Neuroscience, AI and the Future of Education | Scott Bolland | TEDxSouthBank

sometimes the things that affect you the most as a child are the things you become most passionate about as an adult for me it's education I was doomed to fail from the start you see when my mother was pregnant she suffered a series of complications and the doctors told her that there was only a 50% chance that I was going to make it and if I did I'd probably be brain-damaged and wouldn't achieve much in life the best thing that she could do was take me home and just to love me true to the doctor's predictions I was slow to walk slow to talk and slow to learn I struggled to maintain a C average in primary school in high school and I scraped into University with the lowest grade permissible so given this fantastic track record how do you think I was going to do it at uni well I ended up with a PhD in cognitive science the scientific study of how the mind works I ended up with a university medal for outstanding academic achievement and I ended up with a GPA that put me at the top one percent of the student population so from a see average brain-damaged student in high school to the top 1% overnight so what was the secret well the secret is I am nobody special and I hope to thoroughly convince you of that by the end of this talk instead I stumbled across some fundamental truths about learning some basic principles if applied to the education system would change millions of lives and it comes down to basic neuroscience the truth is were all fantastic learners and we all have an innate thirst for knowledge from infancy nobody teaches us how to crawl or grasp objects or to walk we learn such things through self-directed exploration and play interestingly the highest concentration of the brain's endorphin receptors are actually found in the learning centers of the brain we're actually hardwired to learn and to find great joy in learning but if this is the case why is it that so many kids struggle at school and the answer is that there's a mismatch between how we teach and how the brain naturally learns now if we look at the learning centers of the brain they release endorphins with an inverted u-shape with respect to familiarity so things that are too familiar are boring things that are too unfamiliar are aversive but things on the periphery of our knowledge that extend our capabilities are actually highly pleasurable not only our endorphins released making learning enjoyable but dopamine is also released making learning highly addictive so who's checked Facebook today so if everybody wrote the same things every single day on Facebook it would be too boring and be too familiar and you wouldn't do it if it was written in some sort of foreign language that you didn't understand to be too unfamiliar and you wouldn't do it Facebook is pleasurable and addictive because you're always getting new information on the periphery of your knowledge you're always learning and it's this natural thirst for knowledge which is how kids naturally learn and how competencies naturally emerge take for instance how kids learn to physically interact with their worlds a few years ago I ran some experiments using robots to actually test some of these theories and what we did we gave these robots basic vision and here and basic reflexes like a baby but we also gave these robots happiness so that is the preference for exploring on the periphery of knowledge and what happened like a baby their initial explorations were completely random but over time what this was actually teaching and that random activity doesn't really do that much and once it learnt that it got bored of it and actually stopped by having randomly but there were some cases where interesting things actually would happen particularly when you put your hand in front of your face you'd see this white blob so over time instead of acting randomly it started learning hand-eye coordination and this is basically what children do as well but again over time this became boring and it stopped doing it very similar to how a child splashes in the bath we don't do that because we know what it actually does and it gets boring right but over time again there were some interesting things that was happening when it put its hand in front of its face particularly when there was an object present so when there was a hanging object if you hit it it would swing if you hit a button music would play in this case Jorah the Explorer which is very sad fascinating for kids and if you turn around and bit on a chew toy it would activate the sensors in your mouth so over time this robot learnt to master its environment none of this was pre-programmed it learned through happiness in the same way that a child learns finding joy in self exploration and play so we are all hardwired to find learning incredibly pleasurable so none of us if given the choice would stare blankly at a wall what we'd rather do is engage in activities such as reading or checking checking Facebook or watching YouTube or even going to Ted all right so things that that actually activate the learning centers of the brain so if this is the case again why is it that some students don't do well at school well let's fill a classroom with 30 students and see what happens so we have a teacher up the front teaching a set curriculum to a set pace unfortunately every student is different so some students are going to be ahead and bored others behind finding learning particularly aversive very few students are actually getting the information on the periphery of their knowledge that's required for optimal learning and pleasure and as a result of this millions of children are getting left behind one study of North in North America actually showed that 63% other student population are disengaged right and this is not a fault of the student this is a fault of the system so thinking about back about my own learning journey it actually makes a lot of sense in primary in high school I was just one of the 63% bored out of my mind's I was sitting there just waiting for the data end but at home was a different matter it's in home I had freedom to actually explore things and I had quite a natural interest in computers so the the age of 10 I started teaching myself how to code I would read every book there was and taught myself nearly every language that there was and by the time I hit university I was an expert my wife prefers to use the term geek but she's not here so I'll pretend the correct terminology is expert so what happened when I got to university why did I suddenly do well because I had freedom of choice those things that I was interested in which was all the geeky things like cognitive neuroscience philosophy and I could choose those subjects and I as a result I loved going to class and I did well unfortunately my students are not given this opportunity and as a result are stuck in a system that diminishes rather than inspires them so Ken Robinson has said on various TED Talks that the education system doesn't need to be reformed it needs to be transformed we need less standardization and more personalization but the real question is how so given the teaching already ranks as one of the top 10 most stressful occupations how is it possible that a teacher could do even more work and personalize the learning journey for every student in the class it seems an impossible task until now so I believe the answer to the future lies in technology and in particular my own area of research which is artificial intelligence so AI can now be viewed as an intelligent tutor that will sit over your shoulder and give you personalized education and what I'd like to do is give you some very practical examples of this and I'd like to demo my own software that that I'm currently working on and I'd like to talk about three levels of AI the different complexity and how they'll be used so the first level is around rote learning and this is the boring stuff we all have to do right so what is five times seven what's the capital of France what's the French word for happy name this bone but it's applicable to a whole wide range of tasks so how do students usually learn this well we give them an exam and if they're a good student they'll probably cram for it but if you cram for an exam there's actually an exponential forgetting curve so most of what you've actually learned is forgotten in the next few days and as a result the kids haven't actually learned all that much the best way to write learn information is what's called spaced repetition so you actually refresh the information just when it's about to be forgotten and every time you do this that forgetting curve gets less steep and you actually remember more over time interestingly some studies have shown that with exactly the same amount of study time you can double your performance just by switching to spaced repetition the other question is how to study so what to do if your rote learning a lot of students just grab their textbook and highlight and read and reread until they feel that they know it unfortunately according to the science again this is a really horrible way of doing things the best way to rote learn information is through what's called active recall so that's actually testing yourself like for example using flashcards so what is the capital of France and when you do this you remember a lot more so given a classroom of 30 30 students to actually optimize rote learning what I need to do is know exactly what every single student that class knows or doesn't know and present exactly the right material at the exactly the right time for optimize performance and this is impossible for a teacher but actually quite trivial using technology so this is my own sort of system that allows any any teacher to actually create their own tutorials upload it to the web and share it worldwide and what we can do is create lesson plans teachers can use a simple drag-and-drop interface for embedding things like our YouTube videos animations for teaching the material but then we have these flashcards and flashcards can do anything such as teaching you having quizzes such what's the capital of Victoria we can use AI to do character recognition so draw the Japanese symbol for five and even audio recognition so player middle see all the student has to do is log into the system and choose a topic on which to work and the system takes care of the rest the system actually it generates the forgetting curve for every piece of information presenting exactly what the student needs to know at exactly that point for optimal learning and retention and of course being a computer system you can gamify it in whatever way to make it more fun for kids so that's rote learning that's a good start but we want to do better than that we actually want to get into active learning so teaching things like problem solving and creativity and this is where generative AI level two comes in and generative AI can actually create its own questions for the child to answer so for example in teaching music what generative AI can do is not only listen to the student and provide feedback but can actually generate its own music match to the skill level of the student so it knows what the student can and can't do and actually generates the right music for that student and again this is applicable to a whole range of topics from touch-typing through to mathematics it's about generating the right problems for that student at that point in time and this we we can do with with today's technology so how would we use this in the classroom so first of all we get rid of boring lectures and textbooks and we replace this with personalized AI enabled elearning platforms that really optimize the learning process and this will t the children what they need to know but then the the role of the teacher becomes more exciting it's around teaching the kids why is this information important why do we need to learn all this math stuff what are the cool things that we can do once we know this information so it's around creating personalized group projects that the kids can work on that actually link the information to things that they actually care about and again we can use technology social networking for sharing these ideas across the globe so you know a single teacher does not have to do all the work so again this is stuff that we can do now with today's technology so the next level is level three which is really what's coming up in the future and this is integrative AI where we actually match generative AI which is the ability to create its own questions with various technologies so for example in teaching a child a foreign language we can use virtual reality and imbed the the child in particular contexts and really interesting scenarios that teach that subject in the shortest amount of time we can link this to other existing technologies such as gesture recognition so you know the Xbox at the moment has this wonderful system where it can actually detect movement and your gestures we can link this to generative AI so I can teach things such as dance martial arts even sign language so the point is we can use AI for generating exactly what the student needs to learn at that point in time for optimal learning so AI is here we can use it now but it is I guess expanding and improving at an exponential rate there is a tsunami of change about to hit our Shore which I believe will totally transform the way we approach education and if we do this right will enable every child to unlock their hidden potential and to live a full rich and valued life but there are some things we all can do in the mean time there was a quote I found on the internet which sometimes people say so Socrates said but it never did but it's a really nice crush and what it says is education is the kindling of a flame not the filling of a vessel what every child needs is someone to delight in them someone to help them find their natural joy and curiosity someone to help them explore at their own pace topics that fascinate them someone to help Kindle the flame be that someone thank you you

Channel: TEDx Talks

Views: 119,267

Rating: 4.9082627 out of 5

Keywords: TEDxTalks, English, Australia, Education, Cognitive science, Engineering, Learning, Schools, Science, Technology

Id: _cYIvfS-knA

Length: 15min 36sec (936 seconds)

Published: Thu Jun 30 2016