Education in the age of AI (Artificial Intelligence) | Dale Lane | TEDxWinchester

foreign we're not teaching children enough about artificial intelligence now I'm not going to get into technical specifics today for the purposes of this talk I won't be getting any more specific than AI which is an umbrella term covering many different Technologies AI just means using technology to do something that looks intelligent often but not always it includes getting a computer to learn from patterns that can find in data now I'm assuming you're aware that AI is playing an increasing role in the world maybe you've heard that AI is used by recruiters to do CV screening by Banks to do fraud detection by law enforcement to do predictive policing we're using AI everywhere and I don't think that's going to be reducing anytime soon children will be using AI Technologies when they go into the world of work now we've always had to teach children about the technologies that will impact them crucially we've not only had to make sure they know how to use them but we've had to prepare them for the risks take email as an example email is a ubiquitous technology today it drives businesses in every industry well the risks with email include things like the spreading of computer viruses so for years now even the most tech disinterested student has left school knowing not only how to send and receive email but understanding that they shouldn't open random attachments they receive we needed to prepare them for a world where email was becoming a core ubiquitous tool well AI Technologies are on the verge of becoming core and ubiquitous so we're going to need children to understand the risks and the implications of this Tech so that's what I'm going to be talking about today I want to cover three questions what I think we need to teach when we need to start and how we could do it but I'm going to start with the what democratization of Technology this is a term we use in Tech to describe how as Technologies evolve they become accessible to more people printing is a good example of this in medieval times creating a book was a manual process it was the work of trained experts like monks in monasteries who would craft beautiful handwritten Bibles most children weren't taught to read and write but most children were never going to own a book but technology evolved the creation of the printing press in the 15th century led to the mass production of books the development of paper mills in the 19th century made printed materials affordable and as technology evolved curriculums evolved with it we started teaching children to read and write in part because businesses were exploiting these Technologies and that set an expectation for education it didn't stop there the introduction of typewriters in the late 19th century made the production of standardized documents faster and cheaper typing became an essential skill in the workplace office and secretarial jobs that didn't exist before were created schools responded by teaching typing classes digital Technologies replaced typewriters word processors gave everyone the tools to produce beautiful documents this influenced the workplace with every industry needing to produce documents leaflets posters and brochures schools responded by teaching word processing and desktop publishing the evolution continued the internet replaced the need for some printed documents it enabled instant publication of any content this transformed the workplace again every industry needing to publish online schools responded by teaching the languages and tools for creating web pages democratization of Technology the evolution of a technology so that it becomes increasingly accessible to broader swathes of the population printing is just one example this Evolution has been repeated for technology after technology and as this happens education has always needed to respond by changing what we teach children as you heard in my intro I work for IBM at our hersley park site just down the hill from here IBM moved into the site in 1958 and in some of the earliest photos we have of the site from back then you can see computers being created by and used by experts in white lab coats Computing was exclusive using a computer was a specialist domain needing specialist skills in the 1940s the chairman of IBM was quoted as saying that the world would only need five computers there is a chance he didn't actually say that but thinking like that was widespread even in the 1970s the founder of deck was quoted as saying there was no reason anyone would want a computer in their home it was enough then to teach children that computers existed but Computing evolved computers became pervasive all Industries started using computers all industries were transformed by computers now computers were still seen as something created by experts sometimes still in those white lab coats but we started thinking of computers as tools that could be used by everyone the spreadsheet became a key business tool today there are millions of office workers around the world who although they don't think of themselves as programmers they are able to use Excel to slice and dice data schools taught children to use Tech like Excel because they were going to need to use tools like it the evolution continued the barriers to entering Computing continued to lower computers became something that everyone could use to create something new you no longer needed that white lab coat you no longer needed hundreds of programmers with degrees in computer science think about WhatsApp or Instagram or Minecraft or Uber or WordPress or even Google each of them were started by just a few people and each of them have helped change the world schools responded by teaching coding they started teaching children how to make a computer Do Something New by describing what they wanted in code they started teaching not just how to use a computer but how to make a computer do new things what education teaches children about any technology has always evolved in response to how far the technology is in this democratization but as a technology evolves like this how do we decide what we need to teach to all children how do we decide what we can just teach to the older students who choose to specialize in that Tech well the key factor is how ubiquitous that technology will be across how many Industries this is the tech for replacing the transcatheter aortic valve it has transformed heart surgery I'm barely aware that it exists I certainly don't need to know how it works I'm glad there are experts in this Tech I hope they'll be available if I'm ever unlucky enough to need them but as a tech the usage of this has been limited to a specific domain so we haven't needed to teach all children about it well AI isn't like that the usage of AI isn't going to be limited to a single domain it won't be restricted to one field AI is not only going to be used in one industry every industry will find transformative ways to apply artificial intelligence take car mechanics for example a modern car runs on over 100 million lines of computer code so car mechanics already need to use complex software but AI is fast going to become a crucial part of their Diagnostics and maintenance or take Farmers Farmers already use software to Monitor and optimize things like crop growth and irrigation but AI will Fast become an essential tool for predicting things like the best time to plant and harvest crops I could give you examples for any industry we need schools to get to the make with it end with AI it's not going to be sufficient for schools to stay at the acknowledging it exists end but when when do we need to do this how far along should we be now where is AI on that evolution timeline well AI started in the 1950s IBM was making speech recognition systems in 1962. in 86 Carnegie Mellon created a self-driving car in 98 Amazon created systems that would predict what we might like to buy in 2011 Apple created Siri to understand our commands and take actions in response AI is not new AI is not futuristic you don't need a degree in Computing to build an AI anymore we've done the experts in white lab coat spit it's time that education got more involved even chat GPT which has got the public so interested and even got mentions today the core Tech behind gbt is something we've had for years admittedly chat GPT is working at a massive scale and it's got a lot of Polish now some have suggested that we should be teaching children to use tools like chat GPT I do think that is reasonable Technologies like it will become standard tools by the time our children leave school chat GPT or some tech like it will become that Excel of the future that tool that enables millions of otherwise non-technical people people who don't think of themselves as programmers to use computers to do amazing things so okay let's teach our children to use tools like this but that isn't enough that can't be our response that can't be the answer Wayne Gretzky's skates where the puck is going quote is admittedly a cliche but I couldn't think of a better analogy because education needs to catch up education needs to skate to where AI is going and AI is moving fast the need for understanding how to use AI tools is in Industry now it's here today people who can use AI tools are outperforming people who don't today now schools can't help with that schools need to get ready for what's coming next we need to get to the teach them to make with the end I've been creating AI resources for schools for six years and in my experience the best way to do this is to give students AI Tech to make stuff with so they can learn firsthand how this Tech behaves I often get school kids to use a computer game as the basis for creating their first AI they play a game they collect all of the moves that they make and they use those moves to train their own artificial intelligence project they're training an AI to play a game using their own playing to create the training data now some of my students are impatient eager to test their project and their AI Control character tends to play badly because with only a few training examples that AI Control character might not have even learned to navigate The Maze and it gets stuck in a corner or it can get stuck against a wall so they go back and they do more training they collect more examples of themselves playing and with a few more examples their AI Control character learns to navigate The Maze but quickly gets caught by the enemy so they go back and they do more training they collect more and more examples of themselves playing and soon their AI Control character starts playing well by the end of the lesson they've collected loads of training data and some see their character able to evade the enemy forever what these students discover is the correlation between the amount of training data and the accuracy of an AI system this is a crucial principle in applying AI Tech the more data you have to train with the better it tends to perform and students can discover this for themselves through their own projects noticing how they improve as they add more training realizing this opens their eyes to how companies need to find Opportunities to collect training data for AIS they start to notice things like how a company could get billions of people to label training examples like examples of things they might need a self-driving car to learn to recognize another project I've helped students make is their own virtual assistant something like an Amazon Alexa they train an AI to recognize commands by writing examples of how they would phrase those commands by giving them time to play an experiment they can test it out they can see what it recognizes correctly and what trips it up I had one student who tested their AI project with activate the spinny thing as a way of asking to turn on a fan which is inspired but their AI project didn't understand that and turned on a lamp instead so I asked them how they could fix that they said that if they added activate the spinny thing to their training examples for the turn on the fan command then their system would learn to recognize commands like that in the future and that's a really good answer but with a little prompting they started talking about collecting everything their project got wrong adding all of those mistakes to the training they recognize for themselves how this would make their project get better and better at recognizing commands the more that it was used with a little understanding about how AI Technologies behave this is a natural and an obvious thing to do I've seen hundreds of children come to similar realizations but when Amazon was reported to be doing something like this the news media was surprised Tech journalists wrote countless articles about how shocked they were that an AI team might want to do this Primary School students I've worked with not only easily understand the motivation but often suggest similar approaches by themselves now I am not saying that all school children think something like Amazon's actions were acceptable most School classes I've LED have been mixed on this issue some students think it was reasonable others think it was inappropriate and I've watched classes have fantastic debates on this The crucial thing is they're debating it in the context of an understanding and understanding of the technology and an understanding of the motivation of the companies running these systems that is exactly what we should be enabling I found that children even at a primary school age can understand the core principles of artificial intelligence projects if they're given access to the technology in a familiar sandbox and the time to experiment they can learn what the tech can do and their natural creativity means they come up with all sorts of fun ideas for building with it they can understand enough about how the tech behaves to recognize risks and to have intelligent debates about the need for oversight they might not use phrases like AI ethics but an innate sense of fairness and Justice takes them a long way in understanding the issues first-hand experiences in creating their own AI projects shows them enough about how AI projects are made to give them a bit of an understanding into the motivations and the behaviors of the companies applying them at scale we need to be doing this now we don't let children learn to drive without making sure they understand enough about how cars work to use them safely we teach them those risks of email because we know they're going to be using that when they start work we teach them the risks of social media because we know they're going to grow up using that well they are going to be using AI AI will play a major transformative role in whatever industry they pursue a career in AI is going to impact their lives it'll impact what job they can get what bank loans and mortgages they get what news they see and so much more besides they're going to grow up using AI systems we can only dream of today we need to make sure that they're prepared for that we need to demand that understanding artificial intelligence is a fundamental part of every student's education thank you for listening [Applause]

Channel: TEDx Talks

Views: 36,181

Keywords: AI, Children, Education, Education reform, English, Learning, Schools, TEDxTalks, Web, [TEDxEID:51573]

Id: m6dyCRS8EmI

Length: 16min 7sec (967 seconds)

Published: Tue Jul 18 2023