Welcome to Episode 122 of the Think UDL podcast: Artificial Intelligence and Accessibility with Joe Houghton. Joe Houghton is an Assistant Professor at University College Dublin Smurfit Graduate School of Business in Dublin, Ireland. He is also a prolific writer and producer of books, podcasts, and instructive videos and webinars on teaching and learning. He has recently thrown his intellect into the emerging trends in Artificial Intelligence and published his book, Applying Artificial Intelligence to Close the Accessibility Gap. I was glad to have the opportunity to write the foreword for this book and am delighted to have the chance to speak with Joe for the second time on the Think UDL podcast. If you’d like to hear our first conversation, you can listen to episode 91: Reflecting on a Starfish Difference with Joe Houghton and I will have a link to that episode in our resources section on the ThinkUDL.org website. This episode, episode 122, details some of the ideas from Joe’s book on AI and accessibility. We discuss tools that can be used to improve accessibility, the benefits and challenges of using AI to make courses and teaching more accessible, and what ethical concerns we should be aware of when using AI. And finally I ask Joe a bit about what he sees on the horizon for AI, UDL, and accessibility.
Resources
You can register for the UDL seminar series in 2024 from Goodwin University with Joe Houghton and Elizabeth Hitches in July – register for free at UDL for Free Webinar
Joe’s book on Applying AI to close the Accessibility Gap on Amazon – Use this link for the UK and This link for Joe’s book in the US.
Joe mentions Ludia – Free to use AI chatbot to help with all things UDL
Otter.ai – transcribes meetings, videos etc and gives smart summaries – https://otter.ai/referrals/UBCS2N25
Use This Link to sign up for Joe’s AI newsletter -you will find links to Ethan Mollick, Amanda Bickerstaff, Mushtaq Bilal and Beth Stark in the episodes.
Transcript
SUMMARY KEYWORDS
udl, ai, gpt, students, accessibility, people, learning, book, transcript, learner, work, tools, podcast, educators, dyslexia, write, teaching, talking, years, chat
SPEAKERS
Lillian Nave, Joe Houghton
Lillian Nave 00:02
Welcome to think UDL, the universal design for learning podcast where we hear from the people who are designing and implementing strategies with learner variability in mind. I’m your host, Lillian nave. And I’m interested in not just what you’re teaching, learning, guiding and facilitating, but how you design and implement it and why it even matters. Welcome to Episode 122 of the think UDL podcast, artificial intelligence and accessibility with Joe Houghton. Joe Houghton is an assistant professor at University College, Dublin, Smurfit Graduate School of Business in Dublin, Ireland. He is also a prolific writer and producer of books, podcasts, and instructive videos and webinars on teaching and learning. He has recently thrown his intellect into the emerging trends in artificial intelligence, and published his book called applying artificial intelligence to close the accessibility gap. I was glad to have the opportunity to write the foreword for this book, and am delighted to have the chance to speak with Joe for the second time on the think UDL podcast. If you’d like to hear our first conversation, you can listen to episode 91 reflecting on a starfish difference with Joe Houghton. And I will have a link to that episode in our resources section on the think udl.org website. This episode, Episode 122 details some of the ideas from Joe’s book on AI and accessibility. We discuss tools that can be used to improve accessibility, the benefits and challenges of using AI to make courses and teaching more accessible, and what ethical concerns we should be aware of when using AI. And finally I asked Joe a bit about what he sees on the horizon for AI UDL and accessibility. Thank you for listening to this conversation on the think UDL podcast. Thank you to our sponsor Texthelp, a global technology company helping people all over the world to understand and to be understood, it has led the way in creating innovative technology for the workplace and education sectors, including K 12, right through to higher education for the last three decades. Discover their impact at text dot help forward slash learn more, that’s learn m o r e. And I’d like to say welcome, Joe, how move back to the podcast. I’m so glad to have you today. Invitation.
Joe Houghton 03:10
Well, since
Lillian Nave 03:11
we last talked, you have written a book about applying artificial intelligence to close the accessibility gap. And that’s what I wanted to talk to you about today. But before we get there, I’m gonna ask you the same question I asked the last time. And that is what makes you a different kind of learner. Okay,
Joe Houghton 03:34
yeah. And I mean, you when you emailed you said, Do you know, do I ask you a different question as normal? And I came back came back to you and said, well, let’s read the same question. Because, because I think my answer was changed. I mean, it was what I think it was July 22, when we last talked didn’t wait for episode 91. And I mean, in in the meantime, artificial intelligence has kind of hit the headlines with the release of chat GPT in November of 22. And ever since. And I’ve been, I suppose, as a techie nerd at heart. Yeah, you know, I was I started off kind of my working life as a computer programmer and stuff like that. So the whole AI thing kind of really got me interested, as soon as chat GPT arrived. And I plugged myself in to AI. And I started a newsletter, and I’m running AI training courses and, you know, both educational ones and corporate ones and stuff like that. And so, I think this this thing of learning. I think it’s going to change learning for a lot of people and it certainly changed learning for me. Because I regard myself as a perpetual student, you know, I’m always whatever I do I’m doing very often to learn. And that’s, we’ll come to that with the book. But now we’ve got a lot of new tools, yes, that we never had before. And they’re very powerful, and they’re getting more powerful every month. So am I a different kind of learner? I’m a learner that’s constantly aware that there’s new ways of gathering and processing and collating information, and sharing that with other people. So maybe I’m different to some people in that I love to do that, you know, that whole process is something that lights me up. And I think in the last 18 months, I’ve become aware of a lot of different ways to do that in in ways that you couldn’t do before. Yeah,
Lillian Nave 05:51
yeah. So you’ve added kind of all this extra technology that you can input in your learning process, it makes me think of the kind of sci fi future worlds where we’re all cyborgs of some sort sort of seems like we’re on the way. Like, we’re not It’s not inside our bodies or anything so much yeah, of our ability to communicate, and to gather information and to analyze and think is connected to some sort of technology. Yeah,
Joe Houghton 06:25
I mean, there’s a, there’s a lovely quote from Arthur C. Clarke, who is a great science fiction author that I grew up with. And, you know, he said that any sufficiently advanced technology is indistinguishable from magic. And I think when you see chat, GPT work for the first time. And somebody asked it to produce something and it goes, this incredible thing just appears on the page. And, and the jaw drops, you know, of the person who’s watching it. And, and we’re going through a period where, where that is happening for a lot of people. And it’s, it’s an exciting time.
Lillian Nave 07:03
And I see it in my students as well. Right? And what used to be an assignment that went one way, I can now even show them like, well, you could do it this way, or that way, or look at all these different ways. And you’re right, their jaws drop. Like, I had no clue. I
Joe Houghton 07:25
have no clue. And we’re in it’s interesting, because I mean, we’re in what 18 months in. So it’s still quite new. And there’s still a fear. And there’s still a resistance from some people and some institutions and teachers as well. A lot of my students, even up to the end of this semester, really hadn’t got to grips with AI, some had, some are fully embraced it some you know, it’s just part of their workflow, I think we’ve got a limited window when that is the case, because it’s just going to become part of day to day, it’s going to be built into Microsoft Office, it’s going to be built into Google, we’re not even going to think about it’ll just be the one of those things we use. But it’s not quite there yet. And I think as educators, we’re almost it’s almost incumbent on us to open up this box to our learners, because some of them are very fearful. And some of them once they see it, love it, and some of them are still a bit fearful. And even some of the educators are fearful as well, because maybe they’re feeling well, you know, I’m a, you know, tricks and all this kind of stuff. So it’s an interesting place to be as an educator, you know, for students of whatever capability level you’re at, and from a UDL perspective as well. I mean, that’s, that’s absolutely fascinating, isn’t it?
Lillian Nave 08:52
Of course, of course. And I had pulled my students this past semester. And I still did have a good bit of students who were not really interested in it, you know, they’re just not ready to jump on that train. But I had heard of this a couple years ago. And I think it applies to AI, that the new literacy or illiteracy, it’s not going to be about can you read or not? It’s about how facile you are with technology. And I think about when I was helping my elderly parents, and they didn’t know how to use a smartphone and how does click on Zoom during the pandemic, and realize that that is the new thing that’s going to hold people back. And so we need to be teaching our students how to incorporate it or else they will be left behind
Joe Houghton 09:57
and we also need to be teaching our instincts Do you want to facilitate everybody in this? Because there is a digital divide, there is a feminists still. And, you know, just because you can afford to pay for, you know, the fancy version of chat GPT shouldn’t mean that you’re not left behind. Just because you haven’t got a computer but you’ve only got a smartphone shouldn’t mean that you’re left behind. So there’s there’s a number of layers here as well, aren’t they from the educational, the accessibility, the multiple means of representation, kind of points of view, and multiple means of access points of view. All of which you’re kind of lighting. Okay, no, I
Lillian Nave 10:41
know, my eyes get really big when you’re like, oh, yeah, here we go. We’re getting into our UDL talk. So okay, I feel like we’ve been talking about why this is so important. And that leads into my next question, which is, what’s prompted you to write this book at this time? And why was it so important? Throughout the book that you included so many? first hand accounts, like, there were so many teenagers day with ADHD or with dyspraxia? Or, you know, I really got a sense of how many different learners there are. And you really use that to verbally illustrate your book. Yeah,
Joe Houghton 11:28
okay, well, well, the book, in the last probably five or six years, I’ve found that I’ve shifted as a learner. And now that the way that I engage with if you like, a new subject, is that I start gathering information. And particularly in the last 18 months, the information gathering process has become exponentially easier because of the power of the tools. So what I find now is I’m gathering information, and I naturally want to start organizing it, because I like lists. And I like, you know, I like organization, and I’m finding that writing is a good way for me to do that. So, you know, in the last, what, three years, I’ve written six books,
Lillian Nave 12:13
wow, it’s very prolific.
Joe Houghton 12:15
This came about I mean, to a large extent, it’s your fault, because because, you know, when when Jen Lynch introduced me to UDL, two or three years ago, and then I came across, think UDL, and you know, we ran into each other and started to correspond and talk and all the rest of it. And, you know, UDL is, has become a thing for me. But the accessibility bit hadn’t really become such a thing. But my little lad, who is 12, has hypermobility in his skull. And we had him assessed a couple years ago. And they said, he’s in, like the bottom 1% for writing. And this is where he was having a lot of problems at school with his written work. He’s very, very intelligent. But he couldn’t write and, you know, form letters very well, and all that kind of stuff. So now, we went through that process. And you know, the school said, right, well, we’ll give them a computer. And so He now uses a laptop, and he’s fine. And I mean, he’s using Word and PowerPoint, and Canva, and all these things. And he’s 12. Brilliant, loves it. And that kind of opened my eyes to the whole accessibility thing. And isn’t it strange? I mean, I’ve been, I’ve been a university educator for 20 years, and I’ve known about accessibility. And I’ve had, I must have had 1000s of students come through all my different classes, which means I’ve had what 10 20% of those have got some kind of accessibility, you know, whether you call it an issue or whatever, they’ve, you know, they’ve got dyspraxia, they’ve got dyslexia, they’ve got hidden or non hidden, what the what they like to call disabilities. But I’ve never really engaged with that. I’ve always said at the beginning of class, you know, if there’s anything I need to do to make things better for anybody, let me know. But that’s again, that’s after the fact rather than building in this stuff ahead of time, which is what UDL and accessibility should be doing. So I suppose this was a wake up call for me, you know that accessibility was a thing. And I’d seen it, how it affected Danny. So I needed to do something about it. I needed to do something more for me about it. And I needed to educate myself about it. So I wrote a book, because I didn’t know anything about accessibility. And I mean, I’m a total fraud. Because, you know, I’ve got a book out on accessibility. And people look at me and think like, oh, Joe knows about accessibility, you know, and I don’t I just wrote a book about it, because I didn’t know about it.
Lillian Nave 14:49
And I love that we see the whole learning process and you’ve organized it so well. That really if anybody wanted to find out something they could look In your index, which is quite voluminous as well, and say, Okay, I want to see what I can do for a student that I know that I have that has dyslexia or ADHD or something like that. And it really just gives a full understanding of so many different ways that accessibility, or the lack thereof, affects learning. Yeah.
Joe Houghton 15:26
And I think, you know, coming, putting a book together, I mean, I started book and I generally start with a table of contents. And I kind of thought, right, well, there’s three kinds of things I want to touch on. There’s UDL, there’s accessibility, there’s ai, ai actually became a bit of an afterthought, even though that was probably the thing that I was most interested in when I started if you like, but I knew UDL and accessibility kind of conflate a lot of times, but they’re not the same there. They’re linked separate. So when I started, I mean, whenever I put any piece of work together, whether it’s a lesson plan, whether it’s a presentation, whether it’s an article, I stopped before I write, and I start to think I say, Who am I writing this for? Who’s my audience? And I try and kind of create a persona. Okay, so my persona for this book is a teacher. Or, and it doesn’t matter whether that’s, you know, what level of teaching, you’re at a teacher who has got people with accessibility challenges in their learning cohort. And that might be a formal classroom, that might be an online cohort, I don’t know. Who doesn’t know anything about accessibility? Because that’s where I was. And kind of, you know, still I’m to a large extent. What if, if, if I was going to give them a book that would help them through what they needed to know, or at least gives them a starting point of what they needed to know. And some jumping off points to find out more? How would I structure that book? And what might I put into that book to, you know, write, I’ve got a student with dyslexia in my class, I’ve just got an email from the disability office saying, this student is registered as a, you know, as a student with with special needs, or whatever it is. Oh, I don’t know anything about dyslexia? I don’t, what do I do? Yeah. So that that’s where I started from, if you like, and the and the kind of the section in each of the different conditions, if you if we call them that, about, you know, a teenager’s day with was just me trying to think of right, well, what’s, what’s another way that I could represent some of the information in a way that would resonate more than just dry facts and figures or, you know, whatever. And that’s again, I mean, you, you were kind enough in the foreword to say that this is a strengths based book. And I love that you pick that up, kind of because I was hoping that that would come out. Because another thing that I started doing, and when I started the book, I wrote, from the point of view that these things were a problem, that if you had dyspraxia or you had dyslexia, or you had ADHD, or whatever, you know, how do we deal with the problem? And then I thought, well, hold on, you know, Danny’s got this thing, but it’s not a problem. Yeah, yeah. I mean, now we fix the challenge. It’s turned into a massive strength for Danny, and he’s gained all these extra benefits. And he’s, you know, he’s doing computing and he’s typing fast. And he’s using all this new stuff, which is going to stand him in huge stead later on in his career and his education and stuff. And that’s when I kind of decided to start each chapter with positive aspects of, and then challenges in learning from, you know, just Lexia does calculus, you know, whatever, whatever it is, to kind of turn that on its head and make because I think there’s power in words, and there’s power in the order of words. And if you start with positive and then go to negative, that’s a very different equation to starting with negative.
Lillian Nave 19:38
Yeah, yeah. When you mentioned that you were thinking of a teacher, an instructor who would get a letter from their university, let’s say office of disability, and that is a common occurrence. That is starting with the negative, right, the first instance or introduction that instructors having with the student, is that the student has a deficit, or the student is different. And how can you make a change or do something that will bring that student more in line with the average, or the normal, and I’m using air quotes with my fingers. And it doesn’t start with, you have this amazing student who has these incredible skills, and can add value to your discussions, your class, this is really great for you. And then also has these other things that are happening in in their life, just like everybody in the world has all of these things happening in their life. And it really does change the script. When we move from the deficit based language and the deficit based thinking about disabilities, then we see such a more positive and more nuanced understanding of, oh, everybody’s learning, I think, in that situation, I
Joe Houghton 21:18
mean, that it also then moves you towards, rather than making accommodations, building stuff in from the start. Right. Which, which, again, is classic UDL, you know, kind of don’t don’t retrofit stuff, to fit somebody with a with a problem or a challenge, build it right into your process, so that the challenge never comes up. And I’ve started to do that as a result of my engagement with UDL. And now my engagement with accessibility, which is adding a newer and newer, kind of more nuanced layer, like you say, to my understanding of, if you like my teaching practice, which I love, you know, I love the fact that I’m still developing and learning and trying to figure stuff better. I think when you stopped doing that, you probably should hang your boots. Right? I
Lillian Nave 22:13
noticed too, that when you start with those positives, you’re really showing your values. And I believe this about universal design for learning as well is that you value the differences in your students, you value learner variability. If you valued conformity to the typical or normal, then you would want to change that variable person and make them more like the normal, like, we’re going to do it this way. And I’m going to give you this help this tool. So you can act more like the other folks in the class. And so in that way, you value normality, or you value one way of doing things or the traditional way of thing, doing things and not valuing from the very beginning, the learners themselves in their variability that
Joe Houghton 23:20
the possibilities that those bring, I mean, I you know, I mean, I’m looking at the, the kind of table of contents here, you know, I mean epilepsy in so we start with positive aspects of epilepsy. And I mean, you might think, well, epilepsy has nothing to do with accessibility. good it is, because if you’ve got access, if you’ve got epilepsy, then there are there are opportunities. And there are challenges, just like there are with ADHD, just like there are with ASD, and with all these other different things. So some of them health impairments, some of them are kind of mental approach differences, you know, whether whatever spectrum that is that you’re on, or your student is on or whatever, but starting to research and starting to ask questions of, you know, chat GPT and other bots and things about right, well, what what could the positives be for learner who’s got this condition or whatever? I mean, this was an eye opening experience for me. And if it was, and the way I looked at it every time I went, wow, it went in the book. Because if I went well, the chances are that other people with my level of ignorance, if you like, we’d also go well. So that was a kind of like, yeah, that’s a trigger that goes in the book.
Lillian Nave 24:35
Yeah. And I’m just like you because when I was starting out in teaching, I really didn’t know anything about accessibility. And I would really only know anything, if I got a letter from our Office of Disability from any of the institutions where I worked. And it wasn’t until my introduction to Universal Design for Learning and that I I started to understand the positives and learner variability. And it’s, as you know, completely changed the way that I design learning and talk about it. And everything.
Joe Houghton 25:15
Absolutely, it was, I mean, lovely little story from from last term, I got a super email from a student towards the end of term. And he said, I didn’t tell you at the beginning of term, he said, but, but I’ve got, I’ve got not dyspraxia, the other one, dyslexia, dyslexia. And I just wanted to say that this was one of the easiest classes that I’ve ever gone through. Yeah, because you made all the information available on day one, it was all laid out well, on the learning management system, you gave us PDFs, and word documents, which meant that I could actually engage with the material in different ways. And lots of videos, he said, which was really helpful. And how I was reading this kind of with a big grid. Yes, something’s working.
Lillian Nave 26:16
Yeah, wonderful. Yeah, I had that experience, too, with a student with ADHD. And because I scrapped all of my timed quizzes and tests. Yeah. And they didn’t even have to ask for their accommodations for, you know, their time and a half, or some of the other things that they usually get.
Joe Houghton 26:38
Yeah. So so it does make a difference. And sometimes, you know, when I’m teaching other educators now about UDL, one of the things that sometimes comes back is our this is going to take a long time, you know, to do all this extra stuff and turn my all my documents into PDFs, or, or whatever, it’s going to take me loads of extra time. And, you know, I haven’t got the time, can’t be doing all this stuff. You know, and I’m saying, Well, you can do it once, and it takes you 30 seconds to save as a PDF instead of Save As docx or, or whatever. So one, it’s not a lot of extra effort to it makes the material so much more accessible to different students, whether you know that, oh, no, it makes it more accessible. And, you know, then I tell stories, like like the one I just told you, and hopefully some of that goes in. And you know, that there are now tools available. I mean, yeah,
Lillian Nave 27:36
I mean, I wanted you to go into what tools that you think improve accessibility that you think others should know about, and should have a facility with, that’s going to help instructors like me, design their courses. And I know, you go into so many in the book. So I asked you to just say three, but you can tell me the ones you think are important no matter what the number
Joe Houghton 28:01
one, which I think I mentioned in the book, but if I don’t, I should have done, but which my university bought and rolled out across the whole university about two years ago, and I was part of the pilot for it is is a product called ally, a double L Y. You’ve probably heard of it. And what ally does is it’s a plugin that works with all the learning management systems. So I think we’re on Brightspace. Now, we used to be on Blackboard, you know, but it doesn’t really matter what you’re on Moodle, Moodle, everything. And when I upload a document now, so let’s say I’ve produced a slide deck in PowerPoint, I upload that in PowerPoint, and what ally does is automatically produce that as a PDF. Yeah, and as a text file. So all those conversion issues that I kind of, from a UDL perspective was thinking I have to offer this in multiple different formats. Now Allah just does for me, yeah, easy, easy, you know, and it’s kind of just there. So it makes it very easy. Not only that, what it’ll also do is it will scan your documents. And you get this little you get this little gauge, yeah, this little halfmoon and it tells you how accessible your documented. So if you’ve uploaded a scanned copy of an old paper, that hasn’t been OCR, you know, and it’s kind of just basically an image you kind of get a very low score on the on the accessibility thing. Whereas if you’ve done a good job on your PowerPoint, and you’ve pulled out text in and it’s high contrast and you know all the good stuff. Then you get a nice high score, but it’ll also give you like tell you, you need to put out Text on these these different images in your PowerPoint presentation, you need to change the font because it’s not, you know, dyslexia friendly or whatever. And that’s fantastic. Because that allows you both to go over existing material and improve it from an accessibility point of view. But it also trains you into when you’re creating new stuff, you’re kind of now thinking, Oh, I just need to be aware of or, you know, Allah is gonna ask me about
Lillian Nave 30:28
is this gonna pass by Accessibility Checker, which
Joe Houghton 30:31
is so good. It’s an ongoing kind of reinforcement of an increase of your own internal knowledge on this stuff. So Ally is something that I would recommend any institution who has a learning management system looking at, and I have no shares in ally. It’s just that from a, from an educator perspective, and also hearing the students talk about the benefits that they’ve got from from it. Yeah,
Lillian Nave 30:56
that things that I just didn’t even know, weren’t accessible. That will tell me to start thinking in a new way. And I know, there are a lot of folks who’ve been teaching as long as I have that have favorite sources. And like you said, they’ve, they’ve had it for maybe 20 years, and it’s become a PDF or a copied piece of paper with handwritten things on it, you know, all these things. There’s like, This is gold for my class. And then you realize that a student for them, it’s a black box, cannot even decipher it. And it’s so important. And you didn’t realize what you were doing was giving your students a black box that they couldn’t even open when you tried to give them that
Joe Houghton 31:39
material. Yeah, so allies is something that immediately jumped out when you ask that asked me that question. Great. Another one that came up probably about a year ago now. And it’s kind of linked to AI. And I think I emailed you to introduce you to Beth. Beth Stark didn’t know. I don’t know whether you’ve had chance to follow up with her. I haven’t haven’t seen whether you did or not. But, but best doc and Jeremy roster collaborated and created about a year ago. A an AI chatbot called Ludia L U D ay ay. I’ve got an email with links to all the stuff we’re talking about that I’ll send you straight away after the after the recording. So but Ludia is is free, accessible to anybody you just go to? I think it’s poll.com/ludia. And, and you can use this. And it’s a chat bot that’s been trained on all the good UDL stuff. Yeah. All the cast stuff, and lots of articles and best practices around UDL. And it continues to be developed by Beth and Jeremy. And it’s fantastic. It’s like check GPT for UDL, and you can ask it questions about, you know, cost material that you’re putting in or whatever, how do I make this more UDL friendly, and give me five different options for expressing this stuff in different ways to the students and so and it’ll, it’ll give you an it’ll link everything back to the 31 guidelines, and you know, the the main UDL kind of ideas and all the rest of it. It’s absolutely fantastic.
Lillian Nave 33:18
Yeah. So we’re gonna link to that. And yeah, I need to follow up. That’s another episode already. I can tell you another episode.
Joe Houghton 33:25
Beth is brilliant. Yeah. So she’ll make a really good interviewee as well. Yes. Yeah. Super. So Ludia is, is one to have a play with, from definitely more from an educator perspective than a than a student perspective, perhaps. But even if you’re just getting to grips with UDL, and you’re just trying to get your head around how you could link UDL into what you’re doing. It will make those kind of linkages for you and help you so yeah, that’s, that’s a really good one. Yeah, perfect. Okay. And I suppose my third one would be otter. Otter AI. Yeah. otter.ai, ott er.ai. And this is a transcription tool. It’s now been, to be honest, built into, I think it’s built into zoom and WebEx and teams even as well now, but you know, often it has to be turned on, to work in those in those environments. But if you go to otter.ai, you can sign up and you can run this in the browser. So I could have another browser window open while we’re talking now. And I could tell otter just hit record, and it would create a complete transcript of our conversation. Not only that, it would at the end of the transcript, it would kind of turn to for a minute or two, and then it would give us a kind of highlights reel, and also an Actions list of kind of things that you’d agree to do. And I’d agree to Do and you know, because I’m logged in and you’re logged in it would know who. And once once you tell it right, this paragraph was Joe talking and this paragraph was Lilian talking, it remembers your voice prints. So if you do another recording, you don’t have to say That’s Joe talking. And that’s Lillian talking, it’ll just say, That’s Lily. And that’s Joe. So it’s fantastic. And the free version gives you I think, 600 minutes a month of transcription. And then if you pay five or $10, you get another $6,000 a month, 6000 minutes a month or something. So that’s a really good one, because you can use that in meetings. You can also upload recordings, either audio or video recordings into into it, and it will give you then a transcript after the fact. Yeah. So that’s, that’s a super tool, because, again, that goes to multiple means of engagement and representation doesn’t. And full
Lillian Nave 35:59
disclosure, I use otter AI to transcribe all of my episodes. And yeah, once I add the names in it, and spits it out, and it’s pretty good. And only have to change a few things with it. But it’s fantastic. And I’ve gotten a lot of feedback from folks who I’ll say, Listen, using air quotes, again, to my podcast, because it has a transcript, and they read all the transcripts, instead of listening to our voices, which is fine with me. Yeah,
Joe Houghton 36:36
I mean, sometimes, you know, if you don’t necessarily listen to translate, read transcripts yourself. You might think, you know, am I is this a waste of time as I’m I kind of, I recently, I mean, one of my other hats that I wear is that I’m a trainer of photography. So, so I recently finished a a series of 20 talks that I did on Adobe Lightroom. For members of the royal photographic society, I think it was, and, and I always video, record the talks, and make those videos available afterwards. But I also provide a transcript. So I ended up with, you know, 20, Google folders, 20, Google Drive folders with the videos in and the transcripts in. And one of the guys at the end of the course, towards the end of the course said he said, Joe, is there any way of combining the transcripts? Because at the moment, I can’t search through the whole course. And I said, no problem at all. I mean, I’ve got Adobe Acrobat, I can just take the 20 individual PDFs of the transcripts, create a single document. So I did that took me about half an hour. And I ended up with a 720 page document. I mean, it’s just 720 pages of text. But it’s fully searchable. That’s right. I just emailed it to Kevin and everybody else in the course. And they had this other resource that we didn’t have before, which gave them full access to everything. We talked about over 20 sessions. And I had not even considered doing it. But you know, it was and that opened my eyes because that was a UDL moment. Yes. Right.
Lillian Nave 38:22
So bringing in AI back into the conversation, you know, now that’s a searchable database, or any of my podcasts are searchable, because they have the transcript. So if somebody wanted to find out and the over 100 episodes about, you know, talking about AI, or whatever, using a particular tool, it’s searchable, you can Google it, it’ll show up. And that’s hard to do. It’s so much more time consuming if I wanted to go through an audio file, and try to find out where somebody said that that’s just not accessible.
Joe Houghton 39:01
And yeah, and the auto transcripts are also timestamp. Yes. So so when you get the auto transcript, it says 3705, Lillian said. So now you can scrub through the video to 3705. And now I can see you saying whatever it was, you were saying? Yeah,
Lillian Nave 39:17
it’s amazing. This is such a meta conversation right now. Talking about how I put the end of the podcast conversation together so people can find this information. And I thought it would be totally off brand. If I had a podcast that didn’t have the transcript. It would be exactly the opposite of UDL.
Joe Houghton 39:39
Yes. I mean, Louie Louie Lord Nelson is good at this as well, isn’t she? I think you interviewed her fairly which
Lillian Nave 39:45
I did the gears the UDL gears Yeah, the UDL gears. That’s right.
Joe Houghton 39:49
I mean her her podcast you know, she she she again multiple means of representation after each podcast she goes even further does
Lillian Nave 39:56
she does a great job. She has like a little questions you could use Is that as kind of a learning tool? Yeah, I have some things I could improve. They son, Louis, she’s amazing. We learn from
Joe Houghton 40:08
each other. And I mean, we learn from people who are doing this well, don’t we? So I mean, that’s, that’s, that’s brilliant. Yeah, that’s what the community is about, you know,
Lillian Nave 40:14
and it’s such a giving community, you know, and I make so many mistakes. So I really appreciate how others bring me along and teach me and I learned from so many. Yeah,
Joe Houghton 40:24
I mean that that’s a nice segue into into a few new members of my UDL community, if you like that I encountered last year. And I saw an email from one of the boardrooms UDL forums come out, and it was talking about, they were looking for contributors to a forthcoming UDL book. And it just kind of I don’t know why, but it just piqued my interest. And I kind of ping the ping an email back, and it was Diana Rocco Goodwin, universities. And you know, Diana, yeah, I think you’ve interviewed
Lillian Nave 40:59
them and lots of folks that Goodwin they’ve got a lot of UDL going on there. So
Joe Houghton 41:03
she was putting together the new directions for Teaching and Learning edition for Wiley on UDL. And, and I kind of reached out and she kind of was kind enough to reach back. And, and I ended up writing the first chapter of that book during the the intro, which was, you know, huge honor for me, because I was expected to kind of be even able to contribute, but But anyway, I did. And she’s got I mean, she she has this amazing network. She’s plugged in, she knows everybody.
Lillian Nave 41:33
Very dynamic person. Yeah. Yeah. So
Joe Houghton 41:38
again, I mean, we’re, she’s roped me into help the seminar series that they’re doing this year. So I don’t know when this episode will go out. But I mean, that seminar series is going to run probably until August, September time, I think. 24 Oh, great.
Lillian Nave 41:51
I’ll put a link in our resources to it. Yeah. Because this will come out in early 2020. If we get to
Joe Houghton 41:56
nine seminars on UDL, all aspects of UDL. From, you know, a lot of very good UDL people. And for some reason, they’ve let me do one of them. And I’m going to do it in collaboration with them with a PhD, somebody who’s studying for a PhD down in Australia that I run into a UDL conference about two years ago. And we’ve, we’ve maintained contact, and we’re doing little bits of work together Elizabeth hitches she, she works at Macquarie University, in in Australia. So that’s, I mean, it’s just so exciting to kind of sit in my kitchen in Dublin, and have these links to Goodwin, and Macquarie and you know, and Appalachia state and all these other amazing places. And not have to move from my kitchen table.
Lillian Nave 42:50
It’s amazing. The best part about having this podcast is talking to people all over the world and learning from so many different people. It’s just fantastic. And I’m so glad to share it with other people. And I get to do it’s just
Joe Houghton 43:02
amazing, isn’t it? It’s, it’s, and I never get tired of it. I never get tired of the magic. Yeah,
Lillian Nave 43:09
it’s so fun. So can you tell us more about some of the benefits and challenges you have apart on both of those in your book about using AI to make courses and teaching more accessible, that we’ve talked a lot about the benefits. But I thought you could give us some more nuanced understanding of that. One of
Joe Houghton 43:27
the biggest advantages for most people at the moment with the new AI stuff is breaking blank page syndrome. A lot of people have a problem starting Yeah, write an essay on Yeah, write a paper on, put a presentation together on and you sit there don’t you’re just looking at PowerPoint and looking at word. Where do I start?
Lillian Nave 43:47
Yeah, that executive functioning problem about just getting started on a project, the
Joe Houghton 43:52
overwhelm can be significant because you’re under pressure, because there’s a grade attached to it. Because you know, you’re in competition with other people, whatever it is, so so that feeling of overwhelm is a big executive function issue for some people. Yeah. So what I say is take any brief that you’ve been given, whether it’s a one line question, or whether it’s a whole page or whatever, and just stick it in chat GPT and see what it tells
Lillian Nave 44:16
you. Yeah, it gets it breaks that surface. And
Joe Houghton 44:20
it can be chat GPT, it can be copilot, it can be perplexity, it doesn’t matter which one of the chat bots you use. Just stick it into another chat and just see what it gives you. And that will start the juices flowing. That will break the logjam of I’ve got a blank page in front of me. And now you’ve got three points or 10 points in front of you or a couple of paragraphs of stuff and you can read it and you can and it’ll start your thinking and it’ll start allowing you to put some building blocks down and it’ll start showing you a way forward. Now in some cases, it’ll may give you most of the building blocks and it might put even put them in the right order. Not most of the time, but some of the time but at least you’ve got something to start with. And for me, that’s one of the huge benefits. There’s very, very few pieces of work that I do now that I don’t use AI to just at least give me a starting point. I mean, I was asked to structure up a new series on AI for business, and a series of new talks for AI for business. And I was asked to do this by training company this morning. I sat down, and I spent 20 minutes. And I brainstormed a set of ideas, and I just wrote them down. And then I put the question into chat GPT. And I said, right, structure me out as a series of, you know, 15, half day talks on AI, and how that could help business and what what topic areas might I, you know, do a session on and all this kind of stuff. And it gave me most of the ones I’d come up with. And it gave me another four or five that I hadn’t even thought of, and two or three of them a really good, you know,
Lillian Nave 45:57
it’s like, when I asked students to do a think, pair share, exercise, and when you don’t have somebody next to you. And I find a lot of students are very timid, they’re dealing with something new, it’s very risky. And so if you say, Okay, talk to the people around you, if you’re in a, you know, in person class, then they can say, Well, I had this idea. And then somebody else would say, Oh, I kind of thought about it this way. And then you start building together, and chat GPT. And all those other ones, offers that sort of sounding board, you’re still in control, you are the only human. But like you said, it stops that blank page. And instead of avoiding it, like don’t use these things, this is cheating, right? Or something like that. I definitely see it as a tool that we can play off each other. And, and mine, lots of data, figure out what something I might not have thought of. But you’re in control, we have to remember that.
Joe Houghton 47:07
It’s a study buddy. And I think if if you introduce AI in that way. And I mean, I teach mostly master’s students in my former university setting. And I mean, that’s how I introduced chat GPT. To them, I say, I say I expect you to be using these tools from now. And that gets kind of what, what, what, what we’re allowed to, it’s not cheating. And then I explained to them, I say, Yeah, because this is just another form of information access. And I’m not going to say you can’t use Google search. I’m not gonna say Google Scholar is offline, too. Well, why would I say chat GPT is offline to you. Now. It’s out there. It’s a tool, it’s available tool. I want you to use these tools, but I want you to use them properly. Here’s some ideas on how to use them. And, you know, as educators, we need to create safe psychological spaces and safe environments where we give students some guardrails and some ideas of what is okay and what isn’t. I mean, something popped up in my feed this morning, an article that said, you know, plagiarism is not a sin. i It’s funny, I didn’t have time to read on it, but I thought, well, you know, that’s a really interesting concept, isn’t it? Because, you know, in a lot of academia, plagiarism actually is a sin. Yeah, it really is. But is it a mortal sin? You know, in the in the in the overall scheme of things? Probably not, but from an academic perspective it is. So what what are the, what are the guardrails?
Lillian Nave 48:42
Yeah, what makes this plagiarism? Right? And what happens if you’re using a tool, you get some information? You cite your sources, you explain what it is? Right? Yeah, you’ve got the right car.
Joe Houghton 48:55
So I asked my students to tell me how they used AI. And most of them this term, you know, when I’m reading stuff that have come back, they’ve declared the use of AI, I would say, in the in the 200. Assignments, I’ve just finished grading 70% of them declared some kind of use of AI. And of that 70%, probably three quarters, were using it for ideation, they were using it for give me initial ideas, give me a structure, give me points that I should include, and then they were going off and amplifying the points and presenting them in different ways.
Lillian Nave 49:32
Right, you know, we wouldn’t get mad at somebody whose text autocorrects their bad spelling, right? We don’t say, Look, don’t text me, unless it’s all your words. Your spelling, right. It’s just so ingrained, and it’s helpful. And yeah, I mean, for the most part, there’s sometimes autocorrect gets us in trouble.
Joe Houghton 49:56
Yeah, but I mean, I’ve got a lot of my students cohorts are drawn from 15 countries, something like that. I mean, it’s a very international. Now, it’s masters level. So you’re expecting a level of English and we have a, you know, a kind of minimum level of English proficiency and all the rest of it. But a lot of people, even even in my classes, who have English as a second language, maybe their spoken English is better than their written English. Yeah. And a lot of them are declaring that they’re using chat GPT to kind of rephrase and kind of make sure that the English that they’re using is, is okay, so they’ve written in their version of English, and then they’ve run it through one of the API’s, and it’s put it into something that they consider to be better English. Well, I have no problem with that. I mean, it’s still their work. Yeah, you know, so. So I think this way, the use of AI tools to help people represent what they what they think, and their ideas in different ways. I think that’s just becoming the norm now. And it should be it will become the norm. And it’s not always the norm at the moment, but but I’m trying to make it so
Lillian Nave 51:11
we’re seeing a lot of the positives. Again, we’re strength based here. But are there concerns, challenges, and even ethical concerns that we should be aware of when we’re using AI in the context of UDL and accessibility?
Joe Houghton 51:30
I haven’t, when I’ve kind of went certainly when I’ve gone to things like Ludia, I haven’t seen anything come back that I kind of saw immediately as a red flag. And very often, you know, if I’m doing that, I’m probably putting in stuff that I am, you know, reasonably expert in from a subject matter perspective. So I’ve got that filter that I can kind of look in and kind of do a check. I don’t tend to just accept things straight away. Now, from a student’s perspective, that’s one of the things that I say, and we actually do exercises in class where I’ll say, right, go off and use one of the AI tools to find out about this. And they go off, and then they come back, and then I and then we’ll say right now go on factcheck. Go and check that elsewhere. Okay, use Google’s search, use Google Scholar, go and find some citations that back this up. And you know, how many times is the AI hallucinating? How many times is it giving back? What seems to be a valid source, but when you click on it, there’s a 404 error. It doesn’t exist, and that kind of stuff. So just exposing students I think, and other educators to the fact that it’s just because it’s on telly isn’t true. It isn’t true.
Lillian Nave 52:40
What just because it’s on the internet, it’s not true.
Joe Houghton 52:47
I think that’s still an important step we need to take people through. Because it can be compelling. When when it spits out this amazing looking page of information. You read it and you think, Oh, wow. Yeah. And it’s kind of like it must be true. Yeah, but it isn’t necessarily true. So you have and I think some of the tools are better than others. And some of them are better than others at different things. So and they’re changing all the time. So this is quite difficult. And you know, you need to find one or two people to follow in the space
Lillian Nave 53:29
where you are that your subject matters. something out
Joe Houghton 53:33
of Wharton would be would be one person that I follow. He’s a professor at Wharton. Mushtaq Bilbao does a lot of work on AI for educators and particularly for researchers as well. He’s very good, Amanda Bickerstaff is, is very good. And he’s putting together a really good team. And they’re working with a lot of educational institutions and stuff now, particularly in the US. And then Beth Stark and Jeremy Rosner, as well. You know, these, these are people to kind of add to your follow list if you like. And I mean, I regularly cite all of them and other people in that LinkedIn AI newsletter that I put out each week, so that’s yeah.
Lillian Nave 54:17
And I’ve found too, that whatever an AI model spits out, is based on whatever is inputted and depends on where you are like, is this very united states centric? Is it global? It and what sort of bias creeps in to whatever tool you’re using? And so I think being aware of those things, like One early example, as if you would ask one of those creative visual AI’s, show me a picture or have a professor, and they would usually be an older white male, you know? Yeah. Which is less and less the picture of a professor, you know, or who would be a doctor who would be a nurse, you might see males in the doctor role females in the nursery. And just the input of what goes in all these models and where that data is coming from, we just have to be aware of the prompting
Joe Houghton 55:28
stuff. I mean, there’s a whole industry coming up around prompt engineering. And I mean, it’s a conversation you have to have, you need to treat these AIs as another person, and you need to have an interactive dialogue with them. And at the end of, you know, a set of questions where I’m looking at something or researching, there’s a couple of things that I now Now throw in towards the end. One of them is, am I making any assumptions? And if so, what are they? Yeah, because that can really throw out some quite interesting responses back from an AI. And the other one is, Am I what biases either have i or are you exhibiting in this conversation? And how do we compensate for those assumptions and biases, if you throw those into the conversation, you know, towards the end, and they can surface? Quite some quite interesting thing?
Lillian Nave 56:19
Yeah. And we just have to be aware, and know that it’s not 100% Honest, all the timer. Yeah. And, and we’re still the expert, right. And we have to be still using all of our critical thinking, to mold what we get. But it is this fantastic ideation generator. But just like in everything, you’ve got to check your sources and verify it’s getting
Joe Houghton 56:54
better and better and better. And I think I think from an accessibility point of view, to a large extent, this is leveling the playing field, because it’s giving people who never had access to these wide ranging sources of information, the ability to pull on wider datasets and bigger sources, and it will become available on all the platforms on every phone on every computer. And so the digital divide thing will will lessen over time. I mean, I love talking to chat GPT on the phone now, the phone app for chat GPT, where you can just talk to it and have a conversation. I mean, that’s how I use my car time now. And then when I get home, the conversation is sitting in chat GPT is a text file. Wow.
Lillian Nave 57:38
I hadn’t thought about that. Yeah, absolutely brilliant.
Joe Houghton 57:41
I can structure a whole session up in the car, and then come back and kind of pull the pieces together and then throw it into something like gamma. And I’ve got a presentation in five minutes.
Lillian Nave 57:53
Wow, that’s fantastic. So what do you see on the horizon? For AI UDL and accessibility,
Joe Houghton 58:01
I think I mean, AI is just going to become built into everything, we’re going to see that in 2024, it’s kind of just going to become mainstream, it’s going to stop being this thing that we have to go to, and it’s just going to become part of our daily use. So that’s, that’s good. From a UDL perspective, I think it allows us, I mean, we’re already seeing, it’s now very easy in all the API’s, you know, to generate images, to generate text. Increasingly, it’s becoming easier and easier now to generate video. Yes, I’ve seen examples, give me give me a three minute video on x. And it’ll generate this video
Lillian Nave 58:45
for you can immediately translated into Mandarin, Portuguese, whatever,
Joe Houghton 58:50
you’ve got multiple means of representation there. And different people coming at, you know, with different ways of kind of preferring to express themselves will be able to do that in different ways. Now, what we have to do is, as educators is, is allow that, you know, I mean, I was reflecting and most of the stuff that I got back last term is written pieces of work. Two of my students sent in, I’d asked them to kind of, you know, do a process analysis of a personal branding website project that they had to do. So they had to create themselves a personal brand website. And I’d said write up a process. Your process of doing that most people wrote a Word doc, quite a few people did put it into Canva and did it as a nice looking kind of consultancy. Report. A few people used PowerPoint slides, and one student sent me a video. Yeah, and talked through their process in the video. And I thought, great, I love that. You know, it was different. It took me two minutes to listen to the video.
Lillian Nave 59:58
Yeah, it’s important Seeing for you. Yeah,
Joe Houghton 1:00:00
so maybe I need to change the brief. So that it doesn’t kind of direct everybody down the automatic path of this has to be a 20 page report and allow them to be more creative and stuff. But that’s a conscious thing that I need to kind of take it out of my preconceived idea of I’m going to get report,
Lillian Nave 1:00:19
right, you get you have a structured goal, you know exactly what you want. But they’re flexible means to get there. It’s just UDL all the way.
Joe Houghton 1:00:28
And there’s so many tools now that will allow these different ways to represent what it is that you want to tell me. Yeah. So I need to encourage people to adopt those different means. And that means I’ve got to open my head when I’m specifying the goals and the, you know, give them giving them the brief and whatever. And I think that’s a challenge for all of us as educators is just to open our heads a little bit,
Lillian Nave 1:00:54
it really is. And I, before I knew anything about Universal Design for Learning, I had an art history course I was teaching. And it was about women artists, and it was an upper level course. And so small seminar, and and project was presentation, the research, paper and slides, of course, because it was art history, and I had a student who was also a dance major. And she incorporated modern dance into her art history presentation, which was about an artist of the same time as kind of the birth of modern dance. And I remember being transfixed, never would have asked for that. Never would have thought of that. Never could do that. And I was close to kind of tears, watching this. And thinking, Wow, what an amazing connection she made that never would have made. And what an amazing presentation just amazed by her artistry. And I didn’t know that was UDL. But I was totally changed. Yeah.
Joe Houghton 1:02:09
And I mean, one of the reasons I love that assignment coming back whether the personal brand website is that I say right, you’ve got a chance to put a website together that represents you in a different way from LinkedIn, or your CV. And I learned so much about my students, I learned the fact that you know, they’re they are dancers, and they are musicians, and they they play sport, or they collect garden gnomes or whatever it is that they do. And it opens up them as people and which is just wonderful. It’s wonderful for me, but it’s also wonderful when they put that out in front of potential employers. Yes, you know, and get jobs because they’ve they’ve shown another side of themselves.
Lillian Nave 1:02:47
That isn’t asked for but you didn’t know you needed that. I’m a Business School.
Joe Houghton 1:02:51
I can’t tell them that I play the flute. Right. Yeah. But but you play the flute really? Well. Yeah. It’s an expertise. It’s another thing you’re good at. So it shows you’re not just a swat at school, you’ve got multiple levels of talent and stuff like that.
Lillian Nave 1:03:07
Oh, all right. I’m in makes me think of a colleague of mine, an amazing woman who has taught in our university and teaches writing, like rhetoric and composition. And she changed some of her courses. And saw this rhetoric and composition course through the lens of hiking, the Appalachian Trail, which is the whole, you know, eastern seaboard of the United States. It takes about six months, March to October, so to hike it, and she continued to teach her courses while she was hiking, moved part of it asynchronously. Then she had the summer to do a lot more hiking. And then in the fall, connected with students again, started another set of classes. And it was amazing to see her blogging about it writing about it, or the writing process. Still, she would come off the trail, go to a public library, login, meet with her students, and I know finish the entire Appalachian Trail. And never would have known that and inspired her students. And yeah, just something that was so different. And amazing and showcased her thinking invited the students to think differently and displayed that to them. Right this absolutely.
Joe Houghton 1:04:25
Mary’s got to the Appalachians book. That one too. Yeah. Oh, if you’ve never read that book, it’s one of the funniest books that I’ve ever read. The chapter about the bear is just like hysterical. Yeah.
Lillian Nave 1:04:38
And I know that that same mountain range is actually what goes connects in the UK. So a long Yes, eons ago. Those are the same mountains, the same ones on the west coast of Africa, Morocco. It’s the Appalachian Mountain Trail. Well, we call it the Appalachian. It’s a had another name a long time ago?
Joe Houghton 1:05:01
Well, we’ve not we’ve not not normally dipped into geology in this podcast.
Lillian Nave 1:05:09
Exactly. I think you’ve answered all my questions. But was there some anything else that I missed along the way? No,
Joe Houghton 1:05:17
I don’t think so I think we’ve covered everything we said we were going to,
Lillian Nave 1:05:20
which is not the entire book because it’s filled with so much information. And it will give so many folks a clearer picture of what they can do to increase accessibility in their learning environments. So I thank you so much for joining me again, thank
Joe Houghton 1:05:38
you for having me on. Again. It’s been it’s been great to catch up so thank you.
Lillian Nave 1:05:47
You can follow the think UDL podcast on Facebook, Twitter, and Instagram to find out when new episodes will be released. And also see transcripts and additional materials at the think udl.org website. Thank you again to our sponsor, textile. Texthelp is focused on helping all people learn, understand and communicate through the use of digital education and accessibility tools. Texthelp and its people are working towards a world where difference disability and language are no longer barriers to learning and succeeding, with over 50 million users worldwide. The Texthelp suite of products includes Read and Write equates to an orbit note. They work alongside existing platforms such as Microsoft Office and G Suite and enable them to be integrated quickly into any classroom or workspace with ease. Texthelp has changed the lives of millions worldwide, and strives to impact the literacy and understanding of 1 billion people by 2030. Visit Texas dot help forward slash learn more that’s l earn m o r e to unlock unlimited learner potential. The music on the podcast was performed by the Oddyssey quartet comprised of Rex Shepherd, David Pate, Bill Folwell and Jose Cochez, and I am your host, Lillian Nave. Thank you for joining us on The think UDL podcast