Priten: Welcome to Margin of Thought, where we make space for the questions that matter.
I'm your host, Priten, and together we'll explore questions that help us preserve what matters while navigating what's coming.
If you're a teacher, you know the questions around AI and ed tech are hard, and if you're a school administrator, you're facing a lot of big decisions, often with incomplete information, competing priorities, and real consequences for students and staff.
Today's guest is Justin cia, executive Director of the Center for Teaching and Learning at Episcopal Academy.
Over the last decade, Justin has served as dean and director at three independent schools.
And he brings a unique approach that blends cognitive science, human connection, and an experimenter's mindset.
We're going to talk about what ethical decision making looks like from the administrator's chair, whether it's choosing an AI tutor, implementing surveillance technology, or setting cell phone policies.
This is about the frameworks we need when there are no easy answers.
Let's begin.
Justin : So my formal title is the Buckley Executive Director of the Center for Teaching and Learning, and I'm at a pre-K through 12 independent school in suburban Philadelphia.
I'm entering year three.
I had made my way into independent schools as a student because I was good at stopping ice hockey pucks and was a good student.
So I found my way into the world of New England boarding schools.
It was transformative for me.
I came from a big public high school.
School was easy.
school was less easy when I went to this independent school, but I had great support and, figured it out pretty quickly.
As I began to navigate that hidden curriculum that wasn't really present in my previous experience, didn't think I'd work in schools.
Found my way back into a summer session teaching.
Had a great experience and I was a first generation college student, so this was like a transformative experience to go to a school that I, the St. George's where I went to a Newport, Rhode Island.
And really enjoyed it.
And I enjoyed the human side of working with students, coaching and living in dorms and teaching.
Long story, less long I worked at two different independent schools before joining Episcopal Academy in a variety of roles at St. George's.
I was leading the academic program, sort of as a associate head of academics type role.
So I was sort of the, the buck stop there with me as it relates to academics.
I found my way to EA because they have a dedicated CTL and I could really just nerd out about teaching and learning which was really attractive.
In some ways it was a step sideways or even backwards, but I never viewed it that way because I got to do the things I really love about schools the intricacies, the complexities, and the nuance of really good teaching and learning.
ironically in independent schools, those things sometimes matter less than people perceive.
Because it's about the prestige access or other things to lead A CTL that had been in existence prior to my arrival as the new executive director, was really appealing to me.
And I get to have conversations like this and think about AI because I have the bandwidth to do so.
Priten: That's awesome.
I don't know if I've heard of too many other, CTLs at, secondary schools and so that's exciting.
Justin : They are not many and they, and the ones that do exist, all serve in different kind of roles.
, We take a lot of inspiration from higher ed, certainly, and some of the frameworks that exist with their CTLs around things like being a hub.
Being an incubator, being a temple to sort of celebrate and then being a sieve to funnel research.
So we play all of those roles depending on the thing that we're doing that day or at that time.
Priten: Yeah.
That's amazing.
that's very exciting.
So, you started three years ago in this role, you said.
was that your first time with the school or were you at the school prior to that
Justin : It was my first foray into the school, and ironically on the AI front, I was certainly early to the scene, particularly amongst secondary educators in my previous role, and that was largely as a history classroom teacher.
when three and a half first rolled out, I'm like, this is something.
It felt a little bit like that COVID moment of like, Ooh, we're gonna have to deal with this.
it's not gonna go away.
We talked about the paradigm shift of COVID.
I'm like, no, no this is the real paradigm shift.
And I took my students' summer essays, which had already been graded and given back months prior.
I give whole class feedback for my essays in class.
So I read everything and give a one pager and then a couple of dedicated, specific individual prompts.
Because the mistakes, the opportunities for improvement are always similar.
So I did the one class feedback in that summer essay and I said to my students, they're all seniors.
I said, Hey, would you guys be okay if I tried this?
'cause we were sort of dabbling with it 'cause it was this new novel thing it was clunky to upload back then you had to do like file at a time.
Really annoying.
But I uploaded these 16 essays and it gave me whole class feedback that was better than the whole class feedback that I had generated on my own.
And it was a moment of like, yeah, there are so many different layers to this that are really, really impactful and potentially powerful.
So we played with it for the remainder of the year and the students, most of 'em were going off to college and university.
So I started to give 'em a bit of a primer.
when I arrived to EA it was clear they were also dabbling in the AI space.
And I had a little bit, I don't wanna call it expertise 'cause I don't think anybody has expertise right now, but I certainly had a little bit more engagement and, understanding of the landscape.
So what was a bug?
Initially it was very clearly a feature and not something that they had hired before in the leadership of CTL.
So that was kind of a nice moment of entry.
Priten: Um, I, I'm kind of curious now to hear your perspective on the tech, from 3.5 all the way to now, five, and how you're viewing it for this fall.
So, I'll wanna hear a bit more about the journey there, but why don't we just jump to the fall for a second, just because we're so close to it., How are you thinking about, the new, education moves that these companies are making,
Justin : I'm somebody who's more of an open AI user than anything else.
I'm certainly aware of Claude and Notebook.
LM is something that's had a bit of traction for us at our school.
I was in a room in New York City with a bunch of ed tech folks and AI folks the spring before I arrived here.
So I think this was spring of 2023 and they were sort of predicting these things were going to be happening at the timeline that they're in fact happening.
I think one of the things that is critical for schools and for really any aspect of this is like we don't really have a shared experience.
if you're using different levels within a platform, it just fundamentally, it shapes your experience with the thing.
So whether you're a pro user or a free user, you're gonna get a different experience.
Now I've been playing with GBT five for the past few days.
That certainly does seem to the level of playing field a bit in terms of model switching.
But Think Pro is still better than Think, right?
So there's still the discrepancies that exist there, which does I think cloud and pervade a lot of the educational use cases as it relates to teachers and students and their understanding of what it is precisely because we don't have that shared experience.
Priten: Right.
Justin : And it's so different.
it's just what do you have access to and what are you engaging with on a daily basis?
Priten: Right.
and I think the narratives about the difference in access across the country, but the reality is that even within the smallest microcosm,
There isn't a shared experience, right?
So even if there's a particular tool that your school is offering, the chances that every student is only using that tool is is, is null.
They're definitely going home and trying out the new TikTok bot their parents chat GPT Pro account or, talking to that Snapchat, abomination that, you know, is scary.
that's a great point that like no matter how much of a shared experience we.
It feels like in the big picture, like we're all interacting with ai.
the individual experiences we're all having are very, very different based on even just what tools we're using, let alone like our proficiency with them.
What does that mean for how you approach it?
in terms of the school?
Do you get a, um, what, what say do you get in fixing some of this or making some headway towards it?
Justin : That's a good, the way you frame that.
the way our CTL exists in our context is we sort of like orbit the day to day.
Which is kind of nice.
So we seek to have lots of influence while having very little authority.
Priten: Yeah.
Justin : certainly the school does leverage my burgeoning expertise to be like, Hey, what do we do about this?
But oftentimes teachers are looking for simple policies for what is a really complex situation, and policies are gonna be antiquated.
If I write one today, it's antiquated tomorrow in all likelihood, right?
Um, I do think a lot about, I've always said from the start, this is an potentially incredible, powerful tool for teaching and learning if deployed correctly that fundamentally disrupts most assessment, particularly in humanities, but in other fields as well.
I'm a human humanist by trade.
I also have 10, eight, and 5-year-old boys at home.
So I think a lot about their own education and the, the world, they're going to begin navigating increasingly in their schooling.
you know, what does it mean for us?
I, I don't know.
Matthew Rascoff has a great quote at Stanford, "School is learning embedded in social experience."
Which, which I think is so true.
Like the learning is often tangential, and kids love going to school for the most part, I think because of their friends and their sports and their arts and they learn along the way.
and I'm the teaching and learning guy saying that, but I think that's just the reality of things as they go.
Um, I, I do think there's a very clear tension of like, am I teaching history or am I teaching history plus AI use?
Priten: Right?
Justin : Like I, that's a thing that I'm thinking quite a bit about.
And I think there's this sort of belief that.
Students will pick up these tools through osmosis if we are using them well as adults.
And I'm just not sure that that's the case.
Right.
So, for me, I try and do things intentionally to be like, Hey, you're using AI now and I want you to use whatever you want and however you want.
But I want your thinking to be made visible.
So whether that's through sharing transcripts with me, we do use a tool called Flint, which is sort of a walled garden AI approach.
Which I really like from a teaching and learning lens because I can design assessment in there.
I can design really great homeworks in there.
Like if there's any quote unquote solution to huge air quotes, like I, that feels like a really good one to me.
But we don't have a lot of adoption amongst our teachers 'cause you have to change workflows that you've been engaged in for 5, 10, 20, 30 years.
And it's the changing of workflows that's really disruptive.
Priten: How much of the narrative from teachers who come to you for help is about detection and policing versus innovating and figuring out next steps of assessments.
Justin : we tend to be pretty self-critical in our CTL of always like,
What can we do better?
It's really interesting.
I think this is because of the influence versus authority piece, we don't really get anything around detection.
We get a lot around the innovation piece, like we are kind of seen as an innovative space where people are encouraged to try things out and to differentiate we do a lot of action type research, what they call the scholarship of teaching and learning.
In higher ed.
We do a lot of that with our CTL, which is unique to PK through 12.
So like we're gonna do a lot on the AI front that in that way of like, Hey, what happens if I replace one AI homework or one homework with AI homework a week?
Does that improve learning?
small bite-sized studies to understand the impact of these tools in our context.
we don't get a lot of the like, help me with detection and I, that is a voice where like I'm pretty strong pushing back around detection aspects as it relates to AI for the variety of reasons that we know, you know.
Second language, non-native English speakers, et cetera, et cetera, et cetera.
And gbt five like I will say in my early understanding of this, it can really replicate student voice quite well.
I had trained up all the previous versions so much on my own that it was replicating my own voice to the point that I was like, I don't really need to tweak this at all.
But now a couple of essays from a student, you're getting all their ticks and all their things,
Priten: Now I'm curious to hear about the experiments that you're doing.
when do you decide, Which interventions you will test.
what does the process look like at your school or even at the center in terms of, evaluating the safety, evaluating which students are gonna get the intervention versus not get the intervention.
Just kind of what is, what are the ethics of that design?
Justin : Great.
Really great question.
we like to call them investigations and not experiments.
'cause when we say experiments, people freak out.
It's very small scale, right?
Yeah.
Like it's not, we're a small school by and large.
They're sort of vetted through me in terms of who gets the intervention and who doesn't.
we're very cognizant of that.
Like I, I didn't use an AI specific one last year, but I will this year in my class, I only have one class.
Right.
So it's basically gonna be like, we run it this unit, we don't run it this unit.
So like we're never gonna get the randomized control trials, but I'm not sure that they're good for education, um, in schools, right?
But the ethics thing, like you're asking about ethics, that's first and foremost to us.
if this intervention stinks, we don't want it to impact this one class and not this other class, right?
So in a perfect world, I would have two classes and I could run it with one and then flip it with the other.
But I just do it with the control group in front of me and shift whatever that element is.
this is something we talk a lot about, not even related to ai.
Like oftentimes we do things and we don't know if they work or they might not work at all, but we just have done them and we believe they work.
But like, oh, the kids really didn't do well on that.
They must not have done their homework type thing, This is really great 'cause it allows us to go, Hey, here's the thing that we've always believed, and we can test it and be like, oh, that doesn't work the way I think it works.
How can I tweak it in this way?
So ethical stance, it's always like, it's always first do no harm.
And if there's anything that sort of illuminates, an issue, let's not do that.
that's where like the really small bite-sized stuff comes into play.
'cause it's, it's like we're gonna use an AI generated slide for this unit and not gonna use AI generated slides for that unit.
Like even just that tiny.
So then it feels the stakes are lower and I'm not gonna like be publishing these things in journals.
But I will share them on our blog and be like, Hey, gang out there who follow us, be on the lookout for this.
And we present, presented various conferences in that regard.
So the ethics are at the forefront.
We're never going to undertake a major intervention of like, let's give everybody access to GPT five and see what happens.
But we also know that many of our students are well resourced and have access to it, Part of me is like, maybe we should.
That's where Flynn is.
Nice.
Because we have, I can see how they're using it.
I have access.
I can design in there.
And in theory, they're using the AI in there to support, and it's really strong AI right now.
I also know that some students are gonna be like, I'm just gonna open my parents' Pro account and then answer it with the pro account and then let the AI gimme feedback there.
Priten: when you say do no harm, I'm curious, what does harm look like for you in this context?
Justin : I don't have a really strong answer for you there.
I think a lot about this in our contexts, and I've only been in indie schools, but when I was leaving my previous indie school, parents were starting to ask the question, what are you doing about ai?
And that question meant two things.
It meant what are you doing about AI to ban it, and what are you doing about AI to make sure that my kids are not falling behind?
Right, because that's also a form of doing harm.
Jeff Slingo is a very persuasive voice in our spaces and he had a recent piece about ensuring students are ready for the workforce.
And I'll push back there and be like, we're pre-K through 12.
Our kids aren't going from 12th grade into the boardroom or into the operating room.
we have a ton on character education at our school.
So like we're preparing humans to take forth this robust liberal arts education to wherever they might go, college, university, maybe even work.
So I'm not sure that we are compelled to prepare them for the workforce, right?
But I think it does, like my personal stance, and this is, I'm speaking with an I here, I do think we're doing harm if we're not in.
Forming students about these tools and deploying them in some instances to give them a taste and a flavor so much for me is just like, I think so much about the
science of learning then that's not a unified science clearly, but understanding of how learning happens is really, really crucial if you wanna deploy AI tools well.
I think it's unethical to not have an understanding of the science of how learning happens.
having had that and seeking to pervade that throughout our community often informs the thing of like, is this doing harm?
The phrase that I love that everybody's using now is are we allowing students to cognitively offload the task?
Sometimes we are, but sometimes in doing that we're actually gaining something along the way.
But am I assessing their understanding of history or am I assessing their understanding of deploying AI well, to make it looks like they understand history, like that's right.
That question is always at the core of any intervention that we're seeking.
Priten: That last question is the one that I think, um especially for disciplines that are normally not tech heavy, am I doing this because, our school just got a bunch of funding for it.
Am I doing it because, quote unquote, my students will be more engaged.
am I doing it because students are more likely to see this stuff as relevant, if it's somehow like related to ai.
or are they still actually learning the content that I think is important, and I've always thought is important.
Justin : there's a guy at Penn, Robert Gris, who works in their math and engineering department, who's really, I think, using this well, he designs his textbooks with Claude.
He writes it with Claude and he says to students, use whatever you want to help you better understand these things.
'cause when you have to build a bridge, in the field, you may not have access to these powerful ais and they might make mistakes.
the stakes are really high there.
I don't love a blue book assessment, but I still wanna under, I still wanna assess historical knowledge in the moment, like, right, conversational competency is a phrase that we're using quite a bit.
We weren't always a text culture, now we're a text culture and LLMs disrupt the text culture.
Like there's a value of being able to have this back and forth,
Priten: right?
Justin : this is more important for a job interview, for talking to somebody at a bar or a restaurant like this is important.
So can we begin to assess in this kind of space and in these ways, I'll certainly do more conversational, and I've been doing it before ai, but I'll certainly do more and I can leverage Flint in a way that allows for students to engage verbally in the moment.
And even before I left St. George's, I had them have this conversation with me as part of their final exam, and the students hated it.
I'm like, well, then you don't know the thing that you've written 5,000 words like you don't know it.
And I want you to know it, It's funny, like as a historian, I had a piece of my personal blog about like my kid who's 10 knows way more about history than I do, but he doesn't read nearly as much as I did because he's consuming YouTube videos constantly.
And they're good YouTube videos.
he's watching good history videos.
The primacy of text is probably shifting.
It's been shifting for a while.
I still think there's value in reading and writing.
We are a house filled with books, but he knows a ton about history, like, and it's kind of cool to watch.
Priten: is the evaluating of whether or not they're good YouTube videos, something that you think you've taught him that he's picked up on his own or are you doing for him?
Justin : Great question.
it was a lot of training along the way as he really got down the algorithm of crap.
And he has two younger brothers and he will get frustrated with them and he'll let me know.
So I'm doing a lot of metering and gating on my end his critical thinking about what makes something good and what makes it bad is we spend a lot of time on media literacy in that regard.
Again, I'm raising three boys, like in our current environment.
Like, I don't want to contribute to toxic masculinity stuff.
Right.
Uh, he's really good at it.
With that said, as he gets a little bit more sophisticated, he's getting into some videos that are maybe a little bit off.
So we're talking a lot about critical thinking and what it means to think appropriately not just his rote verbalization of things he's just heard.
I'm like, okay, why does that matter?
What does this tell us?
What does the perspective of this person?
Super interested in current events, so like it's a rich landscape of foreign policy right now in our house.
Priten: Very cool.
and then when you think about the tech usage in general, there's obviously a loud course in the last few years of like the psychological impact of devices on students and some debate about how valid that research itself is.
I'm curious about how you're approaching like screen time and what age they got the devices at.
Um, because I mean, obviously they're using a productive loop, but I'm curious what your philosophy on that is.
Justin : Yeah.
I mean they're, but they're not at the start of the summer, our 5-year-old, he was addicted to the iPad and I just took it and put it in the highest cabinet possible and hid it.
And it was, he was frustrated at first.
But wouldn't you know it like two days later he would bring his toys upstairs from the basement and he's like, I like not having the iPad.
I'm playing with my toys more.
Like even he noticed the moment of like this difference.
I'm somebody who's attached to my phone all the time, but I'm attached to it 'cause I'm just reading good things all of the time.
Like I'm just constantly trying to learn and stay informed.
We won't go cell phones before middle school, certainly, probably, you know, we're looking at 13.
I don't love hate's book in that regard.
I think it's probably overstated and the data's not, doesn't really support a lot of what's in there, I think.
Priten: Yeah.
Justin : But we instituted a cell phone ban in our upper school last year.
a new paper came out that I just shared with our leadership team it does seem to support student learning with cell phone bans.
I'm also the person who, might want to eat lunch by myself and I might be on my phone, but I'm reading a great academic journal article.
Right.
And I feel like if I were sitting there with a book, you would be okay with that.
I think this is reflective of society in general.
There's not a lot of room for nuance.
That's frustrating to me, but as a parent, we're not, their screen time is like, they play a lot of video games, but they play really cool indie video games, or they're building Minecraft or they're like building worlds in Mario Maker, I'm okay with that.
They're talking to their friends, they play Fortnite and they play Madden.
But like, I grew up on video games.
it gets to a point where I'm like, Hey buddy, you've been playing video games a long time.
Let's go for a walk.
Let's have a catch.
Priten: When you think about having the conversations you can have with your children is obviously very different than the conversations that schools can have with students.
are you hearing, pushback from students about like the cell phone ban?
Because I've heard stories of students saying like, gosh, like somebody take my cell phone away.
Right?
Like, I, I need, like, they're, they're in favor of these things because they know they can't self-regulate.
then of course there are other students who are just like, this is stupid, what kinds of conversations are you all having?
Justin : I think the trend is more towards the former rather than the latter.
initially there was a bit of pushback, but it was relatively minor in terms of the vocal pushback.
And I think there's an appreciation because they're untethered from things that can be very sort of addictive.
An anecdote that I think supports this really quite well is at St. George's.
We used to have a semester at sea program with our boat and students didn't have cell phones for like four to six weeks they loved it.
They talked about just being there in the moment.
I think there's really great value there.
again, I was like agnostic around the cell phone ban personally.
Because I think part of what we're doing is trying to have them be self-regulating.
but at the same time, if I have ice cream in my house, I'm gonna eat all the ice cream so I don't buy ice cream.
Right,
I think that's a real thing.
particularly for adolescent development, you know, middle school that had been banned previously, they just dropped them at the front.
I think particularly for the students who have grown up having screens and cell phones for much of their lives, there's a real value in taking a moment to just be like, Nope, we're gonna pause this for the day.
And you just need to be a student or an athlete, or a musician or an artist.
Priten: The ice cream analogy is a good one in terms of thinking about like autonomy, because I know some folks are there, there is
obviously folks who think that we need to teach students to have healthier habits and use them productively and learn how to put them away.
but sometimes we as humans do just need a little bit more coercion, whether self-inflicted or externally inflicted to do the right thing.
Justin : We do a ton on pedagogical partnership with students.
in our kind of context.
Students just wanna be told what to do so they can go to the next thing and get into the college university.
And like, I, I think they oftentimes forget that they're a huge part of this.
I could design the best homework in the world, but if they don't do it, it's not gonna have any impact.
when you see them that way, it changes the dynamic.
'cause then they see themselves that way.
I think that's particularly useful in indie schools right now, where maybe there's this pressure from parents and that sort of thing.
but I think it's useful everywhere, like even having conversations with my 10 and 8-year-old to give them some agency and autonomy is really useful.
'cause parents are doing so many things for kids.
Our kids are like kind of latchkey kids.
They get on the bus by themselves in the morning.
'cause we leave work early and they get home by themselves and people are like, they're, they're so young.
I'm like, yeah, they can do it.
It's okay.
It's an electronic lock.
We have cameras, we live in a very safe neighborhood.
They'll be fine.
Priten: I have two more questions for you, and then I'll, make sure I respect your time.
I'm curious about your perspective and in terms of when you're thinking about the science of learning, and you hear folks talk about we need to meet our students, where they're at in terms of their attention spans.
That's one that I find distressing myself, but I don't have the background that you do.
Um, and so it's like, we need to be the TikTok classroom and like things need to be in 60 second chunks.
Like, because that's what our students are used to consuming the high dopamine, high speed content.
is that what innovation in schools is gonna look like going forward?
Is it gonna be like, can we be little TikTok stars in front of our students?
Justin : I was chuckling when you first asked that question because we're doing a strategic plan right now and we send out a weekly internal newsletter that we spin off into our substack as well.
And there was some complaints from the teachers that we are writing too long.
Priten: Yeah.
Justin : we send out three to four sentences per resource that we send out.
And people are like, the emails are too long.
sometimes our adults are complaining about our kids while also not modeling the behaviors then.
Right.
That's, that's a handful of instances, right?
Yeah.
So yeah.
No, I don't think we can.
But with that said, like we're also adopting a micro learning platform called Seven Taps.
Like to have them have a little bit more bite-sized, digestible access into these things.
I think there's an element of realness to that, but I think there's also value in reading deep, rich, complex things or engaging with, you know.
Deep nuanced problem sets.
Pick your discipline.
in terms of the science of learning I think micro learning platforms can have some value, particularly on like retrieval, practice nudges or like some space practice along the way.
Yeah.
I think AI's really good there.
If people aren't really familiar with science of learning vocabulary and jargon and they sort of do it, but they don't know they're doing it, you can be like, Hey, help me make more interleaf practice with my syllabus.
Where should I be leveraging certain things along the way?
that's a great simple blunt force prompt that can let people do things better.
I don't think, if I wanna understand Vicinities, I can't do it in a seven taps.
Like I have to read it and deep dive.
We are sort of tail wagging dog as kids chase colleges and universities.
they feel like they need to do more, more, more faster, faster, faster.
I, I worry a little bit about like preserving.
Time for kids to just be kids.
we're trying to accelerate their development in every capacity, whether it's athletics or academics.
as a dad, I'm certainly pushing back on that quite a bit.
I don't think we can Tik Tok our way into the science of learning things, but we can dabble it and sprinkle it in there as a novel sort of engagement piece.
I think that's totally fine.
And we try and do that.
Priten: Yeah.
How do we get our kids to read more books?
Like that's the more I talk to teachers across the country, I'm just like, folks are like, we give passages now and no longer even require them to read the full books because we know they won't.
Justin : There's a phrase I say all the time, like, you're not stuck in traffic.
You are traffic.
in some instances, we are the traffic.
Again, my context is in independent, not public school.
we don't have the prepackaged curricula.
Our students are being assigned books.
I do think their attention span has shifted, but like I think it's our job to meet students where they are.
And to shift them back if we value it and can articulate the why behind it.
Yeah.
Like a good example of this is I use a traditional textbook now, which I don't love.
'cause I used to use a digital text called the American Yop.
Priten: Yeah.
Justin : Free open source written by college professors all over.
We helped AP integrate it into their curriculum, but I was like, all right, I'm not just doing this to have you screen read.
I'm doing this because we're gonna layer over hypothesis social annotation.
Priten: Yeah.
Justin : It changes the dynamic 'cause now I don't need to do a discussion board post.
The discussion board is embedded in the text every night, and students who are introverts and don't wanna talk in class are writing
these beautiful paragraphs, doing synthesis and analyzing and critically thinking, and then a student's responding to that in the moment.
And then I wake up and instead of reading New York Times at seven 30 in the morning, I'm reading all the annotations.
Priten: Yeah.
Justin : And then I'm tweaking my class in the moment or before I head in that first period to be like, oh, you guys really struggled with this today, so let's talk about this.
Like, let's project, you know, gin's, annotation on the board and like let's everybody have at it, annotate it again now together.
And now let's jump off and talk about Shay's Rebellion When I said at the start making student thinking visible that wasn't for, and people were like, they're reading on screens.
I'm like that's meeting them where they are and they are doing way more.
And that's where there was this instance of causal correlation.
the students who did the most annotations performed really well because they were very clearly close reading.
if they want a PDF bounded version for a nominal fee that's not $300, they can also get that, but they still have to annotate in the thing.
So we have this social experience with one another.
I miss that a ton as a teacher, and I'm using a traditional text now.
any chance I have to have them read digitally and use hypothesis, I use it.
Priten: Yeah.
that's a great example of hitting all of it, right?
Like you're, you're getting them to read slowly.
You're still using the tech but you're also encouraging the social interaction and encouraging them to actually grasp the material
Justin : I have them read JSTOR articles that I read in grad school that are above their head, but now I'm gonna take that and put it in Notebook LM and have that generate a podcast to give 'em a different layer of access to it.
And then I'm gonna have them create their own notebooks of like, Hey, find some supporting resources that would nudge this in a particular direction that you think it should go in.
I'm not.
AI everything.
I'm just like notebook.
LM is a really useful tool for you guys,
And this is a great way to use it.
I did it last year when I gave them five primary sources in small groups.
They were meant to turn and talk in groups about the primary sources, and then come back together and think about the themes we were talking about.
it was post lunch and I had 12 boys and three girls, and they were just like, it was a Thursday.
So that meant they were setting their fantasy football lineup.
They weren't doing the thing I wanted them to do.
Right.
Whereas if I give them a turn and talk in Flint, Flint isn't gonna let them do that because their partner is this AI that's informed.
Right.
And I took all those primary sources and I upload, uploaded em into Notebook, lm, this is when it first came out, and I said.
Produce the podcast.
So we listened to the Seven Minute podcast and they were like, what just happened?
I'm like, yeah.
That they like, how did they know these things?
'cause they didn't know it was AI generated.
Right.
So then there was this moment for them of like, what is this?
We can do better than that.
So the next time we did it, they were like, we wanna compete against the ai.
We can do better.
Like this little novel moment of being provocative with them.
And getting them to hear and think and listen and read like imperfect, but really, classrooms are really imperfect and messy spaces.
Let's try this thing out.
Let's see.
there's something there.
Priten: yeah, we hear from a lot of teachers that what they're getting is, top down guidance that they have to use the tech in some way.
And I think once you're thinking about it from the, like, how do I integrate tech rather than like, oh, what need am I solving?
Obviously that flips the dynamic quite a bit.
yep.
So it's, it's refreshing to hear the, oh, here's, to have the tech helps me achieve a goal that I already have.
and that's just the unfortunate reality of how school systems across the country are and not any individual instructor's fault.
Justin : I have to shout out a guy at Loyola College right now, Patrick Dempsey, who leads their CTL I did some work with him this summer just to help me sort of orient my space and understanding of this.
He's got some really great protocols and needs, finding exercises that we're gonna leverage here to help people think about AI and tech integration and what it's solving for them.
but I also know that When my kids ask me to get down and play with Transformers I struggle at it 'cause I'm 42 and I forget how to play.
But I also know that it's incredible learning that happens for them.
when I ask people to play with this empty chatbot window, it's really weird.
this alien intelligence sitting in front of us.
But I also know that play is really useful and designing some playful experiences because then all of a sudden they're like, oh, I didn't even know I had this need, but now that I know what it's capable of, I can leverage it this way.
Colleague in CTL, he plays the organ for us.
He was talking about trying to find a transition in our schools chapel.
He's talking about trying to find a transition between these two musical pieces.
He was struggling with it.
He could have figured it out on his own in an hour.
we might say that's a great cognitive engagement exercise.
he said, no, like I didn't have time to figure out this transition.
So he just uploaded the music into AI and found the great little bridge between these two things.
And then to his credit, because he is open and thinking this way, he's like, oh, he teaches Spanish.
He goes, I struggle with transitions in the classroom from going from conjugating verbs and doing grammar into this reading.
And he goes, let me think about these sort of 32nd, one minute transition.
So he uploads his lesson stuff into ai.
And he gets ways to leverage the exercise that they've just done into the new exercise.
He could have gotten there on his own, but like that was a good moment.
We talk about this all the time.
School's always gonna look to schools for inspiration.
And I'm like, no, no, no.
We gotta look elsewhere.
We gotta find adjacent inspiration, whether it's restaurants or service industry or sports, whatever it might be.
And this was this nice moment of adjacent inspiration that's gonna allow, that has allowed him to embed better transitions in his language classroom utilizing ai.
But how many teachers think about like, what are my transitions like?
Like we're CTL people.
We can think about this all the time, there are so many moves we might make that can have an impact.
now I'm going like, oh, we should study that.
Does the use of positive transitions support student learning?
Priten: Very cool.
awesome.
I. Really appreciate your time today.
I'll end on like a, a big picture note of when you think about where the tech is going, what is your biggest fear and what are you most excited about?
Justin : Yeah, you know, the fear thing is real.
when I talk and work with teachers, they are fearful and I think, I don't think schools oftentimes do a good job of acknowledging that fear and what that means.
My biggest fear is honestly this, I think we're already there.
The amount of AI slop that's out there like this just.
Terrible road of middle, but in a way, and I think this is in my nature, I'm like, okay, great.
Having a more unique voice and perspective then only becomes more illuminated because everybody's defaulting to this AI junk, everybody talks about the M dash.
I've been using M Dashs forever.
I'm not ever stopping my M dash usage.
But when I see the, it's not this, it's actually that construction.
I'm like, oh my God.
and once you see it, it's like seeing, white cars That's all I see.
my fear is us giving up our humanity and defaulting to this middle slop and society not having enough taste to be like, this thing is excellent and this thing is not.
Right.
So do we have enough knowledge and understanding of the world?
Can we make enough connections and our particular disciplines or across disciplines to be like having taste and being like, yeah, this thing is a masterpiece and this thing is sort of novel and it's wild
Like it's not the same.
But it's a nice approximation or facsimile of that thing.
That's my biggest fear.
And I think just given the amount of digital stuff that's out there, I think we're already there and it drives me a little bit crazy.
, I'm a historian, so I collect quotes and think about other things.
I think it was Aha.
Gold from Duke presenting at Georgia Tech a few years ago.
I say this to my students all the time, you often don't know when you're living through history.
But we are undeniably living through history.
Yeah.
And again, another quote David Alder says, like, the future's not a prediction problem.
It is a design problem.
we have a chance to design whatever we wanna design, but I think we're, we're still going like, oh my gosh, the updates, I can't keep up.
what do you want?
and I think it's come through the way I described my classrooms.
here's what I want.
So I'm gonna just make that thing.
we can make whatever we want.
I was using agent mode today to upload my assignments in our Canvas calendar for the year.
I don't like doing that.
Instead, I just uploaded the calendar and now it does it for me.
that's what I want.
Priten: Yeah.
Justin : Are people just similar, around taste?
are we willing to have enough agency to actually create the thing that we want, as opposed to just whining and complaining against the, the tidal wave that's coming.
Build some walls.
Build some bridges.
Put your house on stilts.
To make the thing that you want.
we feel like.
We're like showing up to this.
So helplessly.
some of that starts with simply learning and understanding.
And I know it's everybody's busy, you know?
Like I get it.
I understand I have the space, but even when I was teaching and doing academic admin stuff, I was still doing it because it was just learning.
And that's why I work in schools.
'cause I like to learn.
So let's just learn and let's design the future as we wanted, as opposed to just letting it be made for us.
I clearly am passionate about this stuff
Priten: That's perfect.
That's what I want folks to hear.
I'm sure that, inspires somebody else to think about it that way because I think there's so much defeatism as well from folks, and I think hearing
your voice will hopefully be refreshing Thanks to Justin for joining us and sharing the administrator's perspective on these impossible decisions.
Justin showed us that ethical EdTech isn't about finding the perfect solution.
It's about asking the right questions and building decision making frameworks that account for context, values, and consequences.
His work demonstrates that when we, his work demonstrates that when we put pedagogy at the center and start.
His work demonstrates that when we put pedagogy at the center and stay humble about what we don't know, we create space for better choices.
Keep listening as we continue exploring the ethics of education technology.
And remember to pre-order my upcoming [email protected].
Thanks for listening to Margin of Thought.
If this episode gave you something to think about, subscribe, rate, and review us.
Also, share it with someone who might be asking similar questions.
You can find the show notes, transcripts, and my newsletter at priten.org.
Until next time, keep making space for the questions that matter.