What's artificial intelligence got to do with learning and education? Actually, as it turns out quite a lot and in the future it may have an even bigger impact on how teachers teach and how students learn.
I was fortunate to attend the Australian Tutoring Association 2019 annual conference in Melbourne, where I heard educator Moya Gibb Smith give a presentation about the role of artificial intelligence (or AI as its generally known) in education and learning and how it's going to affect students, teachers, and tutors.
Today, Moya joins me on the podcast to explain her first experience of AI when she worked as Learning Support Teacher in a New South Wales country school, and how she sees it developing in the future.
Listen to the interview
- How artificial Intelligence (AI) is being used in education
- What the future might be for AI in education & learning
- A teacher’s experience with an AI enhanced online learning program
- Brain plasticity
- Educational neuroscience
- Dyslexia and other reading disabilities
- The Enigma Machine
People & organisations mentioned
- “The Brain That Changes Itself” by Dr Norman Doidge
- Sir Ken Robinson
- Alan Turing
- Winston Churchill
- Professor John McCarthy
- Ita Buttrose of the Australian Broadcasting Corporation
- The City of Parkes in New South Wales
- Bletchley Park, UK
- Dartmouth College, USA
- USA Education Sector Report
- Individualized Education Programs
- Ebbinghaus Forgetting Curve
If you would like to read the podcast transcript, here it is:
Episode 106 of The Learning Capacity Podcast
What's Happening with Artificial Intelligence in Education & Learning?
Peter Barnes: Hi Moya. Welcome to the podcast. You're going to tell us about artificial intelligence in education. How did you get interested in this subject?
Moya: Well, to tell you how I got interested, I need to tell you a personal story. That is a bit to do with my work history for the past 10 years. For nine years, I worked at a government primary school in Sydney, and I'd been there for some years. My role there was to take the students in year one who weren't progressing in their reading, and to give them intensive tuition, three times a week, in a small group of eight.
I loved this job but there came a year where I got two little boys called Robbie and Nicholas, and despite my best efforts of singing, dancing, Play-Doh, Sight Words, you name it, Robbie and Nicholas just didn't get the reading thing. Eventually my principal came to me and she said, "Moya, what's happening with those two?" I said, "Honestly, Maureen, I don't know. I'm doing all my usual stuff."
I went away and thought about it for a few days and I thought, "These two boys are never going to learn to read unless I increase the number of times I see them each week."
So, that was the start of our early birds program. With the help of volunteers and parents, we started giving each child a before school lesson in reading. We used a very well established and research program from a Sydney university, and every day we would have up to 16 children before school come and have a one on one reading tuition lesson.
I did that for three years, and then we had a new principal who wasn't as interested in having a before school class. At that stage, my husband and I were ready for a bit of an adventure, and we decided to move to the country. We moved out to Parkes in Central New South Wales and we got jobs at a Catholic boarding school.
It was there I was once again working in learning support. It was there that we had come to the school a very unusual young man that nobody really knew what to make of.
One day the principal came to me and he said, "What are we going to do about this young fella?" I said, "I don't know, but I've heard of a really good program that is augmented by artificial intelligence and I've heard it's excellent, but I've never tried it."
He said, "Okay, go find out about it." Very shortly after that, we started with 20 students that were chosen from basically the ones that nobody knew what to make of or what to do with.
The first year we had those 20 students, the program was so successful that the next year we did 50 students. As far as I was concerned, every child that walked through the class should have had two terms of this online, computer based AI assisted program. It was so effective.
Then we went overseas. Using this program at the school in the country really made me aware of the difference between what I had been doing at the government primary school and what I was now doing.
Instead of running a whole team of volunteers and all that entailed, suddenly the computer was doing all the work.
And yet I knew that the children were getting what they needed, because the computer program was designed in the same way that the brain likes to learn. I always thought it was kind of like the difference of going from a landline to an iPhone.
Peter: So Moya, the artificial intelligence that was enhancing this program you were using was, do you say mimicking how brains learn?
Moya: Absolutely Peter. There's been quite a lot of research done on this in the United States. Research conducted by neuroscientists. They were actually studying brain plasticity.
And out of their research into brain plasticity, they realised it had implications for the classroom. The advent of the functional MRI allowed doctors and scientists to actually observe a brain while it was working.
In other words, they could have a child with no reading disability read, and they could observe what their brain did while that was happening. And then they could have a child with dyslexia read the same passage, and see what happened in their brain, what areas lit up, what didn't light up.
Then they were able to devise a program that would address the issues in the child with dyslexia, and help them develop the ability to process language.
It got me interested in AI in general. I read some fascinating books. There's a great book by Norman Doidge called The Brain That Changes Itself, but I also started looking into AI in general.
Basically, this is what I was talking about down in Melbourne. I started with the history of AI. But to really understand the history of AI, you need to have a look at the history of education in general, and Sir Ken Robinson, famous British educationalist talks about the factory system and how the education system was based on the factory system.
It grew out of a need, because of the industrial revolution, for people to work in the factory. They needed people who were literate, who were punctual, who could do a reasonable job and basically who would do as they were told.
The education system was set up along factory lines. There were bells. You started at a set time. Children were educated in batches according to their date of production. It was all very regimented, because that's what factories needed.
Peter: They had a production date? I guess you're talking about when they were born, their birthdate?
Moya: Yeah, their birthdate. That's not my joke, that's Ken Robinson's joke, but it's a good one.
And then of course comes the Second World War, and things are changing very rapidly. In England, during the Second World War in Bletchley Park, you've probably all heard about the British team who were fighting the Germans Enigma machine. They were desperate to crack the code of the Enigma machine in order to improve their chances of winning the war. There was a chap there called Alan Turing.
He was working with computers but it was his idea, there had been computers before that, but it was his idea to instruct the computer how to think. He actually said, "I propose to consider how a machine could think."
He wrote the first algorithms that were a set of instructions for a machine. If A happens, then do B. Something like that.
They were successful, and Winston Churchill said that they reduced the length of the war probably by about two years, by the fact that they were able to crack there Germans' Enigma machine.
That was in the 1940s, and that's where that all stayed until around about a decade later in America, Dartmouth College, there was a professor there called John McCarthy. He was the first person who actually used and defined the term “artificial intelligence”.
It was at that time that it began to leak out of the laboratory and into the real world. Prior to then, it had just stayed in the universities with the professors and the scientists.
But a few things started to make a difference. The beginning of data analysis, the advent of the PC, the advent of the internet, and finally the iPhone where we're all walking around with a phone and computer at all times.
Suddenly, the AI that started with Turing was developing under McCarthy, leaves the laboratory and goes out into the world. That's where we're at now.
Peter: Well, that's really interesting, Moya. What can AI, artificial intelligence, do now in education?
Moya: Well Peter, I think AI in education is still in its infancy but the artificial intelligence market in the US Education Sector Report, it's a big name, said that it's likely to increase by about 40% to the year 2020.
Peter: Was that a report done by some education university or government?
Moya: That's right, yep. In the United States. Currently AI can mark short answer questions and multiple choice questions, it can relieve teachers of a lot of administrative tasks, it can analyze student data. It's really good at that. It's great at picking up when a student is starting to fail their course, and alerting the teacher and the student. It can also identify when a teacher has missed the mark in something that they're teaching.
So, if every student doesn't understand a particular concept, then that's good for the teacher to know, and to go back and revise. That's what it can do at the moment.
Peter: You mentioned to me earlier Moya, before we started recording this, that the AI assisted program you were working with out in that country school had some way of adapting the degree of difficulty to each student. Each student, I understand, was working at a different level depending on how their brain was functioning. Is my understanding correct on that?
Moya: Absolutely right, Peter. Each child in that classroom was basically on an IEP, Individualized Education Program. Because the concept would respond to the child. Every keystroke they made, the computer would adjust the level of difficulty. If they were getting a few right, it would make it a little bit harder for them. If they were getting a few wrong, it would drop back down.
The aim was to keep the child in the sweet spot of learning, about 80% comfortable and about 20% challenged.
This is actually one of the ways the brain likes to learn. The human brain has its favourite ways of learning and it likes to stay in that sweet spot.
It also needs to be, if you really want to learn something, you need to do it frequently. You need to do it at least five days a week. You need to practice a skill, and it needs to have a slight time pressure on it as well.
Peter: I remember when I was a kid learning to play tennis, the coach said, "You've got to practice things." I'd go and every day I'd practice because I was quite enjoying it. In the early stages, there was a lot of challenge. But if it was too hard, I'd get discouraged. And then as I learned a bit more about it, if they kept giving me easy shots, I got bored, right? That's what you're talking about, isn’t it?
Moya: It's that skill, keeping a child right there where they're challenged but not overwhelmed. You have to keep refining the challenge and AI is really good at that. It's great at differentiation. Getting back to the five days a week, you obviously put effort into learning to play tennis, and you have to do that.
There's this great graph that I found while I was doing the research for this talk called the Ebbinghaus Forgetting Curve. If you Google it, it's E-B-B-I-N-H-A-U-S, Ebbinghaus Forgetting Curve.
If you want to remember something, the chances are if you go over it for several days after you first learned that thing, you have a much better chance of remembering it. It seems so simple, but this Ebbinghaus Curve actually shows you how quickly we forget things, so worth checking out.
Another way that the brain likes to learn is it likes to engage all of our cognition. It likes music, it likes color, it likes movement and place and time, and when all of those things are combined, we have a much better chance of remembering something.
That was another thing about this computer program that I love. When a child got something right, on the screen, something would happen. Like, a little bird would pop out and sing a few bars or whatever, and that would release a tiny amount of endorphin in the child's brain, which would cement the learning that had just occurred.
The computer program was designed to work the way the brain likes to learn.
Peter: So the computer program, the artificial intelligence component in there, the algorithms that they've developed were what was enabling it to do all this stuff you're talking about, is that correct?
Peter: That's an example of artificial intelligence actually being used today, in the actual learning process, yeah?
Moya: Yeah. They're thinking that it's going to develop, and I don't think anyone really knows just how big or how it's going to develop.
But one thing that I have heard is that because of all this facial recognition software that it may be the case that a computer can scan a child's face and recognize when they're looking confused or bewildered, and then respond accordingly.
We just really don't know how big this is going to be in education.
But when you look around at what artificial intelligence is doing in other aspects of our lives, it's checking your CV, it's checking your credit rating, it's scanning your face at a railway station. There's a lot.
Peter: For all the teachers who are listening today, should they be scared that their jobs are going to go away because artificial intelligence and robots and facial recognition, and all of this stuff that's happening?
Moya: I don't think scared is the right word. But things are going to change. As Ita Buttrose said to the friends of the ABC, a couple of weeks ago, she said in the face of evolution, it's those who adapt that survive.
I think that's what it's going to be required, there's going to be adaptation required.
From a personal point of view, having used these programs, having seen their effectiveness, having seen how they can turn a child around…somebody who has a blank look on their face is suddenly engaged and able to do what you're asking them to do, having seen how effective these programs can be, I actually think it's a bit of a crime that the Australian government is not utilizing them in schools today.
They exist, they're out there, they're available, we should be utilizing them.
Peter: Wow. That's a big call.
Moya: Yeah. I believe it. I really believe it, and I think the whole philosophy of school is going to change.
Peter: As a teacher, when you were using this artificial intelligence, enhanced learning program, what changed for you? Did it take away a lot of ... You said you had 16 helpers in the room, and then you didn't need any. What did it do for you personally as a teacher in terms of the grunt work, the hard-
Moya: Hard yards.
Peter: The hard yards, and did it enable you to use the time, your time, on higher value teaching activities?
Moya: Absolutely. What I found was yes, the computer, the algorithm did the grunt work. All the hard yards were taken off my shoulders by the computer. That allowed me to focus in on those kids who were even on the computer struggling.
So, there will always be children who have very severe difficulty with learning, and that allowed me to give them the extra time to sit with them while they were working on the computer program, to slow them down, to make them think, and to go ahead like that.
Peter: And in terms of your satisfaction and enjoyment as a teacher, did that diminish or made no difference, or did it improve how you felt about what you were doing?
Moya: I loved what I was doing, because it was so effective. Having spent 10 years trying to help kids, like I'm one of those people that is always reading. I love reading, it's always been something that I've enjoyed all my life, and so when I became a teacher, that's what I wanted to do. I wanted to help children get that skill of reading, and particularly for those kids that it's difficult for.
That was my aim, and suddenly I had a tool that allowed me to do that. Instead of all this hard work and frustration and feeling like there was still some kids that I was failing, suddenly I had this great tool that I could use and it's just amazing.
When you see kids, like teachers coming up to me and saying, "This child has never put pen to paper in my class ever, and today he handed in a half a page." I had twin boys come to that school, their mother was a teacher and she said, "Look, one of my boys is not as strong academically as the other, I'm a bit concerned."
I said, "All right, well let's put him on the program. Not his brother, just him." The weaker twin, the weaker academically twin went on the program. After a term, his brother came to see me and said, "Mrs Smith, can I go on the program?" I said, "Why? Why do you want to go on the program?"
The reason was, he was losing his advantage. He wanted to stay ahead of his brother. In the end, they both went on and they did very well.
Peter: Moya, that's fascinating. What would your advice be to educators out there today who are becoming aware of artificial intelligence and even those who are not yet conscious of this revolution, if it is such a thing, is about to happen or is happening?
Moya: Oh Peter, you've used a really good word there, because it is a revolution. We had the industrial revolution and the current education system that was set up for that revolution, and now we have the information revolution. It is going to impact every single teacher, every single student, so it's coming.
It's going to change the way we operate, and we really need to be aware of it and to embrace it when it enhances our teaching abilities.
It's a fantastic tool, but it can never replace the teacher. We always remember that teacher from our childhood who inspired us, whom we loved, who understood us, who got us. Everyone knows the name of that teacher. We never forget it our entire lives. That's what a computer can never replace, but a computer can be a fantastic aid to a great teacher.
Peter: Wonderful, Moya. Thank you for that advice and thanks for being on the podcast.