Video transcript
] Hi, everyone. Great to be here. Great to see you all. So I hope to broaden the conversation a little bit about tech and learning and how those things weave together. We're thinking about the future and tech and what future our kids are moving into, you can't really do that in isolation. You have to stop and take a step back and just think about what's happening in the world around us right now. So you have mass population movements. You've got the disparity of wealth. You've got that dominance of digital and data happening everywhere. Climate crisis and protests around that and global mobilisations, you've got the anonymous hacker movement. You've got a mass increase in pollution. You've got of connectivity and data flowing around the world at a pace that it hasn't before. Both for good and for bad. So we've we've got these seismic changes appearing all around us right now, which is scary, dark, exciting, rapid evolution in front of our eyes. So it's definitely a time of fairly seismic change. And I guess that the challenge for all of us as parents, as teachers, as educators, is to try and navigate this changing world. And we're not quite sure exactly where the destination is, and to help our kids prepare themselves so that they are in a better place when that future unrolls for them, but we're not quite fully understanding it ourselves, so it is a fairly fluid moment. So as a tech person, it's also quite exciting because it's an open world to try and make some changes or to try and improve things for the better. But it really is quite fluid. I just wanted to start with that fluidity. And there's some pretty powerful forces as well. So the one about digitisation, is just open access. I'm from South Africa. Education in South Africa is very poorly funded. There's a massive imbalance between super poor rural schools and some wealthier urban city schools. But right now, any child in any school with Internet connectivity can get access to vast realms of really good quality resources, lectures from eminent professors, Wikipedia, open education resources, open source movements, there's a ton of movements which are freeing up data and information in ways that wasn't available before. Obviously, this isn't learning. This is the information that can make learning happen. But this is a seismic change. And this is just down to digital. Another one that I'm particularly invested in at our work at Babbel is just scale, the ability of creating something that can be consumed by millions. So you can have 10 fantastic educators working together, building a new digital product, Rose was describing a whole bunch of digital products there. And if they get it right, if they get the right combination of learning and tech and desire to learn, it can scale to millions and millions of people, it can influence millions of learners, right, if it's a digital thing, it doesn't matter if it's one or ten or a thousand or ten thousand that there's lots of teams trying to get it right. So not everybody does get it right. But it's the promise of being able to be available in infinite multiples, which is a real, real alert. I'm from Babbel, so what we do is we have a language learning app that we have millions of people using, in a way that's the world we inhabit, right? So we would never claim that the experience of learning a language with an app is better than a one on one session with a teacher. But what we can say is that we can get in front of millions of people and help them all just a little bit. And so this is the area that we operate in heavily. So I want to pick a number and think a bit about that number. 20 million. Think of 20 million students just as a number. So in the UK, at the moment, in universities across the UK, there's probably two and a half million, 2.2 million, something like that. So it's nine times the entire university population across all of the UK. Twenty million. It's quite a reasonable number of students, right? So last year, 20 million students registered to do courses on MOOCs. So probably most of them are not existing H.E. students, they're probably private individuals, who from their own desire decided they wanted to learn something. That's 20 million signing up for some free, probably entirely voluntary, opt-in learning. That's quite a big number, right? Of course, the thing with MOOCs is you don't know how many of them went all the way through and how many of them spent the time. But let's just hang onto that thought for a second. In the last two weeks, 20 million people spent an average of an hour and a half online watching other people playing video games to learn how to play them better. Two weeks. So the particular platform I pick the data from is called Twitch, which you may or may not know, it's the dominant platform for people watching other people playing video games. It literally looks like this. So I have the screen that this guy's playing. He's there, he's talking to me while he's playing the screen. There's a bunch of live channels. It's like live YouTube. But seriously, the average number of the average amount of time people spent is ninety five minutes per day. Ninety five minutes per day. Students, people interested in learning, interested enough to spend that much time, right? Twenty million. Two weeks. So I guess my main point is that digital is there, people are learning digitally. It's happening right now. It's just not happening in school. I just want to start from there. It's not the same. It's different. It's anarchic. It's sometimes, it's for a variety of different drivers. But there's stuff happening right now out there which I think we would all benefit from just trying to get a bit of a better grasp on and trying to understand how to deploy some of this for good, and what the hurdles are and how we can make use of it. So a quick intro to me. My name's Geoff Stead. I'm a chief product officer at a company called Babbel, we're a language learning app. So we are in this digital space. All learners are dispersed around the globe. We are fairly large in terms of a startup. About 750 people, most of us in Berlin, but 50 nationalities. So we're quite a big eclectic mix. And we all have one single product. All of those people work on one product, which is a tool that helps people learn languages. And it's a language learning app, 14 languages, subscription model. I'm not going to talk too much about it, but this is just to contextualise. So I'm coming from the purely digital end of the spectrum, a digital product, people learning mostly by themselves sometimes and sometimes in groups and just using tech as a way to support that. So what I'm hoping to do today is give you a couple of suggestions or a couple of lenses to think about emerging tech and think about what's coming in the future and try to think about how that might be useful to take back into your schools or take back into your own thinking. You'll hear overlaps between what I'm speaking about and both the two sessions we heard before. And really, I just want to cover these four areas, which is my reflections as somebody totally immersed in digital, where that intersect is lying between digital and tech and digital education. So the first is Saul's story about triangles. So I'm particularly enthusiastic about trying out new things and trying out new tech and trying out new ways of learning with new tech and how these intersect. And I've been quite frustrated at how slow the mainstream parts of education are to experiment and be a bit more wild and loose with these emerging techs. And it feels a little bit like one of these Mexican standoffs, I think they call it, whatever it is, pointing a gun at everybody else. And it's not that nobody wants to innovate. In fact, everybody does. But they're held back by the other two who are pointing guns at them. And so if it feels like maybe being overly provocative, yes, I'm happy to hear if this is different in your worlds, but it feels like there's a desire to innovate, but that each step of the innovation is held back by another step. You know, is it the curriculum that we are constrained by that doesn't let us be too loose? Is it the assessment? Is it the publishing resources we use? Is it the frameworks that we ask to operate under by governments? And everybody feels that it's somebody else that needs to change first to allow them to be slightly more open ended or or creative in how they operate. And so there's nothing innately wrong in this situation because it's got there through very carefully thought through steps. But it just makes it a very difficult environment to do more exploratory digital stuff, which is why labs like Rose's exist, because people step outside the mainstream and try and do the more experimental stuff there. It's why a lot of tech companies get pretty excited about what's possible, but actually really struggle to embed it into the mainstream. So I guess it's a trap that we're all in. And I think you need to recognise that trap to be able to understand how you deal with it. A framework I particularly like when I think about how we take tech, can we move it into education, is printed there. And I don't know if you've come across this or not, but I really like it. And the idea is there are different ways of bringing technology into education. And at the very beginning, it's a substitution. So you're doing what you did before, but maybe a little bit faster or a little bit easier. So this might be when you're doing homework and you email the homework back to the teacher instead of physically bringing it back in, or there's a class register and you're using digital for your class register, which just makes it all a bit quicker and simpler, but you're registering in exactly the same way as you did before. So this is the substitution idea and then the augmentation is where it starts improving a little bit. So I lived in California for a while. My sons, the teachers at my son's school would have them do homework, take a photograph of it immediately and send it to the teacher. And the teacher would have it before the lesson started. And the lesson would be based on the homework that had come in that evening from the evening before. So it's changing the dynamic a little bit, not too dramatic. But so this is the augmenting. I think a lot of the tech that I see in schools are in these two areas, the substituting and the augmenting. And that's not bad. I mean, this isn't a bad to better thing, you know. It's not a progression that's getting better and better, but it's helpful to recognise what you're trying to do with it. And then when you drift up into the transformation stage, that's where the technology actually changes the kinds of tasks that are possible and the kinds of approach you have to the extent that really you're creating a task that wouldn't have been possible pre-digital. And you're building skills in a way that wouldn't have been possible pre-digital. So I would probably argue that redefinition: you're not sitting in a classroom at all, students are probably creating something, using technology to do more probing, constructivist type of learning approach, perhaps across several different subject domains at the same time. So it really is a flow. And I think the helpful thing, what I find helpful, is also to be quite explicit about where you are on that flow or what you're trying to do when you're thinking about new tech. Quite often I have conversations about tech and innovation and new ideas. But the people who I'm talking to actually are quite clear that they want to be in this enhancement area, but the kind of technologies they're talking about, or their ambitions are in this transformation area. And I think just being clear about what we're trying to do is a really good way of helping us succeed and move forward. So I guess this area here is more build and make and create, for me. Often, much like for Rose's Lab, the places I come across this are outside the mainstream. It's around the edges of the mainstream, in a couple of places. It can be a more informal learning space, maybe work based learning where you have a different set of constraints. In our area, the business to consumer direct learning that happens because consumers might be more up for paying for some new clever gadgets, even though it's slightly loosened, slightly emerging still. So language learning, our area, health, career, general interests. These are the places where you see that. The stuff on the left is really nice Royal Mail training, VR training, which significantly decreased the amount of dog bites that postman had. So there's, I forget the number, but there's a significant number of postmen and women who still get bitten as they do their rounds. And this was a whole series of role-play scenarios which made a significant decrease to the number of postmen who were bitten on their rounds. This is Liulishuo, they're a Chinese company with some fantastic language processing tech and they have some some really great language learning, practicing apps, particularly this one, an IELTS preparation, where you speak to it, a long chunk of speech, it totally transcribes everything and it highlights phrases that you've got wrong or gives you suggestions about what you could do next. So really clever use of speaking tech. So Babbel, this isn't really a product, this is more playful. So we hosted an event a month or so ago about artists using AI and using technology to create art. So we had artists creating artworks, creating music albums, releasing albums that were all computer generated and they were trying to discuss how they do that, how they train a computer to be able to create something that counts as musical creativity. And we had a whole debate about what that looked like. These are all at the edges of what would be considered mainstream, but are fascinating uses of how tech can be interacting with our lives slightly differently. So I guess we are drifting into the use tech to make things area. Rose I loved your slide, you had a slide of the students in your lab tagged up on things. So this is a slide from one of my colleagues, it's a college project where they were trying to build some artwork. But really, a lot of teachers get this, but the wider education institution doesn't always, EdTech certainly doesn't. There's a mistaken idea that the way to help education is about distribution of content. It's about giving you the right thing, and sometimes it can be helpful. But it's very focused on distributing content, whereas in actual fact most of the more enlightening things that happen in any learning is when you yourself make a discovery. You figure out how to solve a problem. You figure out how to build something. You're stuck and you struggle and then you can rise beyond that. And that's not directly about content being distributed to you. It's about trying to build or make things. And so this using digital to make things is a great area. I was just thinking of some of the tech that's flying our way at the moment. We've got the VR, we've got the 360 videos, we've got the drones, we've got Raspberry PIs and people hacking together new computers to solve new kinds of problems made with all DIY pieces. You've got the 3D printing area, the rise in voice interfaces, not just these devices, but the many ways you can talk to things. So what do you do with this stuff as a teacher, right? Should you be ignoring it, should you be trying to weave it into your teaching, even though it's a bit grubby and informal? Should you be giving precise instructions how to use it? Well in actual fact I don't fully know how to use it. Both my kids have just got jobs for the first time. One's just finished school, one's just finished university in these last couple of weeks. Both of them are doing jobs that didn't exist five years ago. Both of them are to do with the AI area. In fact, one is working on training drones to know where to land for Amazon delivery and the other is working on understanding crowd movements and teaching cameras to understand what they're looking at. This is right now and they didn't study that at school, so I guess my challenge is to try and think of these texts, not as a consumer, not as a thing that you will go and sit and watch a great movie on, but a thing that you can encourage your kids or your students to build something with. What could they build if they had one of these? What could they create? What film could they create if they had some voice interfaces? Could they build their own skills? It's actually really easy to build your own skills for Amazon, Alexa, or for one of these. It's a fairly simple way of starting, but why not get them building or creating? So I guess my call out here is to be really enthusiastic about the creative side and the skill building side. And luckily, teachers know this. So there's a great woman called Jane Hart who every year compiles a list of the different tech that's used most widely across education as globally as she can. In fact, I think tomorrow her 2019 survey comes out. But this is the 2018 survey. And this is what are the top 100 tools for education, according to teachers. And the fascinating thing is they're almost all tools that you do things with, right? There's very few which are passive, observe this content stuff. They're almost all used to create things. You know, I'm building, I'm looking things up, I'm putting videos up there, I'm creating slides. I'm creating quizzes that my students use and I'm recording stuff that's going on my screen. There's a few, there's Ted, and Udutu and there's a few which are about using content, but the majority are about creating and building. And in fact, as a software person, many of these are ones that my teams used to write. These are also the tools of work. So it's a nice reminder that actually teachers on the whole do get this and are seeing that these are powerful tools to use to support learning and education. It's the tools of doing and making as opposed to the tools of more passively consuming. I don't know if you know Alan November. He runs a big event in the US annually. He's the guru of workplace learning. I liked his quote. "We should be designing assignments that students can't Google". And this is the challenge for us, right? It's not about repeating information that you can find. That's like a bit of a fail. It's about helping you solve problems that even I haven't understood yet. It's helping my kids get these jobs that I didn't even fully understand what they were doing, and I'm in tech. So does that make sense? All right, cool. So next one: learner centricity. So the awkward dilemma that people like myself face when we have a whole business, 750 people's salary, the rent of that enormous building we're in is entirely dependent on learners wanting to learn the stuff that we've given them, right? We're a subscription model. So if you're paying your nine euros every month, our lights are on. If everybody stops paying the nine euros every month, you know, we should go home. So we're painfully aware that actually helping your learners want to be there is critical. If you're not helping them want to be there, the whole thing falls apart. So digital only really works if they want to be there. It's different in a class because sometimes you don't want to be there and you come in anyway. And a charming teacher can help turn that into a wants to be there. But when we are only touching them remotely, digitally, it's really tough. Luckily, there's a ton of tricks or approaches or methodologies which you can use to help here. Sometimes you hear a lot of the negative hype about Facebook or a lot of the social media sites about trying to draw you in and click you in and keep you locked in and keep you spending more time in there than you want to spend. And there's reasonably solid science and a bit of mythology around the ways to do that, it's fairly well established. So the trick is to recognise that and understand: do you use it for good? Do you avoid it? What's the way to approach it? So I want to give you some Babbel perspectives here. So we have these billions of users. They spread out all over the world. We don't really have any way of really knowing what they really want and what they are really doing at that moment and why they are really there. You know, we ask them and some tell us, and we have a bunch of different methods to probe and poke and try to understand from this amorphous mass of very different people at different ages, learning different languages, why they're there and how we could help them be there. So we do a few things and we do some of the classic AB tests. So here we would have two versions of almost the same screen and we'd give it to half and half and understand who uses one more than the other to try and get an understanding. We try to track people's usage and we cluster people into different behaviours and then we try these experiments with the different behaviours, the fanatic regulars and the occasional drop-ins. And we try and do that, so this would be a way of understanding. We ask them. So if you're coming into one of our screens, you might find one of these customer service type pop up things that appear. But actually if you answer that, there'll be somebody on on the other side who can have a little bit of a chat with you to understand. We have a wish board, so we encourage anybody who's using Babbel to go and post a suggestion of something they think should be better. So we're collecting all sorts of different data points from these millions of people who we really would otherwise not see. And some of them are quite fun, right? So in some of those points, something pops up and says, hey, you wanna talk to somebody at Babbel? And if you say yes, literally this is what we get. We have a UX researcher who's chatting to you live. So this is a live person, I just covered his face so you can't see who it is. This is a live user, who happened to be using the app and responded to a pop-up, being recorded and talking to our researcher explaining what they're doing at the moment. So we can push new features at them and understand what they think or they can use the product and talk to us about things. And then we get these video recordings. I get a recording of this which lets me hear what they said, the conversation, what they clicked. So these are all touch points for us to try and understand, because the reality is, none of us really know. We try to guess. We try to understand what this large, amorphous group of learners is actually thinking and doing. So that's one side, we try and collect this information, the other side is the murky science and art that I mentioned a moment ago. There's a bunch of different theories and thinkers, some coming from the learning side, some coming from the behavioural economics side like Thaler with the nudge theory, approaches to learning, and flow state. And what organisations like ours try and do is we try and stitch these together and construct something that works for us to help us keep our learners. It's a mix, some of it is fairly scientific, some of it drifts into hyperbolic arts. It's a bit of a blend between those things. But within those blends, you really do see some patterns. So we closely track analytics for what our learners are doing. We try and understand what their motivations are. And we then try and nudge some of them to come back a bit more often because we see that a chunk of our learners use it regularly, a chunk of our learners don't use it very often, but really want to learn. And we know that if you come in only once a month, you're really not going to learn anything, whereas if you are coming a couple of times a week you will. So we try and find ways to steer them back in. If you are new to all of this stuff and want one fairly easy to read summary book, I would recommend this one, 'Make it stick'. It's a high level summary of different approaches that help people engage more deeply with learning. The others are diving slightly more deeper: Nir Eyal, I spoke about Thaler, Fogg is about changing habits, the whole flow piece as well. So spaced repetition, trying to get intrinsic motivation for your learning and not extrinsic. So you're not over rewarding people with stars as you go, it's actually trying to get them to love the learning itself. But it's a mix of different things. And it is really a science and art hybrid. But if you're interested in this, this is certainly the underpinning foundations of a lot of successful digital products, learning and not learning. And so I guess the challenge we tried to take on board is how can we extract the good bits out of these and not get swamped by some of the more negative bits and try to understand how to use our learner's time for good. Does that makes sense? Cool, all right. So my aim here was to try and use these three ideas, learner centricity, tech as a tool to create and build, and innovation around the edges as a bit of a frame that we can think about for emerging tech. We think about how we could bring tech into schools, we think about how to deal with the fourth industrial age, the AI piece flooding over the top of us, just a lens to make sense of that, the digital ethics that Rose mentioned. So I'll do that in one or two examples and then we can take it back to all of us and try and tease it out a bit further. So in Babbel's case, we enthusiastically try and use AI, but we do it in fairly narrow lenses. So Rose was talking about general AI and then the narrow AI, and we use narrow AI. We're very clear about the problem we're trying to solve, we've got oceans of data, and how we can use that data to help us solve that problem. So it's I guess fairly classical, we try to be a bit more personalised, we try to recommend the thing for you that's most useful based on what other people have done and what you've done and how you learn. We use machine learning to try and help us build these profiles of different types of learners. We use the behavioural nudge side. We use quite a lot of natural language processing and computational linguistics because we try and understand what people have typed in so that we can give them a more helpful response. We do speech recognition as well. That's a key part of processing what people have said. So I guess the distinction here is the broad versus the narrow. And we're trying to be very learner-centric here. We're really trying to say: what is a learner's problem in learning a language that we can use AI to fix rather than trying to be broad and say, hey, let's do everything. So an example I particularly like is feedback. Computers are very quick to make decisions and can make them multiple times, but obviously they're not always right and they don't have a very good feedback mechanism if they've got it right. So typically if you're answering a question, multiple choice, voice record, whatever, it's wrong or it's right. It's fairly simple. Whereas if you think of a teacher or a parent teaching language, it's not black and white. You can be absolutely right, you can be absolutely wrong, but there's a big grey area in the middle. So if I'm really only a beginner learner and I get more or less the right words, actually you do better to encourage them and help them keep going than to correct them. And we're trying to understand, how do we get that grey area into the computer? So luckily, we have millions of examples of people who've written typed answers or voice answers to previous prompts in the product. And we can look at all of those and we can see ah actually a third of these, we marked them as wrong, but they are right-ish. Maybe it's a typo, or it's a minor thing that at this level doesn't matter. So we introduced a secondary tier. So in this flow here, she's simulating a conversation and they made a mistake, but it wasn't such a bad mistake. I understood what you meant. It would be better to say this one or this one. So the idea here is how to capture that greyer area that's actually a bit more human. So this is just one example, which I thought might help contextualise it a bit. So really the framing I wanted to leave you with as we went in to talk more widely about tech and learning and schools was first of all, be totally clear why and what is the problem you are trying to to solve? Which tech? Why do you want it? So that whole SAMR model, really what are you trying to do here? And be upfront about that, and accept it. The second was don't fall over-in-love with the consumer parts of the new tech that's emerging, because there's a lot of industry that spends a lot of money on helping you fall in love with that, but really try to think about how you and how your students can use that to make or create rather than just receive. That's a really nice mind shift. And the third one is if you are doing digital things, you need to be a zealot at helping keep the learner in the middle of it. Whether it's about design or engaging, whether it's about really understanding what their problems are. In the digital world, learners only engage if you can really make a real connection with them or they can feel the need to do it. So part of your role as a curator of that experience is to really help them, the learners, be right at the middle of it. So this was the framing, and I'm hoping that's a useful way of talking more broadly. We were talking about what's the future of digital in school with these ninety five minutes on twitch alone every day. Are you students just heading off on a tangent and is school becoming increasingly irrelevant? Is there a chase for relevance in their attention? We describe the attention economy quite often in the app world, people have a limited amount of attention to give and there's some powerful forces - I don't think Babbel are powerful, but we try to be - in the digital world all trying to grab those learners attention. So how do you react to that? Do you fight it? Do you absorb it? Do you take on board some of it yourself? So those are my three framings. And really, that was it for now.