Subscribe
Share
Share
Embed
Tales 4 Teaching shares stories of purpose within higher education. Join us for expert insights, engaging interviews and thought provoking discussions that will inspire your teaching. Presented by Deakin Learning Futures.
All views expressed are those of the Tales 4 Teaching team and may not reflect those of Deakin University.
Ep. 84 – Beyond Generative AI: redefining education in the modern classroom
Welcome to Tales4Teaching, a podcast where we explore stories with purpose in higher education. We’ll share expert insights, engaging interviews, and thought-provoking discussions that will inspire your teaching. On behalf of Deakin University I would like to acknowledge the Traditional Custodians of the unceded lands and waterways on which you are located. I acknowledge the Wadawurrong people of the Kulin Nation as the Traditional Owners on which this podcast was recorded, and I pay my respects to Elders past, present and future.
Joan: My name is Joan Sutherland and this is Tales4Teaching, brought to you by Deakin Learning Futures. Hello and welcome to today's episode where we're going to be talking to
Leon: Furze, a consultant, author, and educator who's exploring how AI transforming writing instruction. Hello,
Leon:. Thanks for joining us.
Leon: Thanks, Joan. Good to be here.
Joan: Add to get us started. Can you just tell us a little bit about yourself?
Leon: Uh, yeah, sure. So I come from a secondary education background. I worked in secondary for 15 years, and I'm trained in the UK as an English media drama teacher. And then, uh, moved out here in 2009 and have been teaching out in south west Victoria since then. In 2022 I finished up in that teaching role and started my PhD at Deakin University. it's all just been a little bit crazy from there on. The PhD was in generative artificial intelligence and it started two weeks before ChatGPT was launched, and it's been pretty wild.
Joan: Wow, that would be intense at that time, I would imagine.
Leon: Yeah, absolutely.
Joan: Now, previous to that, you did did you do some work around writing instruction? IS that your expertise?
Leon: Absolutely. I've been an English and literature teacher for the whole of my career. And in 2016, 2017 became a head of English and I've also been on the Victorian Association for the Teaching of English Council (VATE) I was on that council in 2017 and then left for a few years, and I've actually rejoined in the last couple of years. Um, so I've been running professional learning training around literacy, reading, writing for a number of years. My first book was Practical Reading Strategies, which was a book on close reading skills for English teachers, predominantly. And then we followed that up with Practical Writing Strategies, which I co-wrote with my former colleague Ben White, who is an English teacher as well. And yeah, then started to transition towards generative AI around the start of 2022 was when I started to get really interested in.
Joan: Like the other 100 million users that subscribed.
Leon: Well, I was 12 months, I was 12 months ahead of them.
Joan: Okay, so that's good.
Leon: that GPT came out. But my PhD supervisor Lucinda McKnight who's also a Deakin academic. Lucinda's been looking at these technologies since 2020, 2021. She and I, I think we first crossed paths through VATE because she was on VATE Council a while ago, and she's been on many of the committees and in in May 2022, when I handed in my notice as a secondary teacher, a few weeks after that she and I were talking on the phone and she mentioned the DECRA Scholarship working on digital texts. So, yeah, we got right in there, talked about generative AI. I applied for the scholarship, and then the PhD started at the start of November 2022.
Joan: How great's that. What a great, um, initiative and not initiative
Leon: it's good timing.
Joan: Yeah. Great timing. That's one. So a lot of your work, as you mentioned, I heard practical strategies. I'd imagine people are very drawn to the practical component versus the theoretical component. We get a lot of theory in education. So the practical strategies I'd imagine would be very appealing, especially for teachers.
Leon: Yeah I've always had that mindset when when I run professional development and training and when I run, whenever I write any of my books, it's I even, you know, the most recent book is called Practical AI Strategies, which came out earlier this year from the same publisher. Um, bit of a theme with the naming conventions. I'm not very imaginative, obviously, but it's, uh, it works out. Look, I found that even in the Practical Strategies books, there's an amount of theory, but I tend to I tend to front load the practical stuff and use that as a way to introduce some of the theory. So with the AI book, for example, there's a lot in there about AI ethics, policy, assessment practices. But there's also a lot of how to use the technology so that you can find out those strengths and limitations for yourself. So I think that's really important. You've got to have the practical things in there and kind of learn by doing, I guess.
Joan: Absolutely. So you mentioned around the ethics side of things, and I suppose that's pretty cool when implementing generative AI in any setting, but particularly in the educational setting. And it ensures, I suppose, to ensure, uh, students information is protected and used responsibly. What are your top tips then to ensure that occurs?
Leon: It's been really interesting. Last year I wrote a series of posts called teaching AI ethics on my website and um, originally I focussed on nine different areas of ethical concern which were just coming through as I was reading a lot more and more about the AI and that kind of preparation for the confirmation of candidature. So, um, those nine areas, each one broke out into an individual resource, and that part of my website's gone absolutely bananas over the last two years. It's the most trafficked part of my website now, so people are really interested in it first of all and then in particular, they're interested in the bias and discrimination issues in AI. They're interested in the copyright and intellectual property concerns, and they're interested in the privacy concerns, like you said. So I mean, like my my best advice on all of this stuff is to just treat it like any other digital technology, which is to be critical always, always, you know, anticipate that your data is going somewhere and that somebody is going to use it, because that's the model. That's the business model of all these
Joan: Unfortunately and they always say there's nothing for free in this world and a lot of tools you're always paying for something. paying with it somehow. If you don't paying monetary, it's with your data as well.
Leon: Well, I mean, this is really important because just yesterday 4am Melbourne time, OpenAI released their spring update and dropped ChatGPT-4.0 on everybody. So they've made their most powerful model free for everyone. But that model has image recognition, it's got voice to text and text to speech synthesis. There's a lot of like, real kind of concerning from a privacy standpoint, technologies that are now bundled into this app for free and in the hands of 100 million users worldwide, including students in classrooms, including teachers. So there's a there's a great post this morning on LinkedIn by a guy called Marc Watkins. Marc with a C, he's got a Substack like everyone else. But I think, Marc's Substack is really good because it has that critical standpoint. But he wrote about, why are we not holding ChatGPT up to the same kind of standards as the rest of edtech? Because, you know, it's essentially edtech, but it's not complying with any of those data privacy regulations at the moment.
Joan: Yeah. It reminds me of the AI dilemma by the Centre of Humane Tech and how they talk about those type of things and getting so intertwined that we need to hold them accountable as well.
Leon: Yeah, absolutely.
Joan: So on that. So how how is generative AI redefining traditional teaching methods?
Leon: Um, yeah, it's it's such an interesting conversation because I try to not be techno deterministic, and it's really hard to avoid that kind of language where you say like oh you know, AI is going to change or AI is going to revolutionise or AI is going to threaten, or because AI, you know, won't do anything. It's a technology. It's a platform. The way that we use will be what changes things or not. And what I'm seeing across education, particularly in K-12 where I work the most, there's still not a great deal of adoption right now. Most um educators, if they've used anything, they use chat GPT 3.5, the free version when it was first released. They weren't super impressed. They haven't used it since, um, a lot of students of the same, you know, they've used Snapchat AI, they've used ChatGPT, they haven't really used much else in higher education. I'm even finding that this, you know, this pretty large pockets of people who haven't had the time or the inclination to engage with the technology yet. So I think actually the biggest thing that's the biggest sort of implication in education is that the technology is just rolling on without us, and I've been saying recently, particularly with people, when they're kind of avoiding the technology, at this stage it doesn't really matter if you think it is good or if you think it is bad, because it just is. And I think that's, you know, a huge implication for education is that the technology companies are constantly updating these technologies, rolling them out in competition with each other, um, in a fierce kind of arms race, particularly you know, Google, Microsoft and OpenAI, they're not going to slow down and wait for education to, to catch up. And I think we will find ourselves in a position where students are using these technologies in ways which we wouldn't consider appropriate or ethical, and we would maybe want to offer some more guidance. But we can't do that unless we engage with the technologies in the first place.
Joan: We we can't do that unless we engage with the technology ourselves, but is it at an individual level or do you think, um, it's got to be promoted at a school level or a government level? Like where does that governance I suppose sit?
Leon: I think it has to be at all levels. Regulation has a huge role to play and it is not it's not keeping up. And I think we're getting this kind of, I guess you could call it regulatory entrepreneurship from these Silicon Valley companies where they move fast and break things and then get too big to ban. You know, OpenAI, they've scraped all of that intellectual property and data. They've built an incredibly powerful model, handed it to Microsoft, Microsoft have put it in everybody's hardware and software all the way through and now we're in a state where we can't kind of go backwards, and we we just haven't regulated anything like ChatGPT. And yet, you know, there are child online protection acts, there are privacy laws, there are data regulations like the GDPR and these companies are just rolling over the top, so regulation has a huge role to play. Yeah. But then at the individual level, I think educators have a responsibility to understand these technologies and the implications and then in that middle level, I think it's really interesting. I've been doing a lot of writing recently around the importance of middle leadership in education context and even having a generative AI strategy on a faculty level, because it will impact every discipline and it will impact every discipline differently. So the math department will respond to these technologies differently to the humanities or the English or the science department and I think everybody in a middle leadership position is going to be really responsible for how this is impacting their schools.
Joan: So if we look then at the individual level, what can educators do to evolve their practice in the current landscape. Because as you said there's so much happening and it's evolving so quickly. What are your recommendations to evolve their practice?
Leon: It's really hard to stay on top of all the developments. So, don't bother. Well that would be my first advice there. Um, people keep asking me, like, how are you staying on top of all of these changes? And my answer generally is, I'm not. A lot of it kind of washes over me as well, even though this is my full time job. So that's hugely overwhelming for people with actual day jobs, unlike me and I think that what you've got to do is you've got to find a few kind of trusted sources of information, a few people that you need, you can listen to that might help you guide through this and, you know, there are people like us out there who have come from an education background and who have an interest in K-12 and can work with these technologies. There are, you know, there are plenty of people out there, even across Australia, people like Nick Jackson in South Australia, who's a digital technologies coordinator, Matt Estamon, who's up in Sydney, Miriam Scott in Queensland, there are K-12 educators in the classroom deeply invested in these technologies, pumping out resources to help teachers. And if you can follow them on social media or something, you've probably put yourself ahead of the game. But don't try to keep up with every single change. Just concentrate on what directly affects you and your discipline and the subject that you know.
Joan: It's a tough one because it is over well, it can get overwhelming for people to stay on top and I think especially in the education sector where you hear certain things that are happening and then I'm always intrigued when you go out of the sphere and I talk to peers and they've got no idea what's happening. So I think sometimes you can get caught up in this race as well but understanding the impacts and critically evaluating it to embed it potentially like any other technology, that you suggested earlier.
Leon: And, and, you know, we can't allow this to be driven by edtech and not just the big technology companies. They are absolutely driving the broader narrative, Open AI, Microsoft and co. But there are smaller edtech companies aggressively building on top of AI platforms and marketing in to schools at the moment. And they are driving the classroom narrative, you know, that they're selling their products to assistant principals, principals, directors of curriculum, that kind of executive leadership who, you know, understandably, are looking for solutions, uh, you know, things that can help them out as quickly as possible, things that, you know, they argue that they're reducing teacher workload. All of these kind of hot ticket items are being really aggressively marketed into schools at the moment. And I just, you know, I'm always cautious. I was in secondary education for 15 years. I've seen a lot of rubbish edtech, and I don't think the edtech companies should be driving how this technology is implemented into schools.
Joan: Into the classroom, and into schools as a whole as well.
Leon: And, um, you know, there are great people working in edtech. I don't want to tar everyone with the same brush. You know, there are people out there that have come from K to 12 backgrounds and then gone into edtech. There are people who've made their own Start-Ups and things like that. But in my experience, yes, you know, it's vapourware. It's it doesn't it doesn't really exist. It's it's a product or, you know, a solution looking for a problem or however you want to phrase it.
Joan: Absolutely. So we in higher education, obviously we get students from mature students. We get all different age groups, but we get a lot from K to 12 as well, finishing year 12 and then coming into higher education. What if you wanted to get educators in higher education, uh, know the top three things that they should know about students coming through around their knowledge or around generative AI, is there anything, um, that you would like to impart on them around, um, what they can expect from students that you've found in your research?
Leon: I'm seeing emerging studies at the moment that are doing sort of focus groups with students and surveys and interviews with students. There's a lot of research of that nature coming out through the peer review pipeline now, which obviously is already kind of outdated by the time you get to that point. Um, so, you know, it is worth keeping an eye on the preprints as well for from that perspective. Um, because there's a lot of that happening at the moment. But a lot of this research from students point of view is suggesting that students just want to be helped in understanding how to appropriately use the technology. They're not looking for ways to cheat or game the system. You know, there'll always be, what, 15, 20% of the student cohort who will cheat and who will game the system and we need to change the systems around that to kind of disincentivize cheating. And, you know, there are some serious conversations to be had there around the kind of culture of academic performance through the VCE, HSC kind of pathways into tertiary, because I just don't think that model works anymore. I don't think it's worked for a long time, but but those kinds of systems, they incentivise cheating behaviours and in those cases we will get students coming into higher ed who have very much gamed the entire system all the way through using generative AI. I made some posts when, when GPT 4.0 was released where I was just sitting at breakfast and running through examination papers with GPT 4.0. I uploaded the exam paper without even prompting the model, just upload the PDF, submit, answers all of the questions. And you know, it got 100% on specialist maths. It got 94% on software development. It got 113 out of 120 marks on physical education. It can do methods, visual arts, biology, physics. It's yeah, we've got a competent model now that can handle that kind of content-based knowledge curriculum across every domain. So those are the students that are coming from secondary into tertiary at the moment. They have this technology in their hands.
Joan: So I'm hearing then the application and the practical components going back to your practical strategies, the practical components within the teaching and learning landscape are probably the more critical component because from a content knowledge perspective, yes, you can get the answers, but whether that application is available to them is another matter.
Leon: Yeah. I think we need to be really conscious of what skills we consider foundational and really important and worth teaching, and why. We need to really focus on the kind of human side of teaching those skills and, and not offloading them onto technology. But we also need to work doubly hard now in education to convince students that those skills are worthwhile. Because, you know, I worked in education long enough to have come past hundreds of apathetic teenagers who couldn't care less about me teaching them persuasive language analysis or how to write an opinion piece. And that now is compounded by the fact that they have a technology which can do that at a level higher than most of those students anyway. So given that we have this technology, this this thing exists, students have access to it, how do we come to terms with the fact that some of our kind of fundamental parts of what we teach in our disciplines is now, is now gone? You know, that's and that's a hard that's a hard thing to come to terms with and I think a lot of people are just, uh, ignoring it, maybe, because it is so hard to come to terms with. As a writing teacher myself, I think, you know, what does it mean now that we have a technology that can write better than many adults, let alone the students that I'm working with? And how do I impress upon a cohort of students the importance of writing? And you know, and for every student, do I need to impress upon them the importance of writing, or is this just, you know, have I put writing on an academic pedestal because of the education system that we work in? So these are existential questions for teachers.
Joan: And it won't be answered in one podcast episode, that's for sure. But going back to those skills, I did a podcast recently with two students, our students, and they they were part of IT peer support program. And their, their biggest outcome was not learning any technology, the communication skills that they built from the start to the finish of that program and I thought that was an unintended outcome of the actual program itself but they actually learned how important communication was and in the conversation, one of them mentioned she knew how important communication is and how she's building her skill set now. So I wonder if it will come full circle. Students will start to understand, hopefully with the guiding principles in education.
Leon: I mean, everything is changing and it, you know, the whole kind of employment landscape shifting beneath us to. Education and training is often a slow moving beast because it's, you know, because it's such a huge system in there's so many moving parts and because it's a system that really centres on humans and humans are messy. So there's no there's no easy answer here. But I think that people in positions of school leadership, in higher education leadership, they need to be looking forward to industries where, if anything can be automated, it will be. Yeah. And if things can be replaced with generative AI based technologies, like chat bots or voice assistants or anything for, you know, analysing large volumes of text or multimodal data, they will be replaced in automated. So you know, what roles does that, create for us as well because it does create opportunities.
Joan: There's two lenses isn't it. There's like you can get bogged down and really think, how hard it is sometimes, but there are a lot of opportunities with it as well.
Leon: You know, I'm reading a book at the moment, by a guy called Mike Walsh. The book's called Algorithmic Leadership. And you know, some great points in there and he touches on economics, he touches on education, it touches on a whole broad range of ideas. But with automation, you know, he says, you can look at automation as something which is going to decimate your workforce, or you can look at it is something which is going to elevate your workforce. And really the way to think about that, particularly in education, is you know, there are aspects of education that will be automated away, but that doesn't mean that teachers will be automated away. It could mean, in fact, that teachers become more important because, we've seen in the past different industries, you know, when automatic teller machines, ATMs came into play, we didn't get rid of all of the bank tellers. The number of bank tellers globally increased because people were able to open more banks and service more people. We've got a teacher shortage at the moment. If we automated away some of those parts of the job, the, you know, the workloads that contribute to some of those issues, then maybe schools could expand and we could get more teachers in the system overall doing different skill sets. So, you know, it's a really interesting way to think about things. But what jobs does automation create as well as just what does it get rid of?
Joan: Absolutely. Look, I think you've posed some great questions, that we might answer, but it's been great to have you on this podcast. So I really want to thank you for your time,
Leon:. And as we finish up, is there any final comments that you would like to make to the audience?
Leon: Just that, you know, as overwhelming as this technology is, the best way to get your head around it is to is to get in there and use it and learn the strengths and limitations for yourself. That's just been made a lot easier by OpenAI making that most powerful model free. So now is the time to get in there and experiment.
Joan: Look, thank you for your time. Really appreciate it.
Leon: Thank you very much.