point with this point you raise here.
Without question, there are lingering ideas about AI ("it gets everything wrong," etc.) that are not keeping pace with its development; and I do agree that students deserve leadership on this issue, which is why I think having specific courses led by those with that capacity would be really helpful (and much more scalable than just asking all teachers to get to this level).
But I also don't necessarily think the "two separate conversations" point is fair to the things that are rather universal. For example, students getting very frustrated at a lack of disclosure by teachers and students valuing authentic feedback? I don't think those concerns are going away.
(That said, really appreciate these points and just subscribed to your newsletter—as I'm trying to keep accumulating more perspectives from folks who are going about it in different ways in their contexts!)
I am 100% in agreement with the disclosure piece. This has been a very pivotal year in my AI usage because I've gotten totally transparent with students about when I've used AI for something and how I've used it. I teach an independent research course for high school upperclassmen and it's been great to use as a sandbox for all sorts of AI related things. Honestly, what kicked off my substack was the release of the Deep Research models last winter. These can now write more than a credible "research papers" as good or better than most High School students. What does that mean? How do we process that? How does that look going forward? What does research mean in an AI world? When teachers say "don't use AI to do research" and then students default to Google which ... uses AI, I think we are missing a conversation there. But it's messy and challenging and I don't like where everything comes out. What resonated with me in your post is my biggest issue (at least where I am) - no one is talking to the students about it! At least beyond - don't use it because it will destroy your learning. Anyway, glad to have the back and forth. That's why I'm here.
Of course! Excited to keep learning from you, too—that course sounds like a phenomenal place to be as a student. (Especially given how useless the traditional "research" model looks in most classrooms, I'm sure, an area I know I want to get better at.)
This was the very first thing I wrote on Substack about something I did in a different class with a custom GPT. Once I wrote this guest post I figured I might as well maintain my own Substack.
This really stuck with me. I’ve been working on something similar — not to make AI more useful, but to make its role more visible.
It’s called BootNahg, and it’s not a writing tool. It’s a way to analyze how a piece of work was made — after the fact.
One version goes to the student.
One goes to the teacher.
They run separately, after the assignment is done.
The student side analyzes ChatGPT logs to trace how the work was formed and generates a report on authorship and intent.
The teacher side looks at the final product through that lens — checking for signals of learning, creative effort, and consistency.
If there’s still ambiguity, both reports are timestamped and structured for real human review.
It doesn’t tell you if the writing was good.
It shows you how it got there.
I built this because my college-age kids are already living this shift. They’re not cutting corners — they’re crafting. But if no one sees that process, it all gets flattened.
This makes that process visible.
No agenda. Just sharing in case it’s useful. Happy to send it over.
This was a fantastic read, and I applaud you for having this conversation with your students. The thing that boils it all down for me was this: "Maybe we need to pause and talk with our students a bit more first. They have a lot to say."
I find that faculty are often dismissive of asking for student input and hearing what they have to say on the matter (for the record, I teach English at a CC). However, when students have buy-in, that's going to lead to better engagement and better learning. This all starts with having conversations such as these.
100% the goal here! In a perfect world, I'd love for these conversations to be happening all over (and not just on this topic) with students being empowered to give feedback and share their experiences and perspectives as a normed part of decision-making in education.
Marcus, I appreciate your commitment to surfacing student perspectives on AI and your transparency about your own uncertainties as an educator.
As a critical friend, I’d like to gently highlight a few biases that shape your analysis to deepen the conversation and to help all of us reflect on how we interpret student voice.
Teacher-Centric Framing: Throughout your piece, student comments are introduced and contextualized through your own lens as a teacher, not so much as a learner. For example, when you share students’ excitement about AI making work “easier,” you quickly pivot to concerns about authentic learning and academic integrity. This frames their curiosity as something to be managed rather than explored on its own terms. Offloading can be destructive, but it can also be useful, for example. It might be helpful to let some student voices stand on their own, even (or especially) when they challenge our professional instincts.
Selective Amplification:
You highlight student worries about AI undermining “real understanding” and “making school too easy,” which dovetails with your skepticism about technological shortcuts. Have you spoken about this with them before? Do they know your strong feelings? If any students expressed unreserved enthusiasm for AI’s creative or accessibility potential, those voices seem less foregrounded. This selective amplification can unintentionally reinforce your own cautious stance, rather than presenting a full spectrum of student thought.
Interpretation Through Adult Values:
When you summarize student responses, you often interpret them in light of educational values like “struggle,” “authenticity,” and “process.” These are core values, don’t get me wrong, but they’re also adult-centric constructs. For instance, a student’s desire to use AI to save time might reflect real-world efficiency skills valued outside school. It might even make sense for certain school assignments. I interviewed a high school junior this week for a piece of research I’m doing who uses ChatGPT every day—e.g. to help her with the verb system in a foreign language class. This is something we might miss if we only see it as a threat to rigor. We can’t know the reality without a free zone to speak.
Your analysis, while thoughtful and heartfelt, is shaped by a bias toward traditional educational priorities and a protective stance toward your own professional identity. This isn’t a flaw. Every educator brings their values to the table but it does mean that student voices are filtered, not raw. I encourage you to consider what it would look like to center student perspectives even more, perhaps by inviting them to co-author or respond directly to your reflections, especially those who disagree.
Thanks for opening up this important conversation!
Appreciate this response and of course happy to have this conversation—that's the ballgame right now, and we need all perspectives in it!
I guess my quick response to this: of course I'm biased! Our opening lesson for research with these sophomores is about the importance of owning and considering our biases rather than pretending they aren't there, so without question there are teacher-biases at play here. (Also sort the goal of this Substack, admittedly: a teacher-centered lens to education conversations that too often leave out the voices of teachers.)
That said, I do want to respond to each of your thoughtful points individually:
#1 - regarding Teacher-Centric Framing: this is the one I'm a bit lost with—as I feel like a post written by a teacher...is going to be teacher-centered in its framing? As in, I believe that you're right, but I also don't feel like I supposed or presented otherwise.
#2 - regarding Selective Amplification: I provided the actual data from the activities rather than just highlighting quotes, so while of course—given the results of the data!—I featured quotes that aligned with it as much as I could, I don't think of that as skewing as much as you might. I don't talk about AI at all in our classroom and honestly try not to share my perspective on most topics unless its pertinent to what we're doing, and students answered anonymously on the gallery walk. In the second activity with juniors, that was definitely framed as them providing feedback on my own choice—but there were three other choices where I received much more negative feedback from students than the one presented, so I guess you'll have to take my word that in our classroom students are very much supported in pushing back and disagreeing with me.
#3 - regarding Interpretation Through Adult Values: While I don't dismiss your anecdotes/examples, I feel like it is sort of an attempt to dismiss what students actually shared (just under 180 students across 6 different sections). I don't presume of course that those ~180 students speak for all others, and I think the best possible path forward would be one where we have MANY more of these types of student-centered conversations and surveys. But just reading your response, it feels like you're approaching it with your own bias, potentially? (Again, I'm biased, too!)
I also think you have to place this piece in the context of what the education debate is right now: one that is without question leaning towards AI enthusiasm in education (often by people whose professional identity is tied to AI at this point) and one that very rarely brings in student voices at all. That was my simple goal here: to pose an example of what bringing student voices in might look like, with the hope that there are many more instances of it going forward.
All this considered, I very much appreciate your thoughtful, well-reasoned response here—and happy to clarify/expand more!
Marcus, I love your point that AI enthusiasm in education right now is often by people whose professional identity is tied to AI. EduTech companies pushing AI are NOT the same as actual teachers discussing AI
Yeah, probably why when I shared this on LinkedIn (where pretty much everyone has some sort of AI-badge in their profile) it was mostly crickets 😂
Beyond this topic, too, I try my best to hold my convictions meaningfully but also loosely, as I know there are many instances in my 13 years in which I've been wrong (looking back now) and, consequently, that suggests that I'm probably wrong in all sorts of ways right now!
I'd rather identify as "remaining curious" than any specific position, even on AI, after all!
The bias that concerns me most is the teacher-centric one, which I’m not sure you can see beyond given your use of the audience you are writing for. Were you not gleeful that your students told you they agree with you? Marcus, it’s obvious your students are not going to act against your mentorship. You are beloved teacher. Ask your kids to tell you about people they know who use AI—their uncle, their friends mother. Ask them what they are hearing from their peers. That’s student centric. Enlist a group of students to write a real survey with validity and reliability and poll the school. I wonder if you would have taken the same tone and stance if the data had come out differently. After 20 years of teaching research methods on social science at graduate levels, it is hard to process the deluge of biases for and against. I decided to comment because I saw on your writing the possibility of examining these biases with a hard critical lens on yourself. You put up a good defense. The argument that the audience you are writing for which shares your bias is justification for presenting bias on this issue is problematic. I had high hopes that I would get something to take to the bank and was a bit disappointed that it wasn’t there. I’m sure those who readily agree with you see my comments as o I dunno stranger or out of bounds. I have the feeling you dont and maybe will take another try at replication. Btw, I think the common advice to talk to your students doesn’t mean survey them (teacher-centric). It means talk to them. I could be wrong here but I don’t think so to quote Randy Newman
We're probably at that point of talking past each other—I think you're looking for this to be something other than what I intended it as, which was a reflection as a teacher on two low-key (definitely not scientific!) conversations and surveys I did across the classes. I'm not trying to pretend otherwise, and I feel like this post is very open about that. (And I would love for there to be more/better-quality surveys along these lines!)
One data point I looked up, though, just to offer context: in those "teacher choice" surveys that I've done this year, I've presented seven different questions to students about choices I've made in our classroom with our readings and writings. Out of all seven, this was the only one in which the average response was over a 6.0/7.0 in agreement with the choice—meaning that my students did indeed push back more critically on every single other choice I made (with thoughtful reasons). As much as I concede some of your points, I don't think it's fair for you to immediately dismiss student voices in a classroom you aren't in just because you disagree with them.
So while I do think we've reached the point where this isn't going to find much reconciliation and I still do very much appreciate your engagement—and I'll keep your points in mind for future writings!
I don’t dismiss them because I want the findings to be different. When I read survey results, I’m prepared to accept them if I can rely on them. As you say, you didn’t intend to communicate reliable results—just to narrate an activity in your class. I found the piece enjoyable to read. You’re a unique and approachable voice in a sea of hostile voices. One thing I wondered about. Is your school middle class or private prep? As I look at surveys across time, I’m seeing a tendency among more affluent student populations to resist AI on the grounds you present, i. e., fear of loss of learning. On the other hand, what little I can find about high-poverty populations differs in that more kids are using AI with many of those still left in high school relying on it. That’s the genesis of my concern about issues of classroom culture. I didn’t know how to ask you
about sociocultural and material culture. I have absolutely nothing to take to the bank anyway. Sorry to trouble you in your space, my friend. I admire and respect what you are doing.
100% fair to ask, and I think one benefit of more-widespread surveys on topics like this would help us distinguish with more clarity how affluence and other contexts might impact what students say.
For my own context: definitely not affluent, about 50% give-or-take eligible for free-or-reduced lunch. About 50/50, too, in terms of white versus non-white. Those distributions look a bit different in my two sets of classes, too, in terms of AP Literature juniors versus English 10 sophomores, which is why I wanted to juxtapose them.
Thank you for this exchange, too, as I reflecting on not just our perspective but how we present it (especially for me with The Broken Copier) is always valuable and worthwhile. 🙏
Thanks so much for your patience! I can be very weird at times. You have the same sensibilities I have re: materially disadvantaged kids. My ten years teaching in k12 were in high poverty schools in California. Then at the university I carried that with me. Now that I’m emeritus I can’t seem to put it behind me. The past four months with the effort to dismantle the federal dept of education…I have to take deep breaths and stay centered. 🙏
I have always admired how you’ve built a classroom where students feel heard and take ownership of their learning. That shows up clearly in this post and in how honestly your students responded. You’ve created something meaningful.
Reading this, I kept thinking about how nuanced this conversation really is. I understand why your students feel uneasy about teachers using AI. Their concerns about fairness, effort, and connection are valid. At the same time, I think there are parts of a teacher’s job that students don’t always see. That is not a knock on them. They just haven’t had to think about lesson design, differentiation, or the many decisions that have to be made before a class even starts.
Students are still developing their skills. They’re in the middle of learning how to think critically, write clearly, and express their own ideas. That’s the work of school. Adults, while always learning too, have already been trained in many of those areas. We’re not doing the same kind of foundational work. When a teacher uses AI to help brainstorm or spark an idea, it’s usually within the context of work they already know how to do. It’s a tool, not a replacement for thought or effort.
It also helps to remember that many professions include support roles that are seen as normal and necessary. For example, paralegals often do much of the background research necessary for attorneys. Would students argue that shouldn't be done? I think thoughtful use of AI, especially behind the scenes, can fit into that same category. It doesn’t do the work for us, but it can help us do the work better.
That said, I fully agree that handing over feedback to AI or using it to replace the teacher-student connection is not the direction we should go. That part of the work is too personal and too important.
This post gave me a lot to reflect on. I’m glad you invited your students into the conversation. It reminds me that the more we involve them, the better our choices will be.
Love this response! I think quite often of our conversation about your work in your district, as I'm actually way more sympathetic to that approach—clear and systemic, with lots of support for teachers to be consistent and prepared—then the ad hoc, spaghetti-against-the-wall attempts I've seen mostly elsewhere. In my hypothetical "perfect world," there would be so much shared learning from these siloes of invention that your district is leading with, and we can then move slowly and collectively towards the best norms for students going forward.
My one reflection (aside from pretty much agreement in all other areas!) is that I think what makes the classroom and education unique is that, in my mind, we ought to be role models of what learning and academic work looks like. I've seen too many students lose faith not just with a classroom but school as a whole when educators don't walk the walk, so I think some sort of consistency (or, at minimum, clear transparency outlining the differences) is needed. It doesn't take many making bad choices to ruin the perception of us all, too, unfortunately.
More importantly, I really value the work you and your district are doing—and of course the ongoing conversation around it. Though I wish I could press a button to slow all of this down so that we could be much more prepared and supported across the board, within the whirlwind that we're in, it's essential to find the good people doing the good work, and without question I consider you one of those to lean on for my own learning! 🙏
I think it's great to bring in student voices and I've also seen student skepticism around AI though that is starting to change. Other than lesson plans and giving feedback on student work, I'm curious what other use cases you've shared with students about how both teachers and students might use AI. Deep research reports? Coding? Image generation? One thing I've seen a lot is the tendency to conflate AI with just writing output when it is so much more than that. Student input on AI use is valuable to the extent that they know and understand the technology (especially the power of more recent models and capabilities) which I find many of them (or their teachers) don't.
Without question, I presented a limited overview of what AI can and has looked like in education—primarily because, despite all the reading I try to do on my own time outside of work, I don't have firsthand experience with it in my own practices or learning experiences. (This is also why I am attentively following teachers who experiment with different structures, tools, and strategies around AI.)
I think right now we are seeing the intersection of a teaching profession and education system more broadly that severely under-resources educators in terms of the time they need to authentically learn about new content and development professionally so that they can confidently bring it into their classroom. Throw in the fact that AI is very much a "moving target" in terms of its capacity and particulars, and it feels like an impossible task to try and keep pace for many. (*raises hand*) This is also why I've long been an advocate for having a "digital literacy" core class for secondary students—which I think would be much more effective than expecting all teachers to adapt it into their own content areas with little to no support.
All that said, I do think there is a tendency to condescend to the students who are skeptical from those who are more enthusiastic about AI. Their skepticism of AI, and in some cases outright rejection, was not simplistic. The conversations I heard were thoughtful and nuanced, and raised very-real ethical dilemmas. So while I believe all the points you raise here are valid, I also think many have decided with a fixed mindset that "AI is the way" and, in that mindset, the only solution to critical student feedback is to criticize the students themselves—which is not something I'm going to do.
You misunderstand me - I completely agree on allowing AI student skeptics to voice their very valid concerns and I have also heard similar nuanced and thoughtful views from my students. My point is simply that frequently the points raised are misleading, wrong, and a year or two behind the curve (i.e., AI doesn't write very well - it does, that's the problem). Students need leadership on this issue. I have used the models - almost all of them, frequently, productively, and for lots and lots of different and interesting tasks. But I know I am an outlier so I sympathize and understand how overwhelmed and under resourced most teachers are. It's a growing divide but sometimes it means people are having two separate conversations.
Appreciate the clarification! and I think I'm at an agree/disagree point with this point you raise here.
Without question, there are lingering ideas about AI ("it gets everything wrong," etc.) that are not keeping pace with its development; and I do agree that students deserve leadership on this issue, which is why I think having specific courses led by those with that capacity would be really helpful (and much more scalable than just asking all teachers to get to this level).
But I also don't necessarily think the "two separate conversations" point is fair to the things that are rather universal. For example, students getting very frustrated at a lack of disclosure by teachers and students valuing authentic feedback? I don't think those concerns are going away.
(That said, really appreciate these points and just subscribed to your newsletter—as I'm trying to keep accumulating more perspectives from folks who are going about it in different ways in their contexts!)
I am 100% in agreement with the disclosure piece. This has been a very pivotal year in my AI usage because I've gotten totally transparent with students about when I've used AI for something and how I've used it. I teach an independent research course for high school upperclassmen and it's been great to use as a sandbox for all sorts of AI related things. Honestly, what kicked off my substack was the release of the Deep Research models last winter. These can now write more than a credible "research papers" as good or better than most High School students. What does that mean? How do we process that? How does that look going forward? What does research mean in an AI world? When teachers say "don't use AI to do research" and then students default to Google which ... uses AI, I think we are missing a conversation there. But it's messy and challenging and I don't like where everything comes out. What resonated with me in your post is my biggest issue (at least where I am) - no one is talking to the students about it! At least beyond - don't use it because it will destroy your learning. Anyway, glad to have the back and forth. That's why I'm here.
Of course! Excited to keep learning from you, too—that course sounds like a phenomenal place to be as a student. (Especially given how useless the traditional "research" model looks in most classrooms, I'm sure, an area I know I want to get better at.)
This was the very first thing I wrote on Substack about something I did in a different class with a custom GPT. Once I wrote this guest post I figured I might as well maintain my own Substack.
https://fitzyhistory.substack.com/p/my-experiment-with-guided-reading
This really stuck with me. I’ve been working on something similar — not to make AI more useful, but to make its role more visible.
It’s called BootNahg, and it’s not a writing tool. It’s a way to analyze how a piece of work was made — after the fact.
One version goes to the student.
One goes to the teacher.
They run separately, after the assignment is done.
The student side analyzes ChatGPT logs to trace how the work was formed and generates a report on authorship and intent.
The teacher side looks at the final product through that lens — checking for signals of learning, creative effort, and consistency.
If there’s still ambiguity, both reports are timestamped and structured for real human review.
It doesn’t tell you if the writing was good.
It shows you how it got there.
I built this because my college-age kids are already living this shift. They’re not cutting corners — they’re crafting. But if no one sees that process, it all gets flattened.
This makes that process visible.
No agenda. Just sharing in case it’s useful. Happy to send it over.