AI Just Got Really Good at Transcribing
Like, really good. It took a forty minute stream of consciousness introduction to my revised book and turned it into the below. With almost no editing.
The below transcript has almost no editing from me. The only thing I have changed is when I began a sentence then started again. Everything else is straight out of the box.
I recorded this introduction (not quite complete) into my phone on my walk this morning, sent it to my laptop, uploaded it to Google AI Studio, and gave it this prompt: “I want you to transcribe this entire audio, word for word. Do not miss anything out or summarise. Do not include timestamps”.
The first attempt failed because it took the lazy way out and summarised. The second time failed because it added timestamps. But, third time was a charm.
It’s less about the content, which will almost certainly be further refined. It is more about how effortlessly generative AI can listen, understand intonation, pauses and such forth, and make an almost perfect written facsimile.
The opportunities in education are enormous. I’ll explore those in more detail in future posts. But for now, my (not so garbled) thoughts on the intro to my redrafted book:
This book is probably unlike any book that you've read on education or school leadership. This book does not have long-winded case studies, nor does it have large numbers of citations of other educational theorists and at no time do I put any surname of an educational theorist and the date of publication in brackets after a quote. Yes, of course I refer to other educators, people I work with, people I read, people I like, I refer to writers on leadership and also writers on technology more broadly. But this is me talking to you about where we currently are in our education system, and what we need to do to move forward.
What this book also does not do, is spend chapters talking about the history of education. We know where we are and we broadly know why we are here. We know that we've had the industrial model of education for about the last 200 years and that probably from for at least the last 50 years, that has not been fit for purpose, but I'm not going to bore you with long-winded introductions about the different acts and blah blah blah because I don't think that's useful and I, I frankly don't think it's important.
What is important is that we work out how to move forward collectively in our school system now that we have artificial intelligence landed firmly on our doorstep and no one really seems to know what to do with it. So this is a book for anyone, whether you are a teacher, whether you are beginning your management career, whether you are a school principal, group manager of schools or at the school system level, to take you step by step through what you need to think about right now in order to take your school into an AI augmented future.
The other slightly unusual thing about this book is that what you're reading began its life as me speaking into my phone on my morning walk. I like to walk every morning, to clear my head, my thoughts, and set me up for the day. And I thought well, why not record the book whilst on my morning walk, so that's what I'm doing. So if you're reading these words, they began their life spoken, and if you're listening to the audiobook, then quite bizarrely they began as I've spoken, they turned into writing and now they're being spoken again. So it's gone full circle. Mind you, you wouldn't have wanted to have read the original draft because quite frankly, most of it was garbled.
The reason I wanted to share this with you from the beginning is that I used artificial intelligence to take the transcript and very quickly and neatly turn it into the written word. This is what we can do now very easily with AI. We've been able to do it for a while now, but it's getting more and more intelligent. For example, it listens to my tone of voice, where I pause, and it puts in the punctuation. This is different from original dictation software and this is the beauty of generative artificial intelligence, which is what this book mainly focuses on because it understands language to such a high degree that it's able to work out to a fairly high degree of accuracy exactly what needs to go where.
The other reason I wanted to speak the words initially, is because actually I want this book to feel like I'm talking to you. I want this book to feel like a conversation, albeit one-sided. You see, I spend my time in my career and I have done for about the last ten years, working with schools and groups of schools to help them improve. I've built schools, or rather been involved in the design of the build of schools, not actually physically built the schools myself. I've bought schools, I've managed, led, taught, you name it, I've done it. And I've used that experience over about the last ten years to work closely with schools and groups of schools to help them improve. And now that's what I'm doing with the support of artificial intelligence, because one of the beauties of AI is that it helps you do things faster. You can get through so many of the early stages of a school improvement process far more quickly using AI. I think this is very, very powerful. And I'll come on to that later.
But I wanted this book itself to really feel more like I was just talking you through my process, really just having a chat with you. Hence not going too big into theories or long-winded case studies or boring histories of our education system. I'll leave that to the academics.
This book is about how we craft a strategy for our schools that embraces AI. Why haven't I written about this before? After all I've been involved in education technology for almost 15 years. I led the first sixth form college in the UK to move to an iPad one-to-one model, worked for Apple, supporting their professional development program for many years, and have always been interested in the disruptive potential of technology to liberate us from the current doldrums that we find ourselves in, and breathe new life into our entire education system.
The reason I haven't done it today, is because I've been profoundly disappointed by every single piece of technology which has found its way into our schools. Interactive whiteboards, virtual learning environments, iPads that I mentioned, chrome books, none of them have made a blind bit of difference to the way in which we set up our schools, our curriculum, our timetable, nor our pedagogy.
I remember back in 2010 or 2011, a guy called Ruben Puentedura came up with the SAMR model. The model essentially explained how education technology would enable us to move from substituting one thing for another, for example, substituting a piece of paper or a paper book for its digital version, digital paper or ebook, right the way through to how technology would redefine learning, as the R at the other end, and how it would enable an entirely new paradigm of learning to be unleashed.
I used the word paradigm deliberately there, because we talked a lot about shifting paradigms, and in fact, Ken Robinson back way back in 2008, recorded his seminal, changing educational paradigms Royal Society lecture, which became the benchmark for us all as we sought to do our do our bit to change these paradigms within our small sphere of influence. Of course, nothing happened. It was a complete failure. Our system is no different now, 2024 than it was in 2008, 2010, so on and so on. So why is that? So I did say I wouldn't go too much into the history of education, but I think just to kind of set the scene, it's probably worth spending five minutes just trying to work out why on earth we've had no discernable change in our education system at all in the last 200 years and certainly within the last 50 years.
Back in 1974, so 50 years ago this year, Alvin Toffler wrote Learning for Tomorrow. And in that book, Toffler begins by giving us a really powerful analogy. Toffler gets us to imagine that we are part of a tribe who live on a river and get our livelihood from the river itself. Our whole economy is centered around the river.
And he he gets us to consider what that means in terms of how our young are educated, that we as a river-dwelling people would educate our children to understand how the river works, to understand how to get food from the river, how to use the river for navigation, that it becomes part of the map and the territory within which we raise our young. But what if 10 kilometers up the river a dam is being built? Now, as a river-dwelling people, we don't know this. We have no idea. Our world is very narrow. It's just the environment, maybe a few kilometers radius from where we live, but no further than that.
But the river begins to dry up. Slowly but surely our livelihood is removed, destroyed, but we've raised our young to only know how to make a livelihood from the river. And we have no way, no model, no different approach to how we educate.
So what do we do? This is where we find ourselves in our education system. We have an entire system which is predicated on the industrial model of education. From how we train our teachers, through to how we structure our examination system and then everything sits in… that sits in between. And we have no idea how to change, because we only know one way. Like only knowing how to gain a living from the river.
So, yes you can bring in a shiny iPad, or you can herald the brave new dawn of interactive whiteboards, where we are liberated from our dry white pens or our chalk, but that does nothing at all to the system itself. That's just bringing in one other tool to prop up the existing system. A classic example is moving from a chalkboard to an interactive whiteboard to an iPad. Actually, it's the same thing, only in a different form. You've moved from a black rectangle with white writing, to a white rectangle with digital writing, to a little shiny rectangle that rests in the student's hands. It's still a rectangle within which you are passively delivering content. You could, I suppose, make the argument that with an iPad there could be a degree of interactivity, but let's face it, most schools have not even begun to consider how that might look. They might create resources that the student can access via the iPad, but it's still passively delivered.
So, the reason none of these technologies has made the slightest bit of difference and the reason why we still have teachers standing at the front of the classroom, talking at students is because our entire system has been underpinned by tools that reinforce a one-way, passive mode of delivery with the teacher as the expert and the student as the passive receiver. And whether it's scratching chalk on a blackboard, or uploading worksheets to a VLE makes no difference whatsoever. The teacher is the expert and the student is the receiver.
But there are several reasons why I believe we are on the cusp of a massive shift. I hate to use the term paradigm shift, because it became so cliched in the early 2010s and late 2000s, but it feels like it might be time for a renaissance.
The reason that we are on this precipice, tipping point, whatever you want to call it, is AI learns. That's actually it, AI learns. And AI is learning. And AI is getting more and more efficient, more and more proficient, it's getting faster and it's getting better at doing more and more of the things that we would normally do. An iPad is a rectangle of glass; it will do nothing unless you interact with it. It's inert, it's dead. AI isn't. AI is not dead. AI is very much alive. Yes, it's not alive in the carbon biological sense, but in the silicon sense, it is very much alive. Because of this, we need to understand that AI is not a tool like we have been using in the past. It's not a tool, it's not a platform, it's not an object.
For the first time in four billion years, we now share the planet with a new form of intelligence. In fact, Yuval Noah Harari calls it a new sort of life form. It's silicon-based, rather than carbon-based, but it is exhibiting all of the signs and tendencies of a form of life.
Now there are a lots of critics of generative AI, and most of them have an absolute right to criticize it. Chat GPT has been called a stochastic parrot, it's been called dumb, that it can't reason, that it's not creative, and to some extent people are right.
But this is because they're judging AI up against human criteria. Now I think this is a mistake, because to be perfectly honest, if AI can write faster than us, can plan faster than us, can design faster than us and can in time reason and act faster than us, frankly who cares whether it's done in a different way than the human mind. It's completely and totally irrelevant. And I mean that really firmly: it's completely and totally irrelevant.
I mean, a horse and a car work in entirely different ways, but they can still get you from A to B, and they can still get you from A to B with your goods, your family, whatever. It doesn't matter. You get the idea. But what would I want to take? If I wanted to travel from London to Edinburgh, well, I wouldn't take a horse. In fact I'd probably want to fly because it will be even more efficient than driving, or perhaps take the train.
Horse, car, train, aeroplane, all entirely different mechanical processes to get me from A to B, from London to Edinburgh. I really don't care what goes on under the hood. I literally could not give a toss what happens under the hood. All I care about is, I'm here, I need to get to there, and there are certain things that stop me and there'll be some sort of time cost, resource cost in order for me to get from where I am, to where I need to be. So that's what we're going to focus on in this book. Because essentially what we need to do is work out how to take where we currently are with our education system and then slowly but surely, move it into a system that works for the students and of course for the teachers.
My ultimate strategy, or if you like the ultimate goal of my strategy for my own business is very simple. My ultimate goal is to work with schools to help to create schools that everybody wants to attend, not has to attend. That's it. That's my ultimate strategic aim, my ultimate goal. How I get there of course, forms the basis of my strategy. But I know that everything that I do, everything that I say, including this book, is all underpinned by that one north star. How do we create schools that everybody wants to attend, rather than has to attend. Simple, right? Not really. And this is what we'll find in this book.
Because of course, it's not all about us. If it was, it probably would be simple. If we had our own school, we could do whatever the hell we wanted with it, we didn't have to worry about exams and inspections and parents and our bottom line budget, then you know what, yeah, we probably could. We probably could just mix it all up, throw it all out, start again, create something quite radical, and indeed some smaller schools have done just that. But they are in a tiny minority, and even those schools are going to be subject to certain external constraints. For example, a school that makes a grand announcement that it's going to move entirely to project-based learning and then all of the parents are up in arms about, well what about the exams? What about teaching them the stuff they need to pass their exams? Or the school that says that they're getting rid of exams entirely, and then parents are up in arms saying, but what about universities? How will my children be able to get into a good university? You get the idea.
So all of us are subject to some fairly stringent external constraints. But of course, any good strategy will take this into consideration because the second thing to say about strategy at its most simple, is it's just a way of managing and mitigating risk. Because with every journey that we take, there are going to be risks. Whenever we step into a car, we need to plan ahead and we need to manage risks. Stopping at junctions, observing the red light, using our indicators and so on. Some people do it better than others. Those tend to be the people who last longer. People who manage risk badly, who get drunk when they drive, who ignore the rules of the road, who speed etcetera, often end up becoming a cropper or cause grief and injury to other people.
So when we're crafting our school strategy, we are also focusing on how we manage risk. And AI is about as risky as it gets right now, so clearly when we craft our strategy we need to have the risks associated with AI front and center with every single thing we plan and every single action that we execute.
Darren,
I have been contemplating many of the issues you discussed in your email. Especially, the issues of how to develop an educational strategy that used AI to the benefit of all our students and continued to evolve, while dealing with the constraints of a school in our area. Well written and well done! I look forward to reading your book. When do you think it will be coming out?
Great article Darren. Like the river analogy. Using AI to transcribe it has clearly worked well and I actually like the slight 'stream of consciousness' feel too....