Open EdTech in the African context – A recent interview with Martin Dougiamas

The following is a transcript of a podcast interview of Martin Dougiamas by Denzil Chetty on his Ubuntu Nexus podcast, made during the recent Moodlemoot Africa on 26th September 2024. A lot of it is about Moodle, but it connects into a wider important discussion of the importance of Open EdTech and Open Education in general. Listen to it on the podcast or read the transcript below!

Denzil: Welcome to the Ubuntu Nexus Podcast Series. Bridging African epistemology and technology, in this series of podcasts we embark on an enlightening journey through the rich tapestry of African knowledge systems and the cutting edge world of the digital humanities. We explore the confluence of traditional African epistemologies and modern technological advancements, showcasing how digital tools can preserve, enhance and propagate African heritage and wisdom. We delve into stories of innovation, resilience and creativity from across the continent and globe, highlighting the voices of scholars, technologists and cultural custodians at the forefront of this transformative movement. Join us now as we bridge the past and the future, weaving the threads of Ubuntu into the digital age.

My name is Denzil Chetty, your host on this podcast series produced by the University of South Africa’s African Digital Humanities Project. It’s a pleasure to have with us today Martin Dougiamas, an Australian inventor best known for founding and leading Moodle, an open source learning platform widely used by educational institutions globally. His journey with Moodle began in 2001 and his original programming contributions remain a core part of the platform. For over two decades, Martin served as the CEO of Moodle, playing a pivotal role in its development until January 2024. In a recent career shift, Martin moved to become Head of Research at Moodle, leading the Moodle Research Lab, where he engages with advanced technologies, particularly artificial intelligence and augmented reality. Beyond Moodle, Martin founded Open EdTech, a non-profit organisation based in Brussels, dedicated to fostering open technology in education. He also serves as a director of Open Education Global, a non-profit organisation committed to advancing open education worldwide.

It’s my pleasure today to welcome Martin Dougiamas. Thank you so much for joining us today. We really appreciate you taking this time.

Martin: Thank you very much. Pleasure to be here.

Denzil: Thank you. I know you’ve had a busy day so far and the MoodleMoot Africa conference is really something exciting for us. But maybe Martin, from your side, if you can start, how did you actually start with Moodle? What was the motivation behind it?

Martin: Yes, so I started Moodle 25 years ago, something like that. My first sketchings were in 1999. I was working at a university, at Curtin University of Technology in Perth. At the time, my role was an internet consultant. The internet was quite new back then. I was helping all the staff at that university get to grips with the internet. In trying to get them to use the web, I realised there was a huge gap, a huge hole of tools to take advantage of it. I knew that there was so much more that could be done because I myself had grown up with distance education all the way up to the age of 12. I lived in the desert and I did my school on a shortwave radio, so a little bit like this podcast, talking to my teacher just through audio. Through that narrow channel, we’re able to expand the bandwidth of communication. I knew there was much more possible with the internet and I started building these tools and prototyping it. There were several prototypes and experiments. I was teaching a couple of courses as well as working at the university. I was experimenting on students and I built the Moodle that we are still using today, that a lot of people are using today. Obviously, it’s evolved a lot since then, but that started my journey. After a while, I quit my job there and Moodle became my passion until now.

Denzil: Martin, Moodle is literally all over the world. It’s global, right? In terms of impact, if we really look at the impact it’s making in the education space, how do you see Moodle transforming higher education?

Martin: Moodle in particular, it was at the right time to enable the transformation from brick and mortar style teaching in real classrooms onto the internet. The structure of Moodle is something that people understand. It has departments and the departments are broken into courses and the courses have activities and resources and they have teachers and students. It’s a very familiar hierarchy of learning. It’s easy to transfer what was before online. Two-thirds of the world’s universities use Moodle today, but not just universities. It’s also used very widely in high schools. It’s used in even primary schools sometimes. It’s used in associations and various corporate learning, very, very big in corporate learning, and even the military and government and all kinds of sectors. Because it’s a very flexible system that can be customized for any particular learning scenario.

Denzil: I like what you said, is that it’s not just in higher education. While it’s transforming higher education, it’s also in industry, it’s also in the public sector as well. There’s a much wider impact itself in terms of the elements itself.

Martin: Yes, correct. Impact is a word that is used a lot and it’s like a punch, right? Impact. It kind of implies that it comes from me and goes to you. I have had an impact on you. Moodle is more of a project that is collaborative. It’s an open source project. All of those users are also contributors potentially. A lot of the innovations and the features in Moodle have come from this open source community. We’re all contributing to the impact. We’re all contributing to a shared infrastructure that provides support for our individual institutions, but also a commonality and a sharing of ideas and innovations between institutions and between different educators.

Denzil: Martin, I also know that you’ve held the founding position, the CEO position for some time, but just recently you transitioned to head of research. With your transition to head of research, how has this shift in your role allowed you to explore new frontiers in education technology? I know I was privileged to hear a little bit about what you’re doing and it’s fascinating but for those that are listening to us today, how has this new transition to the research role really propelled you to explore the frontiers in education technology, particularly areas like the AI technology advancements that you’re mentioning?

Martin: Yeah, I’ll set the scene for that. My background is computer science. I regard myself as a scientist. I developed Moodle because I was studying this problem and using technology to solve an education problem and Moodle is really right along that niche in between education and technology. As it took off, I had to work more and more on it, had more people to manage in the community and engage with and I had to start scaling up and I had a very deliberate choice. Do I build a large company to do this and be the CEO of it or do I stay with the research and the development side of things? At that time, I actually picked a second option. Moodle remained a quite small company for a long time. I didn’t have very many people I was directly hiring because they were all on development, they’re all developers and I outsourced the business part with Moodle partners. We now have 100 different companies, some of them are quite large with thousands of people who are our partners and do services around Moodle. I was managing that but still things were growing and having partners became also something, a bit of a risk because partners can come and go and so then you’d have not such reliability on your revenue to pay developers. We started to build our own services, we started to acquire partners and sooner or later, I found myself the CEO of a much larger company doing a lot of business and worrying about tens of millions of dollars a year in revenue and it became a big thing. Now remember, I’m a scientist and I’ve suddenly found myself managing a large business with, you know, I don’t have a management degree or anything, that’s not my… So I taught myself along the way, I think I did okay but it’s not my heart. My heart wants to be in the science and meantime, all of this exciting science is coming along. So I’ve been passionately following it all but frustrated that I couldn’t get my hands into it, I couldn’t really be getting on it. And I found that even as a CEO, I needed in my head to have a clear idea of the next 10 years and 20 years to be able to make any decisions at all. So the frustration got to a point where it was like, “Yeah, I’ve just got to drop this job, I can’t do it.” So it took a year of searching and we found somebody to step in as CEO and he’s going great, he started at the beginning of this year. And the immense relief I had was great and I was freed up to, I now have been freed up to just learn again. So I’m learning, learning, learning. And you know, people who know me will know the last three global Moodle conferences that we had and I’ve always give keynotes at our conferences about Moodle, the last three that we had, I didn’t even talk about Moodle. I was just talking about AI to the Moodlers and saying, “Look at all this stuff, look at all this stuff.” So now I’ve actually had time to really dig in, research with the community about what things they want, what things are needed, where AI works, where it doesn’t. And particularly to look at the far future, which is about really the future of education itself, not even just education technology, but what is education for? When do we need it? Why do we need it? What are the things that we’re going to be learning and why? Digging into all of that in depth, and it’s an ongoing process, but I’ve learned so much more in the last seven months, eight months than I’ve learned in the years before that, that I now feel I have a much better handle on where we’re going. And yeah, so that’s, and it’s almost like I’ve gone back to the beginning of Moodle again and I’m starting again because I’m conceptualizing new products that will probably replace Moodle in years to come.

Denzil: That’s fascinating. I love the fact that you just mentioned the jump from an academic space into really this corporate sector, but there’s also the idea of innovation and entrepreneurship to some extent because you’re piloting something completely new and you succeeded at it. And I think that’s a very positive message for a lot of younger people, emerging scholars in the African context as well, is that the idea of thinking of innovation, as you just mentioned, meeting a need and a demand in society and see how we nurture that. But I also like the fact that you went back to research and I think that’s a critical part for me. I’ve had an opportunity to listen to you yesterday at the Moodle Moodle Africa 2024 conference. You spoke something quite interesting in terms of the technological advancements that you’re projecting. Maybe share a little bit about that as well.

Martin: I will. Just to finish on the last, I wanted to add something actually, and the last thing I said was that in building that company, I also was very focused on building it the right way. And the predominant paradigm that we see with companies is that they are focused on profits, they’re focused on money, they have venture capitalists funded, startups get a series A, series B, series C funding. That’s the predominant thing. And almost the product is almost like an afterthought. It’s like producing widgets to make money. Whereas we’re the other way around. Moodle is very much a product, a project, and the whole business is only to provide sustainability for the open development of things. And I had to do things like turn down $20 million US offers over the phone for Moodle and all kinds of venture capitalists trying to get involved and reject all of that and fight for a much better, open, sustainable way to run businesses. So we can talk more about that later if you’re interested. But I just wanted to add that in. That was important to me and I still fight for that kind of mindset.

Denzil: I think that’s an important point because there’s an ethics behind that. And before we get to the emerging technologies and the rest of the stuff, I think for us in the African context as well is, tell us a little bit about that, maybe expand a little bit more, is that that’s quite important because it’s a business model. Yes, you need a business model. And I think that business model is very much rooted in the idea of African ethics as well, is that we need to be community driven as opposed to being capitalistically driven. And in this case, your vision for Moodle sustained its vision objective over the several years. And if I have to ask a lot of people, if you refuse a $20 million investment in your company, not many people would say that.

Martin: And at that time, we were only 10 people. Like it was very small. And still we were threatening larger players, incumbent competitors, and they were calling me up with these kind of offers because it’s this American way of doing things, it’s very venture capitalist driven thing. So it is very philosophical and very deep for me that I see the world as every person is one person. We are all equal. We all equally share this earth. And that business or busyness, being busy, is just activity. But it’s a way of structuring how we can work together to get things done in society. And so for me, the idea that you should structure organisations so that people can contribute to a common good. And everyone gets on board with that, like it’s good. Whereas the other view of thinking is more it’s extractional. It’s like I run this thing and I’m going to target the market to extract value and produce revenue streams to produce massive profits for my shareholders who aren’t in the business, they’re outside the business. They’re just extracting value, right? And it’s that kind of language. And it’s very, you know, I’m sure we’ve seen plenty of it in Africa with the whole colonialist thinking that it’s not good, right? Some people get super rich and a lot of people get nothing. They get used or they get farmed, basically. It’s a very deep and a very important distinction about the way to organise things that I think I’d like, I really fight to see it. I want more people to see it the other way. I think most people I know get it when they hear that. So yeah, if you look around you, and I love the word Ubuntu because it totally jives with that. And it’s a sharing and it’s about we’re all in it together.

Denzil: Exactly. 

Martin: And I think, you know, in my own country and here in Africa, just look at mining, for example, like this is the dirt we all walk on, we all come from. And for some people to say, oh, it’s all mine and become billionaires on it. That just seems wrong, right? It’s a finite resource. It’s been there for millions of years. So yeah, this is the kind of thing I feel like the world would be a better place if there was more infrastructure building in a collaborative way.

Denzil: And I think that’s very important. You know, and I must be honest, when we normally have these types of conversations, Africa is very sceptical towards technology innovation, specifically those coming from outside, because there’s a tendency to take all our resources, exploit us. And I think this type of model that you’re putting on the table really speaks back to the African dynamics, the humanities of what we are putting in terms of technology innovation. And I really like that business model, because it shows that the people that are contributing are not just, you know, kind of victims in the process, but are participants in the process. And there’s an accountability and a value to it. And that fits very much with the African model and thinking of the way we do business in terms of Africa.

Martin: Yeah, I feel very at home here, actually. It’s very, a lot of sympathy. 

Denzil: Oh, no, but we appreciate your time as well in this place. So let’s get back onto the other part, Mark. Now, shifting from the business models, and where do you see us going? And you had this fascinating trajectory. I think that stemmed from your several months of research as well and projecting. It’s a pity we can’t show the graphs that you showed yesterday, but these emerging patterns that emerge, right? Where do you see us going in terms of the next couple years?

Martin: So AI, artificial intelligence, is an innovation that has taken 50 years to develop. But just those last four or five years of the generative AI, that particular nut that has been cracked, is more important than the internet itself. Now, a lot of you will remember when the internet came, and started replacing a few things. Look how dependent we are on it now. You can’t do banking without the internet. You can’t deal with the government. You can’t deal with the school. You probably talk to your family most of the time over the internet. Like it’s part of the fabric of society. AI is going to be even more impactful. Like in the future, there’s going to be billions of AIs, probably more than humans, running around, connected on the internet and connected with us. Some of them are going to be, many of them will be living in the cloud. They’ll be on machines. And many of them will be embodied in personal robots. So there’ll be robots serving you drinks at a bar. They’ll be helping you at the supermarket. There’ll be doing tons of work around the place. And they’re going to be so cheap and so available that they’re going to be, you’ll be buying one for home. And if you’re the kind of family that can afford a second car, you’ll go, well, instead of a second car, let’s get another robot. And that robot will be cooking dinner and cleaning the dishes and doing the laundry and all those things we don’t want to be doing. Instead of buying lots of special purpose tools, these things can use our human tools. So they can pick up a drill and fix something on the wall, or they can pick up a lawn mower and mow the lawn. And they’ll be cheap enough that everyone will have them. So this seems like science fiction, but it is literally around the corner. And there are already, all these things are already happening in labs, and they’re already being developed in factories, but they’re just not around so much yet. But that’s coming. And that is transformational. That is a world where the entire economy is turned upside down.

Denzil: But that’s one of the biggest fears that people have today. I mean, the moment you think of AI, the moment you think of robots, even people who do not understand the impact that AI and robots has, the first thing is, is this going to take my job away? So what do we say to those people?

Martin: You need to be voting for people who are thinking about this stuff, because it’s going to take a complete transformation at a government level in every government. It’s not going to happen tomorrow. It’s going to happen slowly. But you’ll see these robots hitting the streets five years at most, probably less. It’s going to depend upon regulations, because there’s so many things to solve. This is nothing we’ve seen before. I could buy a robot and send it around to your house to rob you, and you won’t even know where that robot came from. So what’s going to happen?

Denzil: That’s an important question. I mean, two things that immediately come into my mind, and we’re shifting a bit here, but I think because of the humanities angle, I would like to bring this in. Firstly, the moment we see robots coming in, there’s a change of how we see society structured. And I’m thinking, for example, you mentioned some critical things here. For example, the roles of being at home, making those gender roles that have also been challenged. And I know some of our feminists say, yes, because we don’t need to do those things anymore, and we can really find those roles. But that means defining those really social community roles that we’ve taken for granted for years. So that’s the first one. The second one, which you just mentioned now, is you can send a robot home and rob a person. But I’m thinking about drone technology, which is already in use. And there’s an entire ethical dilemma behind it as well. When we think of bombing and wars and the rest of the stuff. Don’t you think this is a bit too fast for us, that society is really struggling with these things?

Martin: Yeah, I’m not saying I want it to be fast. I’m saying it will be fast, because there’s a lot of little tiny micro decisions going on that are just going on. I mean, there are a lot of drones around, for example, already. People are buying one for 500 bucks. So look, there’s a few bad things we’re mentioning here. And I’m only drawing attention to them. I am still very optimistic. I still feel like we can have a utopia here. If we do this right, we can actually create a post-scarcity world. And a lot of what is problematic today is that there are limited amounts of things. There’s limited amounts of energy. There’s limited amounts of food and water in many places. Basic survival stuff. Not everybody has a nice house and cars and all those things, because we don’t have enough to go around somehow. But if we’re able to solve all the production issues, AI applied to every single field of science to improve the efficiency of it, to improve the practicality of it, better batteries, better food distribution methods, better way to grow food and make food, better ways to get water around the place. We can solve so many of the world’s problems right now. Then you don’t need to work. Watering was a thing we did to group together into larger entities to get stuff done. A single person can’t distribute water around a city. You have to build a corporation of some kind to maintain all that infrastructure and so forth. So you have to hire a lot of people, because that’s the only intelligences we had. But now we have these other intelligences that can take care of that. So it’s a real mind shift. We’ve totally been in this hive mind kind of idea that we’re all going to do all the work, but now we’ve actually got another species we’ve created, basically, that can do it for us and with us. So there’s a lot to get right, and there’s a lot of changes. It’s literally the biggest thing humanity’s ever faced, but I think we can do it. And I think we can work our way into it. And as for what we’ll do and our roles and our identification with our work, a good example is to look at very rich people. Because people who are right now very rich, they’re already living that lifestyle. They don’t worry about their post-scarcity. Whatever they want, they can get it. They have nannies to look after kids. They have cooks to do the cooking. They have a travel agent to organise their travel. They have drivers. They have all of that stuff. And what are they doing with their time? Well, some of them– let’s talk about the best examples of rich people. Some of them are investing and making changes in society that they want to see. It’s like Bill Gates. That’s a Bill Gates foundation, and he’s trying. Maybe he’s not always successful, but he’s trying to do good things. There’s lots of charity work and so on. There’s a lot of community stuff going on. There’s a lot of– I’m sure they’re all hanging out with their friends, and they’ve got great family relationships, and they’re all doing things like that. So that’s all stuff we could all be doing. And I feel, actually, you have a greater chance to be more human. So less of being a robot working for someone in a job, being told what to do, and more being a person with your own autonomous capability to decide whatever you want to do. And I can see a very lovely society coming out of that, and less angst, because there’s not us versus them anymore. It’s not rich versus poor. There’s less friction, because, hey, we’re all rich now.

Denzil: That’s actually quite fascinating. I’ve always heard the argument on the other side. Technology is going to challenge who we are. But what you are saying now is technology has the ability to bring us back to what is actually the foundation of being human, to having that connection, to not being overworked, burned out, families being destroyed by overworking. Now you see this kind of synergy between technology and human. But the idea is this relationship actually redefines what we mean as a human community, human society.

Martin: Absolutely. And look, it’s not like you won’t be able to– you won’t work. Of course, you’re going to work as much as you want to on whatever you want to. You want to learn to play guitar? Go for it. You want to be an artist and sculpt? You might work all day hammering away at some piece of marble, because you want to. You’re enjoying it. And that’s work. So you’re going to create your own work for your own self-satisfaction. Beyond that, I’m very interested in the philosophy of India, Buddhism, and so on, and this idea of enlightenment. So if you’ve got enough time to really dig into yourself and to be thoughtful about the world and yourself and to achieve this kind of a state, all of those become a bit more possible, perhaps, if you’re not constantly daily thinking about a job. In my dreams. If we have an enlightened species on Earth, that would be fantastic. So this is like the utopia that we could possibly have if we all wanted enough and if we organise it. There is still a lot of profit-seeking behaviour in the world, however. And there’s a lot of control-seeking behaviour in the world. And there’s a lot of people who want that. And we have to push back on that, because that kind of thinking will seek to control the technologies that do the work and continue the landlord mentality that we have now, which is subscriptions, rental of things. You don’t own anything. Actually, there was at the Davos Summit a few years ago. They said, in the future, nobody will own anything, and you’ll be happy. It’s like, yeah, but who owns the things? We’re all renting, are we? We’re all renting to two people who own everything. And the dark side of Bill Gates is, why is he buying up farmland everywhere? Why? I really want to know what the thinking behind some of these people, what they’re doing. So yeah, there is definitely a battle ahead, and I feel like it’s something really worth having. And so if you’re listening to this now, I’m glad you are, because I really want more people thinking about these things and in your own areas, fighting for the good outcome.

Denzil: So it’s really pushing people out of the comfort zone, who are listening to it. Let’s bring the discussion back into education, AI education. AI, augmented reality, virtual reality, extended reality, how do you see it? I mean, you’re in the research lab, right? How do you see it changing? Because obviously, I mean, this is going to transform the AI education environment completely. We cannot be doing education, even in the African context, and what’s our deficits, the same way we’ve been doing it. How do you see this transformation taking place?

Martin: Some ways things won’t change that much, because all of education is structured around how people learn. And we will continue to learn in similar ways. The brain still needs to do the same kind of gymnastics to learn. So you need to work at things. You need to read things. You need to practise. You need to do projects. You need to have feedback loops around your learning. All these things still exist. And whether or not you’re doing it for a job, if you choose to learn something or whether you have to do it to get a job, it’s going to be much the same. So I feel like a lot of that will be similar. Where AI can really help, though, and where my research has been leading me, is AI can help increase the quality of courses. Right now, we’re depending on one person and maybe a learning designer or something to structure a course, to optimise the experience so that you get from start to end and you end up learning things. There’s assessments and gamification and so on. But it’s a one-size-fits-all thing. It’s one course. We’re all going to take the course. And maybe there’s a lot of collaboration along the way and discussions, and that’s great. But it’s one course. And AI can really help make that course better. It can optimise the experience. It’s bigger experiences. It can make the content better. It can help create content, find content. So maybe you made a course this year, but it’s in a fast-moving field like AI. So you want to update that course probably on a monthly basis. So that’s too much for teachers usually to do. So that’s something an AI could actually assist with, is keeping things up to date and making sure the quality’s there. So that’s the first part. The second part is then AI can really help an individual’s experience with the course. So it can be personalising, customising, transforming it as you need. So somebody might be approaching a course, and they’re a full-time student, got all the time in the world. They’ve got a fancy big computer, and great, they’re using it as intended. Online course on the web, they’re interacting, blah, blah, blah. Maybe you’re a single mother, and you have limited time, and you don’t have a computer, you’ve just got your phone. And you don’t speak the language of the course. You speak a local regional language because you’re out somewhere in a village or something. So AI can transform the course and help be an intermediary between you and the course, transforming text into audio perhaps. I showed an example at my keynote of, I had some text, and I turned it into a podcast discussion between two people like we’re doing, and it sounded like a real podcast. And it was enjoyable to listen to, and you could still pick up the information through your headphones while maybe on a bike ride or on a bus or something. So there’s so many possibilities for AI to help transform and customize it. Another really good example is fitting it into your calendar. So you’ve only got a few hours a week you can do this. So when you have that time, the AI might notify you on the phone and say, I can see you’ve got a bit of time now. You’ve got an assignment next week, why don’t you do a bit of work on it here? Here’s what you need to look at, and here’s what you need to do. So tons of potential there for AI. And you notice none of that replaces what the student was doing. So they’re still doing the learning, and it doesn’t replace what the teacher was doing either. It’s just augmenting and assisting teachers and students. And I still like that human connection. I’m not a fan of one giant AI being my tutor, because I just, it’s just not as fun. I like the flavor, the flavor of people. I like that everybody’s got a different flavor.

Denzil: Exactly, that diversity.

Martin: Yeah, yeah, yeah. I mean, it’s again, back to Ubuntu, right? We should be strengthening human connections in the face of an alien species we’ve created, perhaps.

Denzil: Yeah. But let’s get back. I mean, I love the fact that you’re talking about Ubuntu now, bring it back. And I know, from the African context, we’ve always put this, that we firstly, we’re behind technology developments. We’re always running here. I mean, technology development for the West, because of our infrastructure, financial situations, we’re running behind these developments. But we tend to also find ourselves being colonised by some of these developments. How do you see Africa responding to these developments in such a way that we shape the developments as opposed to running behind it? And I know you mentioned something, for example, about the AI impact and how Africa can leverage that, whether it’s building our own AI models, whatever it is, please speak a little bit about that.

Martin: Absolutely. So the American hegemony over things is frustrating to me too. I’m an Australian, you can tell from my accent. And even I feel it. Like Australia is quite aligned with the US in many ways. It used to be with England, but maybe a bit more with the US now. And even I feel it. OpenAI will release some chat, their chat bots with like 10 voices and like nine of them American accents, and they’ve got one English accent. And that’s it, right? Where is everyone else in that picture? They consider themselves a global service. That’s a tiny example, but it goes deeper. It’s in all the training data, all of the materials used in the training and all of the benchmarks that they use to measure their work are based on this particular framework and culture, which Americans have been doing for many years. Look at Hollywood pushing out and their marketing and their fast food and their products, all around the world. And you can’t walk into a mall anywhere in the world without hearing American music on the speakers. That’s what they do. They’re probably the world’s best marketers and they happen to market their own culture and it just is very successful. But that does come at a cost, of course, of local cultures. And what I love about computers in general and AI specifically, is that it is something that everybody can do and contribute. So even if you take a model that’s produced in the US, say some of the open source models like Lama or the many others, you can still fine tune and train it and give it all the extra information it’s missing and produce locally relevant to the local culture services and tools. And it’s not rocket science. It’s all well established techniques. It’s not that expensive. So I really feel like there’s so much opportunity here for everybody to be part of the revolution. Not just sit back and go, “We’re going to use this dot com service,” but take the core technology and build it yourselves over here. It just takes willingness. And everything is, you can learn it on the internet. And the beautiful thing is you have AI now to help you learn anything you need to learn to do it.

Denzil: And also build as well.

Martin: And also what?

Denzil: Also build it as well.

Martin: Yeah, and build it. Yeah, yeah. You can use AI to build this stuff.

Denzil: Exactly. So we’re all empowered. So we need to do it. 

Martin: And I’m very interested to work in Africa on helping, because all these universities are using the Moodle platform, we’re looking at AI in an educational context. Well, let’s work together on making AI language models that are very customised and local.

Denzil: Martin, this is quite important for us, because you’re really saying, “Listen, there’s a developmental agenda, there’s a skills agenda, but we have the potential in Africa to do this.” Totally. How do you see open education technology transforming this space? Because I know that you’re advancing this quite much in your own space. How do you see the transformative impact of open education in Africa, open education technology in Africa?

Martin: Yeah. So I actually started an association called Open EdTech, Open Education Technology. And that’s an association of organisations that are interested in open education technology. It just makes sense for this space. As I said, I have this sort of academic background. Academia and science works because academics publish their work in the open. They do some studies, they publish it. That open sharing of science is what has enabled every development, including AI. It’s not companies making AI particularly. 90% of it was done by publicly funded research that developed all their things, and it’s just someone just commercialised it on top and put it under a dot com. Even with all the dot coms, 90% of all the technology they use is all Linux servers and open internet technology and it’s all open stuff. Open education is especially important. Education is how we transmit culture and knowledge from one generation to the next. Why would we put that at risk of some dot com that can come and go on the whims of an investor? It doesn’t make sense. It’s got to be open. It’s got to be infrastructure that we can trust and has low risk. So we need to always build things that way in my opinion, and that’s why I’m a very strong proponent for it. I can almost not even imagine why you would choose a commercial product to actually build a university. It just seems wrong to me. And honestly, I don’t care if it’s Moodle or something else, as long as it’s open, that’s what matters. Openness is the thing.

Denzil: Yeah. And I know you mentioned this previously in part of your keynote address, I think one of the questions that came up as well. It’s critical for Africa to produce open content. And you mentioned that also to bring about the balance of the AI.

Martin: Correct.

Denzil: Explain a little bit about that. What do we need in Africa? Because one of our challenges in Africa, I think it’s also a mindset shift that we need because we tend to still think about copyright. We still think about intellectual property. But your perspective is quite interesting. You’re saying push the open content, challenge the AI models that we have in the system with data sets.

Martin: Yeah, correct. The AIs are a summarization of us. That’s what they are. The training data for most AI now is just wide scale web scraping that’s going on. They’re just scooping up everything off the internet. You need a lot of data to train AI and get this level of intelligence. There is a movement towards more synthetic data, like constructed data, but the raw data that’s on the internet is still super, super powerful for the training. And if in a web scrape, you pick up like 50% of the internet, roughly, I heard. I would love to, I just heard it recently and I can’t verify that, but it sounds about right. 50% of the content on the internet is in English. There’s also a lot of like Chinese content, which just gets ignored usually. They just go, “Oh, Chinese, we won’t use it.” But there’s a whole world, a whole parallel universe of things written in Chinese characters and Asian languages that don’t count. But English is a huge proportion, so that outweighs, and so with that comes a lot of American, let’s say, frameworks and culture and ideas. We need to balance that. So we need to counterbalance that. And that’s an intentional decision to go, “Hey, we have all these things that describe what we’re doing and what we think and how we think things should work. And our stories and our history and culture, if you hide it away in books or if it’s hidden away behind firewalls, it won’t get into the AI. And then some future student who’s talking to that AI in years to come, the AI won’t know about it, and so it won’t come out. And so it’s going to be biassed. So to counterbalance that bias, we have to be openly publishing.

Denzil: That’s exactly, that’s a problem that we have because, for example, with content that you generate on Africa, you still have much of those bias positions being replicated. And this brings me to the next point. So do you think from an African perspective, should we play the game at a global scale or should we start developing our own AI models and our own AI datasets inside Africa?  Datasets particularly?

Martin: Yeah, I think a lot of stuff is in common, right? The world, the way the universe works, a lot of science is common. Gravity works pretty much the same here in South Africa as it does in Australia. So a lot of that is common, and so that’s not a problem. I’m talking more about human culture. And so the AI models have an inherent kind of understanding of the world that they build from what they read, and that’s pretty common. So you don’t, and building an AI model from scratch is expensive. It does take millions and millions of dollars, and it puts it out of the reach of most things. So I’m not suggesting we go right to the core basics and immediately, you don’t need to, although it would be nice if there were more people making foundation models. The good thing is that once you have a foundation model, you can do that extra training. So it’s almost like an education program. You’re actually educating that AI on things. Sending it off to school, “Hey, read this, read this, read this, read this, read this, learn this, learn this, learn this.” And you’re just getting it to learn all these other things. And so then you produce a different kind of entity that can act in different ways.

Denzil: So basically what you’re saying is we’ve got this infrastructure that’s set out. Countries may not have the cost to bring their own really foundational models. So let’s get there and dominate the data sets that we’ve got.

Martin: Get more data sets, more variations.

Denzil: The data.

Martin: Exactly. Like, you can’t complain about colonialism if you don’t have a clear idea of what’s being colonized. Let’s work out what it is we want to protect. Let’s define that. Let’s write it out. What is it?

Denzil: Exactly. So in the Australian context, in terms of indigenous communities, indigenous language, has there been any intentional attempt to rectify that in an AI perspective?

Martin: This is an area I’m very interested in, but I haven’t looked into too much yet. I’ve just started today! Over lunch, I was starting to do a bit of research. So yeah, not much that I know. I know in Nigeria, there was a project started in April, I believe. And they’re intentionally training up, I think using the Llama model as a base, and they’re fine tuning it on five different local languages. And there’s several other projects that I haven’t looked into yet, so I can’t really give details. But I’d love to see that stuff happening. I want to support that. Where I come from, I grew up in an Aboriginal community, and the Aboriginal languages are very marginalised in Australia. And the population of Aboriginals in Australia is like 200,000 to something like 25 million total. So it’s a very small population. And so they just don’t appear in the media very much. So there’s a real danger of all that culture disappearing. So it’d be nice to see more saving of those things. The human ecosystem, we shouldn’t throw away anything, I don’t think. I like to save things.

Denzil: Exactly. And my last question, Martin, this particular podcast really focuses on the African digital humanities. So when you think of preservation of knowledge, preservation of culture, and we had this conversation, and digital storytelling, how do you think, from your perspective, what kind of technology should we start rethinking about in the African context? And we start thinking that if we do not digitise some of these traditions, they’re going to be lost to the next generation. And the idea of the African digital humanities is really curation, archiving, preserving those voices. So what kind of technology? I know you’ve got this fascinating technology that you’ve also mentioned, and this is the last one. How should humanities scholars, social sciences scholars, approach the technology space with a specific agenda of rethinking preserving African digital humanities?

Martin: Well, look, I’m sure a lot of them, and a lot of the audience, are probably writing research papers and sharing them widely. And that is excellent. I mean, the scientific method, still the best way we have to try and nail truth down, the review process, the whole rigour around that is pretty well established. But it’s being attacked from every side. A huge proportion of papers now being submitted to journals are written by AI, and they contain mistakes. So we’re actually polluting the scientific literature with errors and inaccuracies and so on. And not to say that everything ever written by every academic was true and accurate either, right? We all make mistakes too. But with duplication and replication and multiple people working on the same things, you tend to have ways to increase your confidence in the truth. So keep doing that. But I would say whenever you publish papers and it’s to some journal and they want to control the result, and they do, like all these large publishers, there’s walls they’re putting around academia, fight back against it. Always get permission. And you just need to ask, right? A lot of people don’t know this. You just need to say, “No, I want to publish this on my own website as well.” Or, “I want to be able to publish it wherever I want.” I published some papers and they said, “No, you can’t.” And I said, “Yeah, I need that.” And I got it. So push back. And the reason is because then you can put it into open education places where it’s freely accessible. It’s not behind firewalls. And the more open it is, the more exposed to search engines and AIs it is, the better. The more it’ll get into things. The second thing is, apart from that, just putting it out there on the web, right? Anyway, any way you can, right? Just get it out there. Because I saw a stat, it was something like 40% of new content is now AI generated. Again, with inaccuracies. Just on the open web. Just on… I mean, have you looked at Twitter lately? How much trash? Just a sort of trash flow, right? There’s no sense of truth there. That stuff is what Elon Musk is using to train his AIs. He’s going to use Twitter. Christ. That’s the wrong thing to do. But we need to counteract that with as much good data as we can. So even if it’s not an official paper, if it’s just a preprint, get it into the archive. That’s the service for publishing preprints. So you don’t need peer reviews. Even if you don’t have that full scientific rigour around it, I have faith that AI will help us sort through all of the noise and still be able to extract truths from it. But we can’t look at it and do those things unless it’s available. So you just got to get it out there. So be open. As open as you can. Publish freely. Keep learning. Use references. And also if you are using AI, make sure you identify you’re using AI. Because that helps us judge quality. Those are the basics.

Denzil: Thanks, man. This has been an excellent session. I mean, you’ve touched several different dynamics as we went through. And I think what you’ve really done is pushed the thinking that this is not something that you can put in a box. I think it’s gone far beyond us boxing this up, this conversation. It’s really for us to think outside of the box, apply our own minds, and also the fact that this environment is progressive. It’s changing all the time. And if we are static in our thinking, we’re going to become irrelevant.

Martin: Correct. We’re all connected.

Denzil: We are all connected.

Martin: And it’s a fallacy to be too precious about your words and your writings. Because to say, “oh, I wrote that. It’s mine”. No, it isn’t. You are part of the culture. Everything we’re writing and saying is part of what you have absorbed as an intelligence around you and you have summarised or you’re reacting to. So we’re all in this together. We are a kind of hive mind. So we need to push more good into that ecosystem. So that’s my main message, I think.

Denzil: No, powerful. We end on a very beautiful note. Knowledge creation in the world of AI, if we really want to shift the African discourse, AI needs to partner with us to get the agenda. Thank you so much, Martin. We appreciate this. I mean, it’s always a pleasure hearing you as well as innovative ways that you are thinking. And it’s also seeing where you are going as an individual that will shape education on a global scale as well. Thank you.

Martin: So we appreciate that. So thank you once again.

Denzil: Appreciate it too. Thank you so much for the opportunity and the space.

Martin: Thanks.

Denzil: Thank you for joining us on this episode of the Ubuntu Nexus podcast series, Bridging African Epistemology and Technology. We hope today’s discussion has provided valuable insights into the transformative potential of emerging technologies. Stay tuned for more enlightening conversations with scholars, technologists, and cultural custodians who are driving the movement for epistemic justice and transformation. Until next time, let’s keep the dialogue alive and strive towards a world where all forms of knowledge are respected and integrated into our collective understanding. Thank you for listening to the Ubuntu Nexus podcast series. I’m Denzil Chetty, and we look forward to reconnecting with you in the next episode.