The Edge

Teaching at the Edge: AI, Accessibility, and Equity in Education

• ISTE • Season 4 • Episode 4

Share episode

The ISTE Edge with Sean Arnold

Georgia Terlaje: [00:00:00] It's time for The Edge, a podcast brought to you by ITSD community leaders. Whether you're a teacher, an administrator, or someone that simply loves education, fasten your seatbelts because this podcast is tailor made for you. Join us as we dive into the world of education, sharing inspiring tales of innovation, passion, and change directly from those on the front lines.

In the episodes to come, we'll highlight the incredible dedication and creativity that drive transformation in the field of learning. Buckle up and brace yourself for an adventure. Coming up today, we're discussing a timely topic, equity and accessibility and AI in education. I'm one of your community leader hosts, Georgia Trilahi.

I'm a former TK 5 instructional coach and an educator of 36 years. And I'm here with my favorite partner in crime, Jessica Peck. 

Jessica Pack: Thank you, Georgia. I always love being here on the edge with you. I'm Jessica Peck, a middle school teacher. I'm a high school teacher and an ISTE author. Today's episode is going to be amazing because [00:01:00] we are going to deep dive into what equity and accessibility look like in the current landscape of education, especially amid the ongoing artificial intelligence boom.

We are fortunate to be joined by our fellow ISTE community leader, Sean Arnold, to discuss the issue. Sean, welcome to the edge. 

Sean Arnold: Hey, thank you both for having me. I enjoyed being here. 

Jessica Pack: Can you introduce yourselves for listeners, please? 

Sean Arnold: Sure. My name is Sean Arnold. I work and have worked in New York City public schools for a little over 20 years now.

For most of that time, I worked in what is known as District 75, which is citywide special education for students with higher support needs disabilities. Uh, now, my current role is the, uh, Director of School Pathways in the Office of Student Pathways at New York City Public Schools, which is honestly a bunch of words that don't mean anything of themselves, but, uh, it, [00:02:00] it, my work is future readiness initiatives.

That's computer science, e sports, uh, financial literacy, artificial intelligence, college career readiness. technical education, all that stuff to help kids be ready for whatever is to come in this new paradigm in which we find ourselves. 

Georgia Terlaje: Well, we're so glad to have, have you here today. And we always like to start out with kind of asking your ITSD origins story, how you became, came to be a CL.

So could you kind of tell us in a, like a Netflix pitch, like how you came to be part of the ITSD community leaders program? 

Sean Arnold: Went to my first ISTE in 2017 in San Antonio. Um, I was blown away by what was there. I was like, I went to a couple sessions though that weren't super amazing. And I was like, well, I could at least be not that good.

And so the next year I tried some, some sessions and none of them got in. [00:03:00] And then eventually they started this like audience choice and I got a really cool audience selection where it was like this gamified learning session that I got to do. And I just tried to find ways to implement meaningful and innovative things and things I've learned at ISTE into my craft as an educator, and ever since I've been trying to guide and lead others in that effort, and they've invited me to participate and have fun with ISTE.

Before there were CLs, there was a thing called learning networks, and there was an inclusive learning network that was about, you know, teaching all sorts of people, especially those with disabilities, and I was a key member of that group. So there is my segue. 

Jessica Pack: Awesome. Maybe that 

Sean Arnold: movie isn't getting made.

Maybe that was too long of a pitch. 

Jessica Pack: No, I think that was just perfect. Um, you're coming from a background of, of inclusive learning. So maybe we could just start our conversation by talking about, [00:04:00] um, how Equity and accessibility are what they look like in sort of the AI era. So do you have thoughts on that?

Sean Arnold: Sure. I I'll say this, this, somebody asked me earlier, you know, where was my foray and entry point into AI? You know, and obviously we all have like our Siri and all the things on our phone that say I. But I remember it was maybe seven, eight years ago. I was at a CS for All event in New York City. Um, and I was presenting on some equitable computer science initiatives we were doing.

It was part of something put together by a group known as New York City Men Teach. And amongst other cool people that were there, uh, Joy Boulamwini. Speaking and joy is and has long been an AI equity spokesperson expert and she gave a wonderful presentation on the equity issues within facial recognition [00:05:00] systems and how it doesn't recognize everybody equally inclusively and some of those systems have since been improved but it shows that these tools are made by people people who are flawed and you know yes Transcribed There's the old adage that like a hammer can be used to build a house or can be used to hit somebody in the head.

But these are very complex hammers and we don't get to see how they work on the insides. And in that way, how do we know what this tool is doing and who it may or may not be harming in all it does unless we Um, I mean, it's a lot of work and it's a lot of time and it's a lot of work to dive in and figure out, you know what's inside of them and understand them better.

And so there and has been my long effort to try to have a better understanding of these systems and how they work and how they affect our students us and everybody else that uses them in the world. 

Georgia Terlaje: So what are some of the most pressing challenges to achieving equitable access to the to AI educational tools?

And how [00:06:00] can educators like begin to address them? 

Sean Arnold: Um, I mean, I would say right now one of the biggest challenges that it is facing is literally there are people who are leading a backlash against educational equity currently, whether it's because of political agendas or fears of social progress, you know, certain terms, whether it be WOCUS or CRT or DEI get weaponized to attack certain initiatives, you know, and and so there are groups that are Typically trying to dismantle programs to bring those forth.

Anything that discusses race or gender or systemic inequities. Efforts to ban books, you know, cut funding for these. Um, so yes, that is a big challenge. And those are misguided in and amongst themselves. But, but even inside of all of that, you know, there's the long standing systemic inequities in these. So, you know, [00:07:00] You go to a company, let's take Google as an example, and look at the workforce of people who are the programmers and designers, and that is not a group of people who exactly looks like the general population, right?

It is a self selective group. And not to say that anybody is purposefully doing anything to bring harm to anyone, but, you know, we are all living our own experiences. We all have our own sort of implicit biases and things that we prefer and things that we don't. And if the room that is creating it is not reflective of the room outside of it, you're going to miss things.

In a case in point, um, brought up with Google years ago, issues that they had with some of their accessibility features. The fact that, you know, Chromebooks didn't allow for switch access for students who had physical disabilities, and That's since something that's been corrected and they've gotten better about all of those things.

And I've gotten to be on panels at Google to help fix those efforts. But the fact that we had students [00:08:00] who needed other devices who couldn't move above their neck and they needed to still interact and interface to learn, um, say a student with cerebral palsy, who cognitively is at a very high level, but physically would be challenged to use a keyboard or computer, you know, If there aren't people in the room representing those folks, who is going to be building those systems for them?

And even more so, all these other deep, uh, machine learning systems where we don't see exactly all the things that are physically being done. And it isn't just about You know, physical access. It's about social and community access and all the other ways that there are barriers to, uh, us getting the materials and the learning and information we need.

Jessica Pack: This kind of puts me in mind of another sort of group that needs to be considered, and that would be Those learners who have English as a second language and have a really diverse, you know, cultural or linguistic background. [00:09:00] Um, how can we ensure that the development of AI in education kind of considers that and avoids those unintentional biases?

Sean Arnold: Well, so we say AI in education. Obviously, there's the part of that that's the systems and the designers who are creating the applications and everything that, you know, have to be reflective and build those. But then there's the other part of that are the people in education. So that's the thing I slightly have more control over.

So I'll speak more to that part of that. It's about training our users, be that teachers and students, you know, robust professional learning experiences. Thank you. Get them to recognize when there are bias or flaws or just incorrect information that is the output from these systems and, and learning behaviors, uh, that are best practices where it's not, okay, I throw in my question, uh, pops the answer.

And that's what I put on a page because we learn, Oh, you know what? Yes, it saves [00:10:00] me time. It saves me effort. It saves me energy, but I still have to go in and check to a degree to make sure all this information is appropriate and accurate beyond that. I mean, these systems can conceivably be helpful for those experiences too.

So, you've got a class of 30, 40 kids even, hopefully not, but then you might have a subset of those students, maybe even 10 percent of them, who are multilingual learners. I don't speak. Bengali. I don't speak Portuguese, perhaps. Maybe I don't speak or do. How do I make this accessible to them? Well, with digital systems, I can get it to them in their native language.

I can get it to them with visual imagery and picture symbols that can create understanding for them. It can translate and speak to them out loud. It can also differentiate. Cultural, uh, stories so that it's aligned and personalized and individualized for students. So, so in a case scenario, these AI [00:11:00] systems are actually helping to personalize the experiences for each of those students in the way that, you know, individual teachers don't have the time or energy to do so on their own.

That's a good case. 

Georgia Terlaje: I realize this may be a hot button issue, but maybe you could give some advice. You know, what could educators do to kind of voice their concerns with policymakers, especially in this landscape, to be the voice of what's going on in the classroom and You know, what's, what's all students need?

Not just what, you know, this students that are, I don't know what the right wording is, but that they tend to just slap, you know, everyone looks the same and this is what everyone needs to do. And we know that doesn't make any sense. So do you have some advice for teachers, like how they can be heard or what they can do?

Sean Arnold: So, you know, everybody's role and situation is obviously different. You know, somebody who is that first year teacher in [00:12:00] Pocatello, Idaho, doesn't have the same experience as somebody who's, you know, gotten the new role is like digital learning innovation specialist, you know, in their earns, right, because they're in a position of higher.

So obviously, you know, there's no blanket advice that can be given to anyone, but but it literally starts with recognition of the issue, right? Recognition and acknowledging those sort of systemic inequities that exist because not everybody is even there. So literally seeing it, noticing it, then giving voice to it is, is like step one through five.

Um, then, you know, getting data, because a lot of times people in authority, if it's simply hearsay, if it's simply feelings and vibes, you know, not that anecdotal, information is an important and meaningful and it helps to tell a story and you should tell that story. But, you know, use the data that you have available to expose any of those and see, you know, there might be [00:13:00] like superficial things that are, I don't mean superficial in the sense that like they're not deep and embedded and important, but you see the surface.

context of how it's occurred, figure out what the root causes of why, you know, all this subgroup of students is coming to school and not achieving as well because they're exhausted. Oh wait, they have to take a one hour long bus ride from a different community that we didn't recognize and we're not accounting for that, you know, and so the issue isn't like, hey, these kids aren't paying attention in class, the kids are exhausted and we need to deal with this transportation issue, right?

Um, Another big piece of that is once we've gotten past the basic issues we can get into like deeper learning experiences like culturally responsive teaching that speaks to the communities of our students inviting in family inviting in the community itself to participate in that learning experience to share their history and their culture.

And, you know, in the best case scenarios, maybe share some food and some cuisine so we can all snack, you know, you know, [00:14:00] and give us all those multiple perspectives, you know, it can start with a simple like book club, right? Student led, family led, you know, and that's the other piece of it to like invite students in to be a part of these discussions on how it works, elevate their voices and let them share what it is that they need from that environment to be conducive to their learning experiences.

Yeah. Build a whole community of folks. But it starts with literally just seeing it, recognizing it, and calling it out. 

Jessica Pack: So we used to kind of frame the conversation of, um, equitable access kind of just around equitable access just to tech or just to Wi Fi. And we made really big strides in the pandemic, um, in those particular areas.

But now we have this sort of emerging conversation about Providing equitable access to AI powered learning resources. So for you, what, what does that really look like? What would you recommend? [00:15:00] 

Sean Arnold: Oh, you know, in one way of a digital divide. has been a real thing. Um, you know, we had a whole system because you spoke of the pandemic of how do we get devices into the homes in a population of a million students here in New York City.

And, you know, at one point it was literally buying, oh, I think the initial purchase was like 350, 000, uh, Not just Wi Fi enabled, but like LTE enabled iPads that were delivered out to homes and like we help parents and stuff set it up and, and we're there in shelters if that's where students are and making sure that they have access wherever they are, right?

Yes, that is an initial step to make sure people have this access and there's lots of programs available out there to help get free Wi Fi in communities and stuff. But beyond that, it is training the teachers to recognize that broad array of needs specific to AI. [00:16:00] We're teaching a lot of tools in a lot of cases, but I don't know them in many spaces.

We're teaching enough AI literacy, right? And the general understanding of what these systems are, what they do, how there can be Bias or misinformation or misinformation built into these systems, but also beyond like what it is these systems are good at and what they are not good at right and there are certain types of question the material that they can do and that they can create really well.

with Fidelity, and there's certain things that they are not meant to do. Uh, well, even just in a, in a personal context, you know, you're working with students with disabilities, and there's like, IEP generator AI, right? And, well, to get help writing some ideas, but it's called an IEP, I hate acronyms, I use them all the time, but like, Individualized Educational Profile, right?

individualized and how does this machine know that [00:17:00] individual student? Are you feeding all that student's personal information and data? Well, then we have a whole different issue because that's a data privacy issue. So, you know, again, these systems. Can be utilized meaningfully and well and equitably.

But again, that requires the training of the teachers to know what things they need to look out for. And then we get to the students and, and start to show them broadly the bias in the systems. Like in New York City, we have lessons for computer science and, and AI training built around these. It's all part of like a broader digital citizenship curriculum that, you know.

Obviously is level depending on the age of the kid, you know, for younger kids, it's, it's about, hey, this is what, you know, dealing work might look like for older students. It's understanding with depth, you know, here's how good prompting occurs. And here's certain structures we want to use. But here's different case scenarios of ways in which that output has been harmful to people and ways in [00:18:00] which it's been helpful for people, right?

And yeah, an overall understanding and training and you gotta have, again, depending on where you are and what your school is, broad systems to make that happen. But that's where it began. 

Georgia Terlaje: So we've been having several guests on recently talking about AI and AI training and we're kind of learning, Jessica and I, what different states, what different school districts are doing around training educators.

And I'm just curious, what do you see going on in your district or your state as far as that training piece? for teachers and educators with all the great things you're talking about. 

Sean Arnold: I mean, they're, my, my district is huge. So there's like no one thing is going to work for everybody. But we, we started, we had, well, honestly, we started where a generative AI was banned in New York city.

So that's where we started. Um, and then the recognition that banning something isn't going to help people. Learn how to navigate it better. [00:19:00] So there's a whole plan and a structure in place. Now we had we had an AI policy lab and we've helped create training for teachers. For example, there's a starting one that's been a basic AI literacy training.

It's asynchronous teachers all across the city in their own time, whenever they have available can go ahead and engage with it using our platform. We learn, um, yeah. But then on top of that, just last week, you know, we have individualized training. So there, there was a really cool one, uh, that was an AI literacy and tech justice that we did in partnership with the Kapoor Center.

Um, that's deep diving into equity within AI environments and trust in spaces and not trusting in certain other spaces, uh, partnerships with like Common Sense Education, Common Sense Media, and their AI curriculum, or, or even just, uh, hands on with students during Day of AI or CS Ed [00:20:00] Week and their Deep diving into an exploration of computer science and artificial intelligence using Minecraft, you know, so, so there's a wide array, you know, we, we built AI equity lessons on our CS for all equity site, um, that teachers can use and borrow and we share.

So, so there's, there's a range of levels and recognizing that not everybody is going to be starting and engaging in the same place and, and you need to speak to all of that, but it all starts again with just a basic understanding of AI literacy. 

Georgia Terlaje: But I appreciate you sharing that because I think, you know, if listeners, we know it's these international and you know, whether it's administrators or educators around the country, if nothing's happening in their area, it may give them a place to like, start or, or, or no, like, how could I like, start this locally to help get people trained?

So I appreciate you, um, sharing what you guys are doing. 

Sean Arnold: Who is it? Google has a decent AI training for teachers. One of the people who helped put that together is a former [00:21:00] New York City schools teacher. Um, uh, Code. org has a decent AI teacher training and, and as I said, Common Sense had, has a really expansive one, mostly for older students, middle and high school.

Uh, but they do have in a few individual lessons connected to it for the younger kids too.

And of course, ISTE has its whole array of training. I mean, I should speak to other things that we're doing because we just had our New York City Schools Tech Summit a few weeks ago, which is like our big, imagine like a New York City solo. It's the conference right just in New York City with our teachers and stuff and we were at Brooklyn Tech High School, which is the biggest in terms of population and literally space high school in the whole country, at least country meaning the world.

Um, it's, it's huge, but Big event, Richard Coulada, who's like the president of this, he was one of our keynote speakers, uh, [00:22:00] discussing, you know, AI in our classrooms, uh, he just gave out a lot of copies of his book, uh, there were sessions all over the conference, on a, on a wide range of tech topics, but obviously AI played a huge role in a lot of these sessions, yes, some from vendors, but some, uh, from teachers and their own individual practices that were being shared and connected, so, you know, Even physical in person ways to learn about, uh, all the work that we're doing, which is cool.

Jessica Pack: So I feel like, um, one of the concerns that we sort of hear boots on the ground a lot from other educators is this sort of nervousness that AI has the potential to really kind of like replace the work that we're doing. Um, could you maybe speak to Sort of critical human interactions and how we can balance, I guess, AI to support student learning and make sure that we're still preserving those human interaction pieces.

Sean Arnold: Trying to remember [00:23:00] where it was. I want to say Chicago, but I could be wrong. Um, where they did try to totally implement a full AI teacher in these virtual classrooms. Um, and it, and it didn't go well or for very long, all on its own, you know, it sort of crashed and buried because kids were putting in random and.

Information that spit back out stuff that, that again, was not conducive to a learning environment. Um, there was no guidance from this bot. I'll also say that I think I read an article just a couple of days ago about how like Microsoft did a study where in people who are using AI, at least in this specific context of their study, very, very regularly at work were becoming less good with critical thinking tasks themselves as human beings.

As if, like, dumbing them down. So, I think some of that reticence is, is justified, and I think it always should [00:24:00] be. Only because we should be critical, always, of all things. Not to say that it can't be amazing, and there can't be wonderful use cases of this. In the same way that we are all, like, super excited when social media first came out as this new technology, right?

And it's gonna connect us all, and the whole world's gonna be together, and we're gonna hold hands and sing Kumbaya. Yeah. And There have been moments, and not to say that over the years I haven't enjoyed the poking of people and whatever it was that we did back on Friendster. Um, but it's also changed and morphed in in ways because of who's in charge and who's dictating, you know, as a power so like we've got, you know, tech companies, you know, are benevolent big brothers ish, but that are also buying up.

Large swaths of energy companies because AI itself uses so much power generation and, you know, whereas in, you know, new AI that's being generated in countries other than the [00:25:00] United States seem to use a lot less energy and power. So, you know, that's the whole environmental conversation around it too. And considerations that we have to have, um, which, which is an equity piece because some people live in flood zones, right?

All that said, we have to make sure that our lens remains critical, but we also remain open to possibilities for good and change in the spaces and places that we can control. So like, I don't have the power to decide, you know, what the US military or massive corporations are doing with their dollars and funds.

I can decide, you know, what is happening in an individual classroom and how I'm preparing individual students. for the future and giving them the tools that they need to succeed and showing them how to use it properly but also to judge ways in which it might be used improperly and how to prepare themselves for situations and environments they may encounter where that's happening.

You know, so critical lens but open [00:26:00] mind. 

Georgia Terlaje: That, that's a great, um, catchphrase. We should have that on a t shirt. Seriously, because that needs to happen. I'm um, you know, just thinking like this whole subject. It's crazy how it's such a wide range of Experience and knowledge just as educators, you know first year teachers that are just figuring stuff out But also, you know seasoned educators that maybe are just dipping their toe in so maybe for some of those people that are just kind of getting started with AI.

What's some advice you would give them around keeping their minds open around accessibility and, you know, being an inclusive environment, um, that won't, won't make them afraid to try, but just give them a lens to kind of think through. 

Sean Arnold: I think again, like we said, step one is simply acknowledging the importance that all students [00:27:00] deserve the chance to learn in their best capacity is a wonderful beginning.

Um, because again, that's literally not where everybody in our world is recognizing that each student deserves the opportunity to succeed. And then we can take those steps, you know, get to know our students well, obviously, as well as you can in every situation, depending on how many students you have, you know, if you're seeing 400 kids in a week, that harder, you know, but the better you can know your students and I don't just mean academically, but like personally, socially, personally, like the, the connection that you can build, then you can know which is best suited for them in every situation.

And then you can also recognize what sort of systemic or, or cultural or, or challenges might be facing them in their journey. You know, Yes, some of them might be listed [00:28:00] on an individualized educational profile, but some of them, you know, literally might be written in their spirits and their skin and their bodies, you know, in the long history of life that they've experienced.

Right. And know that in all of those kids, there is no single story, you know? There is no single, sorry, for this subgroup of human beings or people, and that each is an individual with their own individual strengths and weaknesses and hopes and fears and dreams. Um, and then from there, you know, once you know them, you, you, you can sort of try to connect with their family, connect with the community, build up a respectful communication.

In this group, and in that way, you'll at least know more information on where to begin and how to help, you know, again, because the specifics of helping any individual group, whether it be multilingual learners, whether it be a student with a physical disability, whether it is a student who is coming [00:29:00] in who's from another country who hasn't had any formalized education at all whatsoever, to this point.

Those needs might be individualized, you know, and the steps you take might be individualized. But, you know, the step of knowing them and connecting with them and understanding that there are individual needs you should meet, you know, is a great place to begin. 

Jessica Pack: I really like what you're talking about with like understanding each individual's story and journey and, and letting that shine a light on their needs.

It's very design thinking. That's kind of what that's putting me in mind of. So thank you very much. Um, this has been Yes, absolutely. A universally designed world for them. Well, this has been such a great conversation, Sean. Thank you again for being here today and sharing your expertise. We really appreciate your time and your message.

Before we let you go, where could listeners connect with you? 

Sean Arnold: Oh, you know, all the social medias are so weird anymore, but like Sean [00:30:00] M. Arnold on most of them and, uh, brave in the attempt. com, uh, is where I share my stories and a blog and whatnot. So yeah, those are probably the two best places. 

Jessica Pack: Wonderful.

Well, that wraps up this episode of the edge podcast. We hope you had a great time. My name is Jessica and you can find me at pack woman two zero eight on X threads and Instagram. Georgia, you're muted. Matt, Matt, 

Georgia Terlaje: Matt. I just realized that. And I'm Georgia Terlahi, and you can find me at Georgia Terlahi on Blue Sky, and both of us at storytellingsavestheworld.

com. 

Jessica Pack: On behalf of everyone at ISTE's The Edge podcast, remember to keep exploring your passion, fostering your creativity, and continue taking risks, all things that can bring you to the edge. Thank you, Sean. So the, 

Sean Arnold: so the 208, is that like your class number room? [00:31:00] Is that the classroom? Yeah, 

Jessica Pack: for, for 20 years, it was my classroom.

And I just moved about a month ago. And my kids have been like pestering slash heckling me. You're in literally the 

Sean Arnold: same room for like 20 years. 

Jessica Pack: It's mind blowing, totally mind blowing. I was the only person in the school who had the same space. But now my kids are like very invested in what number will you choose now?

And I'm like, I'm more intelligent than I was. That many years ago. I'm not going to pick a number. 

Sean Arnold: That's so cool. I was hearing it and I was like, maybe that number has a more significant meaning. I don't think it was the 208th person with that. Yeah, okay. 

Jessica Pack: No, it's like back in AOL days where you'd put like your birth year after it and you'd be the only one because it was so new.

Sean, can you do 

Georgia Terlaje: me a favor? Will you, um, put the link to, um, what the, your [00:32:00] site in the chat for me? Cause I'll link it to show notes for Margaret. 

Sean Arnold: I mean, do you want me to link any other stuff 

Georgia Terlaje: that I've talked about? Yeah. Anything that you would like to share with people? Sure. Sure. Here is. I'm gonna stop the recording