AI and the Future of Education: Teaching in the Age of Artificial Intelligence

[00:01] Announcer:

Welcome to Principal Center Radio, helping you build capacity for instructional leadership. Here's your host, Director of the Principal Center, Dr. Justin Bader. Welcome, everyone, to Principal Center Radio.

[00:13] SPEAKER_01:

I'm your host, Justin Bader, and I'm honored to welcome to the program Preeten Shah. Preetan is a serial entrepreneur whose projects aim to drive change and innovation in education. He holds a BA in philosophy from Harvard College and an MED in education policy and management from Harvard Graduate School of Education. He currently runs two education ventures, United for Social Change and Pedagogy.Cloud, and he's the author of AI and the Future of Education, Teaching in the Age of Artificial Intelligence, which we're here to talk about today.

[00:46] Announcer:

And now, our feature presentation.

[00:48] SPEAKER_01:

Preeten, welcome to Principal Center Radio.

[00:50] SPEAKER_00:

Thank you for having me.

[00:51] SPEAKER_01:

I'm excited to talk about your book because this is a difficult topic to write a book about because it's changing so fast. What did you see happening in the field of education and specifically with developments in artificial intelligence that prompted you to write AI and the Future of Education?

[01:06] SPEAKER_00:

Yeah. So this is technology that I've been really excited about for a few years. I did some work in college on AI and language learning. We've been talking about hypothetical uses of AI technology ever got there for the last few years. And then suddenly, I guess a little bit over a year ago, the technology started to get there and I became excited about the potential for using it in education. Except a lot of our clients and my fellow teachers were really scared.

[01:28]

And a lot of the reaction was fear, anger, and concern. And I was disappointed. I was hoping that when this technology finally became possible, we'd be talking about exciting things. But I also understood there's really good reasons to be scared. There's really good reasons to be upset. And the book is kind of to talk about both those sides.

[01:44]

I hope that we can all think critically about these technologies, think about how we can make sure they're safe for our students, and how we can make sure that our goals in education don't become wasted, but also so that we can make sure that we're racing it to better meet those goals. The book takes a middle line stance between those two.

[01:57] SPEAKER_01:

In a way, many of these concerns go back to the dawn of Google and the ability to look information up online. And as a profession, we've adapted to the easy ability to Google things by having students show their work and assigning more complex tasks that can't just be quickly looked up online. or the kind of calculator proofing that we've had to do for many decades now since students have had calculators that they can use to help their math homework. What is similar and what is different about AI compared to, say, calculators and Google in ways that impact education and the kind of tasks that we give students in the course of their learning?

[02:35] SPEAKER_00:

Yeah, I mean, those are really early precursors to the kind of problems we're facing, right? There are external technologies that can actually do the work that we're asking our students to do. But I also think that there's some big differences here. I think that partly the differences are becoming in terms of the rapidness with which these technologies came about, and the rapidness with which they're evolving. And so this is where, while we can think about scaffolding the things we want to teach in education the way we scaffolded math, and so we wouldn't allow kindergarteners to use a calculator to do basic arithmetic, we definitely allow our 12th grade calculus students to use calculators to do even more complex math, because that's not exactly what we're trying to test at higher levels. And there are very similar things we can do with AI in terms of writing, in terms of research.

[03:12]

But at the same time, I think the bigger problem folks are facing is what does out of school work look like? And how do we justify to our students to continue to do this? And again, very similar analog to calculators. We've had to explain to students why they should memorize their times tables if they have an iPhone with their calculators on them. And we'll have to similarly continue to explain to them why they need to still write well, why they still need to think critically about how they're doing research, where information comes from, how to synthesize information, all those things that AI can, quote unquote, do now. We need to now do a further disciplinary process we need to have with our students to explain to them why it's still important for them to do it.

[03:42]

At the same time, we kind of had some time to evolve after each of these things. And so the calculator came out and the technology didn't have to dramatically shift in the next coming months. Google search came out and there was a moment where we had to rethink a lot of these things, but then it plateaus for a little bit. With the AI technology, we see it every month. The new interventions come out. that make whatever solutions we came up with irrelevant.

[04:00]

And so last year we saw folks saying, oh, it can't, it doesn't know anything up after 2021. So maybe you can assign things that are more related to current events or it doesn't write like a particular student's tone. And so we can make sure that we have some in-class writing we can compare it against that. But these things change overnight, like the within two months, they had access to the internet. And now as you put custom instructions that you can write in a particular tone, knowledge will continue to evolve at that rate where we won't be able to just come up with these short term solutions. The way we have in the past, we've really had to do a fundamental rethinking of what we're doing.

[04:30] SPEAKER_01:

Yeah. And I think especially the ability for ChatGPT to produce college level writing, I think really caught a lot of people off guard that this essay that previously would have been hours of work for a human to do is now 10 seconds with ChatGPT and then maybe a little cleanup on the student side. Also some fact checking in case the output is fabricated. And certainly there are some issues that are still to be resolved with that, but I expect that in time. The factual accuracy issues will be resolved. You said something about our ability to adapt because of the timeframe, like calculators were designed and invented, but they were very large, very expensive, and only gradually became cheap enough for students to actually carry to school with them.

[05:12]

And yet somehow graphing calculators are still $120 like they were when I was in high school. And what we've been able to do as a result of that time that we've had to adapt is figure out how to still give kids tasks that are meaningful, that don't allow them to over rely on technology in a way that prevents them from developing the skills that they need to. How have you seen that catch us off guard when it comes to generative artificial intelligence, the ability to do things vastly more complex things all of a sudden. What do you see as some of the key points of our caught-off-guard-ness?

[05:49] SPEAKER_00:

Yeah, I think a huge chunk of this is the kind of things that we've been assigning to students that they can do outside of the classroom. I think that the take-home essay, take-home short answer questions, all those kinds of things that they can now do overnight, like you said, at pretty high levels of accuracy, at high levels of mimicking the students' own writing, But at the same time, I think there's some work that we need to do better at, which is justifying to our students why they're learning these things in the first place. And I think that that's where I think we really got caught off guard, right? So I think at the end of the day, there's really good reasons what the technologists can do with these things that's in the classroom. There's ways that we can bring those things into the classroom. There's other interventions we can maybe do.

[06:23]

But the fundamental thing is that we're students to be convinced that they need to learn these things, right? And that's not just me preaching here. I know that's an impossible task, especially in large classrooms with a diverse set of students, convincing every single student in your classroom that learning how to write your essay yourself is a good idea. It's not an easy task. But I do think that we need to do a shift towards doing more of that participatory work so that our students are bought into the idea that maybe the process of writing an essay is important. And I think there's also fundamental things that we can change about how we structure incentives in school systems in order to do some of that work as well.

[06:53]

And right now, a student at 3 a.m. with the essay due the next day has a really high incentive to cheat and use chat GPT to write that essay because they don't want to fail the assignment the next day. And if we can shift some of that emphasis from being, what is your final work output? And we're going to grade that final work output to how can we incentivize the process? How can we incentivize the journey over whatever this final work output is?

[07:11]

I think we can get closer to making our students understand why we still have them write a five paragraph essay at the end of the day.

[07:17] SPEAKER_01:

So focusing more on the process, building in accountability, building in checks, like what is your topic? What is your topic sentence? What are some of your sources? Things that a lot of writing teachers have done. And I remember taking college writing classes where there was process to it. It wasn't just turn in an essay.

[07:32]

And because even before technology, you know, there was always the option of have somebody else write it for you, your big sister or paying somebody or whatever. bullying somebody into doing your homework for you. So like we have precedents for thinking about these things. Take us into, if you would, a little bit of the technology itself, because I know in the book you help us understand what AI is, what machine learning is. What do you find that some of the key misconceptions that educators tend to have are and just how does this stuff work at a fundamental level?

[08:06] SPEAKER_00:

We have seen a variety of exposures to the technology and the right exposures. to even just digital literacy. And so we see misconceptions at every level. And so we see folks who still don't know where to put prompt in. So we have folks who are trying to put the prompt into the Google search bar instead of into the chat messaging system within a website. And it starts there.

[08:24]

And I guess the folks who think about limitations of the technology, I think those are permanent. There's still folks who think that it can't access information after a certain date. There's still folks who think that it can't mimic a particular student's writing. Oh, I always know when it's written by chat GPT is something we hear very often. And that probably was true in the early days of the models that started to go away. So I really think it's about staying up to date with the technology that I think we're having.

[08:44]

Folks are just like, there's so much information coming out at them that try to figure out exactly what the capabilities are. It's definitely where we see a huge amount of misconceptions. And then I think the other part is there's an idea that this is a fad. That, oh, this is a technology that came out overnight. No one's going to keep talking about it. It's going to go away.

[08:58]

This is just like the cryptocurrency again. And this is very different, right? So while cryptocurrency had a similar surge of interest, especially across the tech industries, It didn't fundamentally change how we worked, how we thought, how we produced information. It really did work in a very isolated context. Whereas with AI technologies, it has an impact across industries, across different types of work that we do. And so I'm hoping that at the end of the day, folks can be convinced that this isn't going anywhere and it's only going to get better.

[09:22]

Those are the two truths that I think we need to operate with.

[09:25] SPEAKER_01:

Yeah, yeah, not a fad, not going anywhere. And just as calculators are not going anywhere, Google search is not going anywhere. Do you see the trajectory of change slowing down a bit? Have we already witnessed the big hockey stick of leapfrogging capabilities, or is that just going to continue to get wilder and wilder?

[09:43] SPEAKER_00:

Yeah, this is something I've been thinking a lot about the last couple of weeks, especially as the anniversary of ChatGPT came out. I think folks who are just trying to figure out what is this pace going to be in 2024. And I think that the pace of development of that technology will continue to be pretty rapid. There's really good literature in computer science about why as AI technology gets stronger, the pace will actually quicken rather than slow down. But I think that in terms of the effects it has in our daily lives, I think that pace might start to slow down a little bit. That gap between where the technology is in the computer science and in lab laboratories and in research institutes versus what we're seeing in our everyday lives, the gap will widen a little bit.

[10:16]

And that gap existed pre-Chad GPT-2. So this wasn't the first large language model. This was like there was a third version of even just Chad GPT's or GPT's rather model. And it took a bit for that to become accessible and available to everyday folks. And I think we're gonna see another gap again, where folks are working really hard in these research institutes to develop the higher levels of technology. They might taper access to the public.

[10:38]

They might not feel ready that it's ready for public access. And there'll be accessibility problems. This technology is really expensive. It consumes a lot of resources. And the next step is going to be the technology that already exists. How do we get into more hands in ways that are practical for more folks and sustainable rather than what is the next big thing that we can get everybody to use?

[10:55] SPEAKER_01:

It was interesting to see how a lot of the initial concerns from educators, especially college educators, were about cheating, having ChatGPT do your essays for you. And now we're seeing, I think, a little bit more balanced interest in actually everybody making good kind of calculator and Google style use of these tools. So talk to us about the educator kind of productivity side Where are some good opportunities for educators to think about using AI? Cause I've spoken with a couple of people who are extremely enthusiastic, very early adopter, very using it for everything. And most people are going to wait a little bit longer, but there are some hard to ignore opportunities. What do you see as some of the most promising opportunities for educators to make better use of AI?

[11:39] SPEAKER_00:

Yeah. And this is like where we like to start with most folks as well. I think folks have an easier time seeing how it fits into their workflow. In a way that's not damaging to student interactions. And I think that any type of content generation is definitely where we ask folks to start. And so everything from creating lesson plans to essay prompts to portions of your syllabus, all of those kinds of things can really be offloaded to AI systems.

[11:59]

And while the book has lots of different prompts and different ideas about every little thing that you might ask an AI system to do for you and some variations on those things, I think that the easiest way for most folks to start if they have waited out trying these technologies out and they're listening today and they want to figure out what do I do first because this is you know, interested in it, is to open the ChatGPT website and then ask it to do your very next writing task. Whether it be an email, whether it be a lesson plan, whether it be a worksheet, whatever your next thing you're about to write to see if ChatGPT can do, please get you 80% of the way there so that you can do that final 20%. And I think that's where folks will start to see that they'll save the amount of time it will save them is enormous. At the same time, that scares folks. And so this is where every time we tell teachers, this is what you can do for yourself. Teachers immediately say, oh, that means our students can also do this for themselves.

[12:43]

And so this is where making sure that we figure out like, The difference is in use between industry professionals and folks working a job and students. I think it's going to be part of the narrative that we also need to really focus on. While the goal for a teacher, for a lawyer, for a business professional is efficiency and getting to the final work product as fast as possible, for a teacher to find ways to make the lesson plans faster doesn't really impede. There's no cost to that as long as they're going in and doing the human check of the final work product. The process is very important for students. And so efficiency is not the goal for students.

[13:13]

And I think that's where we need to, if teachers can start to conceptualize that, I think they have an easier time explaining that to their students as well.

[13:18] SPEAKER_01:

So you're saying, actually, as a teacher, think about the assignment I'm going to give and then ask ChatGPT, for example, to do the assignment the same way that I'm going to ask my students to do it and see what it comes up with?

[13:29] SPEAKER_00:

That is an option as well. It can craft the assignment for you. So if you have an idea for, I have this creative idea for this project I've always wanted to assign my students, but I didn't have enough time to come up with the five different variations I needed to come up with examples I needed to come up with the rubric I needed. All those kinds of things can be offloaded to the system. But you can also give it an assignment you have made and ask it to modify it. Say, hey, I have a student this year.

[13:50]

I have a classroom this year with a lot more ESL students than I normally do. How about you modify this assignment that I traditionally assigned? Or how might I integrate technology into this assignment I've been doing for the last decade in ways that's more innovative and gets the students to be more engaged? Or solve this assignment as if you were a student and see what kinds of questions you might have if you were a student doing the assignment so that you can preempt the kinds of mistakes that students will make. So there's a variety of ways that you can approach using these systems and just like during your prep period, sitting there trying to make your workflow more efficient and get better work output that will better serve your students.

[14:20] SPEAKER_01:

I love it. Yeah. One, one thing that I've found it very helpful for is just revising, revising for reading level. Like I don't need to be using college level words. I can revise this writing down to a fourth grade level. I can translate it.

[14:34]

I can summarize. And one of the things I have to do fairly regularly and fairly soon is submit conference proposals. And those have a word count, little things like that. Take what I wrote and get it down to 50 words. Lots of neat things that can be done there. Well, Preeten, one thing that we have long heard about in education technology is the digital divide, right?

[14:55]

The availability to access devices, internet connectivity, high-speed internet, and now AI. What do you see as some of the challenges around that digital divide where some students are familiar with and have access to the latest tools and some students don't? And what implications does that have for us in schools?

[15:11] SPEAKER_00:

There's so many aspects to the divide here that I think are really concerning. And I think we saw some of those even very early on. When the school district started responding to the advent of the AI technologies becoming more accessible, the New York City ban on chat GPT within their school devices was a great example for me of how the device could get worse. So we had students who only had school devices who no longer could access these systems. And then we had a bunch of students who also had other devices at home, their parents' devices, their cell phones, their siblings' device, that they could go home and now access these technologies that they heard about through the ban. And I think that we'll continue to see those kinds of things where we'll have students who have some sort of access, either whether it be outside of the classroom, whether it be schools that embrace the use of AI in their classrooms.

[15:51]

And so we'll have teachers who embrace AI literacy, figure out ways to incorporate it within their assignments. And then we'll have schools who continue to enforce BANs, enforce viewing AI only as a plagiarism tool. And then we'll start to see a gap with the literacy skills to use the technologies between these students. But we have even more fundamental problems. And so if we have problems between basic device access, so it's not just about whether or not you have a school device or non-school device, there are lots of folks who don't have any device. And so there's an internet bandwidth gap across the country where not every district, not every town has the high level of internet that major metropolitan areas do.

[16:22]

And there's also literacy gaps between what kind of conversations the teachers can have with their students. And so until professional development for student teachers across the country becomes a little bit more caught up with the level of technology there is, we'll have teachers who can't think about how to approach these technologies and explain to their students the use of these. And so the gap between those who are prepared to use these technologies in meaningful ways in their careers, which I think can be taken for granted as a necessity as we even see industries adapting from the last year, let alone by this time these students graduate high school and college, and then we'll have students who are entering the workforce extremely underprepared to utilize those. There's also concerns about when we start implementing these technologies as helpful for students. If that process isn't scaffolded, we will have students who are skipping some rungs because they can now offload some of the tasks that they really need to learn how to do before to these AI technologies. And one of my fears is that it's very easy to think about, okay, access to devices, access to internet.

[17:11]

We also need to think about where students are in their learning journeys before we implement AI technologies and allow them to access these technologies and use them productively because there will be students who will never learn fundamental critical thinking skills, their fundamental ability to think, reason, synthesize, because they've offloaded too early to AI. And I think that will only, again, further the gap, because the students who already had those skills will learn how to use AI more productively, learn how to use it at higher levels and do more with it. And then we will have students who find themselves disadvantaged in the long run.

[17:35] SPEAKER_01:

Yeah, just as we saw with calculators. If you never learn your multiplication facts, you never learn how to do things by hand, and you use a calculator always and forever, you don't develop the ability to do the higher level math that those foundational skills are supposed to empower. As you've seen different schools and different types of educational organizations adapt over the past few months to the sudden emergence of AI, Thinking not just about students and student learning and the work of teachers and developing curriculum and planning and how we design assessments for students, what are some of the bigger picture opportunities to think at the school level about how administrators can use AI and how administrators can think about how AI is used in their schools. What are some of the things that we need to be thinking about as leaders who are hopefully going beyond just banning things and saying, don't use this or don't use that?

[18:29]

What do we need to be thinking about as leaders?

[18:31] SPEAKER_00:

Yeah. And I think this is where some of the divine starts already. And I think we've seen districts, most of the folks who we work with, where the leadership is very interested in figuring out how to innovate, how to get their teachers prepped for this. They're implementing professional development on AI at every of their PD days. They're providing teachers access to these tools, paying for subscription plans for some of these tools, starting to have conversations at the school level about what this means for how we grade things, what this means for what kind of assignments we have. Are we going to enforce plagiarism policies about AI?

[18:57]

Do we want to pay for Turnitin's AI Detector or any other company's AI Detector? Do we think that there's value there? Are we recognizing that those are not effective? And those are all fundamental questions that I think will come from the leadership. And I think teachers are waiting for some leadership to make some of these decisions and have some of these conversations. And so I think that starting those conversations is definitely the first thing that most leadership needs to start by doing.

[19:16]

Because I think teachers are waiting for it. There's definitely a survey after survey from Education Week from all various surveying companies has come out and said teachers are seeking PD on this. They're finding that it's not enough what they're getting right now. And so figuring out how you can best support your teachers by providing them the resources they need in order to navigate this next era is important. But this is also a really fun time for a lot of leadership, I think, in the sense that you can rethink what you want your school to be. What kinds of things do you want your students to learn?

[19:42]

How do you want to motivate them? What are the long range goals you have for your students, not just getting past the series of standardized testing. What kind of students do you want to put out into the world in the long run? And what does that mean for how we rethink things if this technology is as prevalent as it is right now and will it continue to get more capable? And so we've seen especially smaller schools where they have the resources and the ability to have these conversations, start thinking about more about, okay, so if we're not thinking about exactly preparing XYZ student for XYZ job, We can take a step back and say, what are the fundamental things we want our students to know about the world? How do we want them to engage with the world, regardless of where the technology is, regardless of what careers exist and don't exist?

[20:17]

Those kinds of conversations, I think, are going to need to happen. And I don't assume that this will lead to a massive change in the 2024-2025 school cycle. These conversations need to start happening now at these levels so that in the next couple of years, when these technologies become even more capable and more prevalent, we're ready to start shifting things towards a more sustainable model of schooling that's not workforce-centered.

[20:36] SPEAKER_01:

Talk to us a little bit about the work that you do with schools, because I know you have two organizations that are supporting different angles on this work in schools. Tell us a little bit about that work that you do.

[20:48] SPEAKER_00:

Yeah, so we have pedagogy.cloud, which is a very technology-centered company where we build tech solutions for schools that are custom-built. So we build custom LMS systems, custom AI tools. We also have some very out-of-the-box things. And so we have teacher tools that you just can kind of go in right now. If they don't want to really learn how to use ChatGPT, you can still make a lesson plan using AI.

[21:07]

You can still make an exit ticket using AI. We have lesson plans for teachers to think about how to integrate AI within their classes. We have a set of 10 English teacher-focused lesson plans coming out of the school year. we put lots of webinars on we have an accredited three graduate course for those who want to explore ai and education so this is a whole range of pd opportunities and tools that we offer through pedagogy.cloud we also have one of the first student bots that kind of allows teachers to monitor the student activity and kind of confine the students conversations with the ai bot to a particular topic which is socrat.ai and we're continuing to do to think about what the next versions of these things can be as technology gets stronger and as appetite for them becomes stronger And then United for Social Change is the nonprofit that I founded.

[21:47]

And there we're doing a lot more of the long-term thinking. And we're thinking about what kind of people do we want to put out into the world in the long run through education systems? And what do they need to learn to do that? And so we have a very heavy interdisciplinary liberal arts focus. We try to bring some of those things down to the K to 12 level. We kind of operate on the idea that the idea of becoming a liberal arts college is amazing.

[22:05]

We think that's exactly what we need in order to have folks who can better navigate societal challenges, to have meaningful relationships with other humans but we currently put a massive barrier to entry on it you have to go to some an elite really expensive or your college before if you really want to access a true liberal arts education and we think that kind of creates fundamental and inequality in a country that's not really sustainable and so we kind of bring some of that to the k-12 level and that's the goal for yet for social change we try to bring interdisciplinary liberal arts education to the k-12 level for civics

[22:36] SPEAKER_01:

Preetan, if people want to follow your work and learn more about what you're up to, where are some of the best places for them to go online?

[22:44] SPEAKER_00:

Yeah, so we like to start everybody off with our Instagram. It's at Pedagogy Cloud. We post lots of things about the book, about courses we offer, webinars we do, but also prompt suggestions, tips and tricks, news updates that are relevant to teachers. All of that gets posted on there. We post almost every day with some sort of content and we find teachers are really engaged with that. And then we recommend folks check out the website, pedagogy.ai, which has links to all the versions of the book, the e-book, the audio book, the book in every store, as well as courses or webinars, and also PD opportunities, both in school and individually for teachers.

[23:16] SPEAKER_01:

So keep us posted on new developments. And thank you so much for sharing just the latest and greatest. So the book is AI and the Future of Education, Teaching in the Age of Artificial Intelligence. Preton, thank you so much for joining me on Principal Center Radio. It's been a pleasure.

[23:32] SPEAKER_00:

Thank you for having me.

[23:34] Announcer:

Thanks for listening to Principal Center Radio. For more great episodes, subscribe on our website at principalcenter.com slash radio.