.jpg)
Preparing for AI: The AI Podcast for Everybody
Welcome to Preparing for AI. The AI podcast for everybody. We explore the human and social impacts of AI, including the effect of AI on jobs, safe development of AI, and where AI overlaps with sustainability.
We dig deep into the barriers to change, the backlash that’s coming and put forward ideas for solutions and actions which individuals, organisations and society can take, and how you as an individual can get ready for what’s coming next !
Preparing for AI: The AI Podcast for Everybody
FACTORIES OF RESILIENCE: How AI is changing the role of schools with Julian Fisher
In part two of a three episode deep dive into the education industry, discover how AI can revolutionise learning by providing customised experiences and alleviating teachers' burdens, allowing them to focus on what truly matters—igniting creativity and fostering meaningful connections. Guest Julian Fisher, founder and CEO of Venture Education, lends his unique insights from his journey in international schooling, highlighting the balance between technology and human touch in cultivating resilient, well-rounded students. In the shadow of a rapidly changing and uncertain future should the focus of schools be on creating a more resilient generation of children?
Ever wondered if AI could shift education from job training to life skills? Let's delve into this fascinating possibility, emphasising the irreplaceable role of human teachers in fostering emotional and social growth. Drawing from the Montessori model, we explore how AI can complement, rather than replace, human interaction in education, preparing students not just for careers but for life. Julian's experiences in China provide a window into how different educational institutions are embracing AI, from minimal use to full integration, shedding light on both the promise and pitfalls of this technology.
Is technology the future or a threat to traditional education methods? We weigh the benefits of AI in easing administrative loads against the risks of over-reliance, especially in tasks like essay writing. As we reflect on the slow technological adoption in education, particularly during the COVID-19 pandemic, we underscore the need for a cautious yet optimistic approach. The importance of resilience in the digital age, practical strategies to balance tech use, and the evolving landscape of performance assessment are all thoughtfully examined. Join us as we navigate these complexities, ensuring a balanced and enriched learning environment for future generations.
On Exactitude in Science: Borges.pdf (kwarc.info)
Venture Education | Empowering Education in China
Welcome to Preparing for AI, the AI podcast for everybody. With your hosts, jimmy Rhodes, and me, matt Cartwright, we explore the human and social impacts of AI, looking at the impact on jobs, ai and sustainability and, most importantly, the urgent need for safe development of AI, governance and alignment. Urgent need for safe development of AI, governance and alignment. No more pencils, no more books, no more teachers' dirty looks Out for summer, out till fall. We might not come back at all. This is Preparing for AI with me, matt Cartwright, and me, jimmy Rhodes, and this is the second episode of a kind of short series that we're doing on the education industry. So we'll have an interview coming up with Julian Fisher, but before we do that, me and Jimmy are just going to have a bit of an introduction and a bit of a chat about our views on how we think AI is going to shape education over the next 2, 5, 10, 15 years. So, jimmy, do you want to start off?
Jimmy Rhodes:with your views. Yeah, sure, thanks, matt. So education for me is one of those areas where I think there's a. I think I've only got reason to be optimistic here now. I don't know whether there'll be jobs for kids in 10, 20 years, which is a different challenge, but in terms of education, I think it's a massive, massive opportunity. I think that ai can potentially bring better opportunities for education to maybe people who wouldn't normally have access to that level of education. I think it has, because you know it's, there's opportunities there for massive customization and for custom learning, depending on your learning style, depending on your level.
Jimmy Rhodes:You know, I think I think that I'm not. You know, I don't know if education institutions are kind of adapting to ai that quickly, but I just think there's huge, huge opportunities there. Hallucinations aside, I guess, because that's where I suppose that you've got to be a little bit careful. You know that you're not, you're not learning from an ai that's making stuff up, but like with young kids, for example, starting with for young, from for young kids. One of the one of the things that I know you you've used ai for is is just sort of like really encouraging imagination, and that's something that I find is like incredible with ai. Um, I don't know if you want to come in on that yeah.
Matt Cartwright:So I mean I, I mean I use it occasionally. I try not to use it too much, but with my, my daughter and to a lesser extent to my son, we use it for kind of storytelling. I guess at times when you know I've read two stories and she says, can I have one more? And I'm like I need to do something else. And then you let you know, chat, gpt or Claude or whatever come up with a story. But it allows them to think of the theme of the story but also, kind of, you know, add to it. And then when that story comes out it's like the next time we'll say, okay, well, remember that last time, but let's add something into it. So it does kind of stimulate their imagination. I think it's definitely a good use of it. I mean, it's a tough balance, I think, with with kids, particularly young kids, with you know, the pragmatist which thinks, okay, we need to get them ready and to understand these tools. And to you know, I mean they don't necessarily, I don't necessarily think they need to be using them at two, three, four, five years old and and then they won't be able to understand. If they don't, they won't be able to understand them in future. But, you know, understanding that this is your future, you're going to have these tools, are going to be a massive part of your life, but also balancing that with the fact that you want them to. You know, take up sports, and it's probably more important than ever to have hobbies that are not looking at rectangles, as as you often say. So it's a kind of balancing out, but I do agree. One of the areas that it's really good, um, is creativity, and the other one, which we did talk about a bit with Eddie on the last education episode, is being able to provide assistance when parents and maybe even teachers can't do it. So particularly things like homework, where you've got a specialist piece like, let's say, your kids study in algebra or calculus and you can't remember any of it from when you studied it, rather than trying to make your not to lose face, but to tell them you know that you understand it. You can use the tools then to do it. So, yeah, I mean there are definitely good uses.
Matt Cartwright:Now I'm pretty optimistic about AI in education as well. Part of the reason I'm optimistic is because I think it's another industry where it will actually sort of augment people, particularly in the next. You know two, three, four, five years, anyway, I think back to you know my. My mom was a teacher. My dad spent part of his life as a teacher. I had friends who were teachers and all of them. It was the lesson planning and stuff like that that they didn't like doing. And in fact my dad said to me I can't imagine how much easier it would have been. I hated all the time we spent planning for lessons and you know, ai tools will help you to not have to spend so much time on that and then potentially, like we said with with you know, doctors and health care, free up some of that burden and allow you to do the stuff that you not only like doing but where you add the most value yeah, totally, totally.
Jimmy Rhodes:I've I've briefly been a teacher and lesson planning is definitely not your favorite thing when you, when you're enjoying teaching kids, um, and then you have to go and spend, you know, nearly as much time faffing around, uh, producing documents and plans and stuff like that, which it can help with.
Matt Cartwright:It's sort of like the front and the back end of the teaching experience, isn't it? It's the bit that no one, really, I think no one wants to do. I'm sure there are some teachers out there that love it, but I think they're the sick minority.
Jimmy Rhodes:I'd be really interested to hear from teachers whether they've used it for marking. And just to clarify here I think we're talking about with everything we're talking about ai assisted, which is where ai is right now. So it'll probably very quickly knock you up a framework for a lesson that you might need to tweak a little bit. It'll probably very quickly mark some papers, but you need to double check it's working and double check it's it's um output and and hopefully the third episode, which I think it'll be three and done for for now, for education.
Matt Cartwright:But I think the third episode will be with a teacher.
Matt Cartwright:So we've got you know different perspectives someone who's worked in in ai with education, julian, who is the founder of a sort of education consultancy, and then a teacher who's kind of right in the at the heart of of teaching and what their views on education are.
Matt Cartwright:You touched on it, jimmy about, about jobs and that there could be losses. I mean, I think, because when we do the industry episodes, we are focused, you know, not not only on the impact on having a job, where it's the impact of how a job is as well, but I think we, you know we always look at whether it's going to lead to, to widespread job losses. And I'd be interested where you, whether you think in the short term, there will be an effect, I mean even even a sort of, you know, positive effect for the teaching industry and that, potentially, if we see other industries start to, you know, really, really start to kind of shed jobs and teaching, doesn't that we drive, you know, high quality people into education, because my understanding is there's a bit of a crisis in recruitment in quite a lot of countries at the moment.
Jimmy Rhodes:Yeah, I mean potentially. I certainly don't think that AI is going to have an immediate impact, if an impact ever, on teaching jobs in terms of job losses. Um, the reason I say that is because it it needs it. It needs that human touch for the foreseeable future. Like part of teaching is socializing kids, socializing, socializing people and and you can't do that with a robot and also who's going to listen to one like there's an element of discipline and class control and all that kind of stuff.
Matt Cartwright:So I think how to tailor things to different kids right, that that that seems like it's. It's beyond any kind of ai in the I see, immediate, the foreseeable future to get to that level of being able to, you know, eyeball six, seven people in a class and decide, well, that one needs me to be harsher to them, that one needs me to put an arm around them, that one needs me to tell them to shut up. You know that that element of being a teacher, I, I, yeah, I don't see how you can replicate that no, it's a human connection, isn't it?
Jimmy Rhodes:and, uh, I don't there's, I mean, that seems something to me that no ai or robot or anything is ever going to replace. I don't think or be able to replace, so, uh, certainly, if that's the case, it's a very, very long way off. But yeah, for me, for me, teaching is an area that, for me, that can only really be upsized, in my opinion, right now, like I've used, I've used AI to teach myself stuff and again, you have to kind of fact check it and keep an eye on it, but it's so much quicker than using Google some of the time to try and find information that you need Not 100% reliable, but I wouldn't say Google is either. You always need to sort of have a critical eye for everything online.
Matt Cartwright:You made a point right at the beginning where you said maybe you say you made a point is the wrong word, but you said something that kind of piqued my interest.
Matt Cartwright:You said that maybe the kids that are being taught won't have any jobs.
Matt Cartwright:At the end of this, I wonder whether and we're talking about education here, and I guess education, we often think about education as leading to a job but I wonder if we're looking at this the wrong way, in that AI kind of helps us to stop looking at education as purely being about how you get somebody into a job and to make the most money or to move as high as possible up that chain, and actually focuses more on skills to exist as a human being in the world, because, you know, I mean man like the problems that we've got in the world at the moment are mental health, and you know a lot of them.
Matt Cartwright:A lot of issues that we've got, I think, in society come down to the inability of people to get together and solve problems or, you know, be resilient to change or to sit down and have different opinions. Is there a case with education that ai frees up? You know a way to spend time actually teaching people to be good, productive human beings in the kind of wider world, rather than just purely being focused on getting them ready to go and have a career yeah, I mean, I think I'm not 100 sure because I'm not fully versed on it but is that not the montessori model?
Matt Cartwright:my daughter goes to montessori and and even though she's been going there for two and a half years, i'm'm still not a hundred percent clear on what the Montessori model is. I like it and I know it kind of teaches you about, you know, the young kids. It teaches them to follow the older kids and the older kids, they teach the younger kids and they learn about relationships and stuff like that. But the Montessori model as such, I've you know this is kind of shameful, but I've never really looked it up. But I'm very with, uh, with the nursery and the job that they do. So, yeah, I mean, they're definitely spending time on teaching them skills that are, you know, not just maths and and you know, counting and letters and stuff, but I think it's more for me that's more about like they're learning real life skills. There's still things that are more sort of practical and translatable.
Matt Cartwright:I think my point there is like, can you spend much more time on physical education and can you spend more time on well-being and can you spend more time teaching people to be, you know, resilient and, like I say, being able to kind of problem solve between people? Learning about not, it's not learning, it's almost like relearning, because I feel it's something we've lost over the last 20 or 30 years I mean, maybe it's shorter than that, but the ability to have a different opinion and to have a constructive engagement with somebody on that, rather than just going well, you think this, you're you're, you're the enemy, kind of thing yeah, and just to clarify, like I mean, I mentioned the montessori thing, but I totally am for this and agree and like and and hear what you're saying.
Jimmy Rhodes:I I think like, if, if it can, if we can get to a point or get to a model where it's less about I think, the focus is less on capitalism and the focus is more on um the human side, then I think that's a good thing and I would definitely be pro that kind of model in education so you don't like capitalism, jimmy?
Matt Cartwright:are you a communist, jimmy? I'm not a communist, you're a socialist I'm not.
Jimmy Rhodes:I just don't think that. I think the fact that money makes the world go around is a kind of narrow-minded view when, like it doesn't, like we're getting into stuff that we don't normally talk about on the podcast, but you know, it's money's not related to happiness. It's been proven time and time again, and all this no, I just I.
Matt Cartwright:I'm not a capitalist. I don't know what I am. I'm not a communist, I'm not a socialist, but I'm not a capitalist. I'm uh, I'm my own individual self your own breed.
Jimmy Rhodes:In all seriousness, like I think I think I'm all, I'm all for that. The best advice I can give to people going through education is find your passion, find what you enjoy and just do that and pretty much ignore everything else. Like, if most people have something they enjoy at school and through life and all the rest of it and and um, if you can just do that and not worry too much about whether it's going to turn into a job or money, um, then actually I think most successful people do what they're passionate at anyway. Um, and you can, you can make money doing anything.
Matt Cartwright:So, and in an ai, utopia, then that's what everyone is hopefully going to be able to do pursue their dreams and not have to be a slave and join the rat race right, yeah, and I hope we get there.
Jimmy Rhodes:Like I uh, I'm not sure I'm optimistic about that and in a in any sense, but I am hopeful, I'm definitely hopeful yeah maybe that's. That's how I describe myself hopeful, but but not optimistic. Yeah, I think. I think optimism is a bit too strong for me.
Matt Cartwright:Right then. Well, jimmy has used his income from the podcast to buy 20,000 NVIDIA GPUs, so he's off into the back room to train a super intelligent AI. I'm going to slip into my silk dressing gown and go into the other studio and interview Julian Fisher, no-transcript. So joining us today is julian fisher. Julian is the co-founder and ceo of venture education. So, julian, welcome to the podcast. Hi, great to be here. Do you want to just maybe a bit of a self introduction? But maybe in that self-introduction you could include anything that you have been sort of using yourself or have seen coming into education industry from an AI perspective already.
Julian Fisher:Sure. So I worked in international schools for about 10 years and then in 2013, I co-founded Venture Education. We started off doing a lot of student facing stuff, summer camps, and we had a learning center, but we started a market intelligence consultancy in about 2017. So you know, we do reports and we kind of provide data and intelligence on what's happening in the Chinese education sector to governments and companies and universities around the world and basically in about early 2023, so about a year and a half ago I had a conversation with a friend about AI a kind of drunken conversation, if I'm honest where he sort of convinced me that it was going to change the world.
Julian Fisher:And immediately the next day I posted a job advert saying I was looking for an AI specialist one day a week in my company, really with no idea of what exactly I wanted them to do. I just sort of realized I needed to do something and I got three applications and two of them were honestly pretty nonsensical. I couldn't actually understand their cover letter. I really didn't understand. They were clearly very technical people. And then the third guy he just kind of seemed a bit interesting. You know I'm based in Beijing, he's an American guy, and so I called him in for an interview. He was an hour late for the job interview, which you know isn't typically a good start. And then, you know, one of my first questions was like right, so what have you done in ai? And he's like nothing. And I was like okay, so what have you done, just more generally? And he's like well, I reverse engineered my xiaomi air conditioning unit and I've done this and this and this. And I was like okay, so like you, are this really bright person who clearly problem solves and kind of just uses technology. You know the way you want it. And so he started with us and I'm I'm fairly convinced he's a genius. He's now full-time with us and, yeah, he kind of completely transformed our organization.
Julian Fisher:So, you know, he started off working internally, so he sort of built us some some kind of tools that we could use internally and I think, especially with what we do with market intelligence, being able to analyze especially open questions. You know, that's where AI has been transformative. You know, when you do a survey with 10,000 parents in China doing that, historically you'd have to go through manually, you'd have to group things. Now, obviously you can put into a large language model and like it will tell you like these are the four main trends according to this open question. So he did a lot of stuff around that, you know, building our own internal systems.
Julian Fisher:But then, as that kind of grew, we were like you know, I think there's a product here that we could actually kind of sell outside of our company. So that's what he's been working on for the last seven months now and it's basically a tool to allow you to compare schools, not just in China. We're gonna start with Asia and then, you know, the product will go global and, of course, because of what AI can do, it's not just quantitative stuff. You know what are school fees and what's the student enrollment. You know that kind of stuff. It's also like what language are schools using about themselves? What news are they reporting through? You know social media and stuff. You know what can that tell us about trends in the sector? So really, really exciting stuff. But again, it all started with a kind of slightly shaky or slightly strange job interview no-transcript.
Matt Cartwright:So if we, if we look at timelines again I mean, you know, two to five years maybe how do you think that ai is going to transform the sector? I mean, what are the things that you are seeing that either you're excited about or you're worried about? I mean it can be from from both perspectives. But if we look at a kind of fairly short timeline I guess it's medium term, two to five years what do you think are going to be the biggest impacts?
Julian Fisher:so I'm I'm sure you've heard this expression before, but I'm, I'm, I'm pretty sure that we're going to kind of overestimate this in the short term and underestimate it in the long term in terms of ai and its impact.
Matt Cartwright:Literally exactly what.
Julian Fisher:What, yeah, two weeks ago, is literally exactly what I said underestimating the long-term impact, but overestimating the short term, yeah and I think and I think you've seen that in education I mean you know even what you just said I mean, in the last six, I think six months ago, every single conference around education globally was like AI, what's it going to do? And you can see, even in that time it's kind of settled a little bit now. So I think one thing I would definitely say and I'm probably going to repeat this over and over, but let me tell a story. I once did an activity. I went to a training and it was led by a school counselor. I went to a training and it was led by a school counselor and she basically got everyone in the room to go up to a whiteboard and on one side, put the positive memories they have from school and on the other side, put the negative memories that they have from school and at the end she basically said look, you know, on the positive side, every single one of those is about human interaction. It was like connecting with lifelong friends. It was a time that a teacher supported you and I think that still stands. So I think good education is still fundamentally about human interaction, and I'm not sure how much of an impact AI is going to have on that.
Julian Fisher:But that being said, you know, when we look at schools particularly, I think it was the teachers that adopted it first. You know, being a teacher I'm going to give a shout out now to all teachers. You know it's incredibly hard to be a good teacher. You know, the comparison I like to make to people who've never taught before is like if you have to give a presentation in your company, you probably spend half a day a day preparing for that presentation to make it engaging and persuasive and all that kind of stuff. You know a teacher a good teacher is doing that six to eight times a day. So I think you saw teachers very early adopting AI, especially chat, gpt, for things like lesson plans, you know, developing PPTs, even things like writing reports that you know when you've got to write reports or marking homework, you know that kind of stuff and yeah, I think in that way it really helped that side of things. Obviously, the next ones who adopted with the students and the students probably weren't quite as smart about how they were using it, because those guys were just plugging in and getting it to write an essay for them and they you know teachers tend to know when you have a student who's kind of, you know, struggling to kind of construct a sentence and they sort of submit an essay. That's like, you know, like PhD level. So I think I think that was the kind of next side of things, I think you know.
Julian Fisher:Then there's a sort of school side of this. So we as a company survey heads across China every year. You know about 60 to 70 school heads and we interview them as well and we asked them this year about their engagement with AI and we kind of came up with four categories for how schools are engaging with AI. So the first one was minimal and actually in some cases they'd almost kind of outlawed AI. They basically told students and staff they weren't allowed to use it. I'm not sure that's the best approach, but that's certainly how some schools are looking at it.
Julian Fisher:I think the next level up was kind of compartmentalized and you see this a lot in schools. So you know you'll have kind of an innovation center where kids go once a week to like study innovation for an hour, which obviously isn't the most effective way of doing it, and sustainability. You know they'll plant trees once a year and in this case they're kind of doing AI one hour a week or like an after school activity. I think the next level up was kind of integrated. You know this idea that it really is something where teachers are using it, students are using it. It's part of a part of the school ecosystem.
Julian Fisher:I would say that the top level that very few are at but a lot of them, I think, probably think they're at is kind of embedded and that's pretty difficult for any school to get to about any topic. So I I think that's probably we're not there yet, even though there is some belief. You know from some schools that they're nearing that. Um, I think probably what we've also seen with schools as well that's that's the most positive step is just like codes of conduct and I think that's really saying look, we don't really know how this is going to work, but at least we can kind of come to some agreement between parents and teachers and students and the school about like when we can use it, when we can't use it, when we have to admit that we're using it, and so I think the best schools, that's probably the level that they've got to yeah, I think one of the really good points there is.
Matt Cartwright:I mean, when we started this podcast, we were focused purely on on the impact on jobs and and not just the impact of, you know, losing jobs, but on how it would affect the way people work. And I think one of the things that we we found whether there are a few industries and teaching is one health care, you know, particularly for kind of doctors and nurses, was another one where in other industries the threat was there that, you know, ai was going to take jobs away, but actually for teachers it's really having an impact on reducing the burden. So, you know, my, my, my mom was a teacher. I know one of the challenges for teachers was always the amount of time that they had to spend on administration and the amount of time that was not spent, like you say, on the face-to-face stuff, the actual interaction, the thing where you need a person, and that already AI tools are helping to reduce that burden. So, like you said, lesson planning, marking, you know, no, I don't think there's any teacher who enjoys marking or enjoys lesson planning particularly.
Matt Cartwright:I think if you get into teaching, you value the interaction. So you know it's great that it has taken some of those things away. I'm interested your view. When you say you don't think that it all those positive memories around, that interaction, and you don't think that that is challenged by AI, I mean, I think I probably agree with you, but do you think that there is a case that you know AI could be used to take some of that teaching time away? I mean, one example I had maybe less so for schools, but in education in higher education is if you have a lecture and you're watching a video lecture or you're attending a lecture where someone is just speaking at you for an hour, I mean, does it matter that it's an ai avatar doing it?
Julian Fisher:rather than a, you know, a real person. Yeah, it's a really good question. I know there's been a lot of discussion around this with marking. I guess one of the one of the fears with ai that I potentially have is that we end up just ai talking to ai. Um, you know, I know that this is true with kind of job applications at the moment. I know a lot of people who are like oh, you know, I got AI to write my CV and then it's like sure, but you do realize that the thing reading is probably AI too, so in this sense it kind of sucks all of the meaning out of it. And I think if kids get to a point where they understand that teachers are actually not marking their work, it's being marked by ai, it will probably lean them more towards the point where they were like then I'll just use ai tools to write it as well at that point what's yeah, what's the purpose of this entire thing.
Julian Fisher:So I think that's probably the danger here a hundred percent.
Matt Cartwright:And and you bring up a really good point it's not education as such, but the one about cvs. I mean, you know the last few months when I was applying for jobs, and the process of applying for a job now is just gaming the computer, gaming the algorithm, you know it's just trying to create. Well, what buzzwords I mean. I hate it. You know what buzzwords do I need to get in and I need to push them further up to the top so they get picked up easier.
Matt Cartwright:And the idea of you know assignments and essays being like that and you're just trying to game the ai, like, like you say, it becomes a point of, well, have I got a better ai? It's just, it just becomes who's using the better ai rather than who's actually producing the best quality work. And again, this is this is something that I think it's it's maybe a stronger argument in higher education, where there's at least a degree of choice to be there, that if you're, you know, if you're getting a large language model to write an assignment for you, you're kind of cheating yourself because you're not learning anything. I think maybe that's more difficult an argument to have in. Let's say, you know primary and secondary education, where you're there not necessarily because you want to, but because you have to.
Julian Fisher:So you're not I will say I am, I, I had that, sorry. I. I had a conversation with a um, uh, somebody in school recently and they said something really really interesting on this. They were like you know, now that we've gone down the kind of ai rabbit hole, we've done training for staff, we've kind of done all these workshops with kids, we've sort of got to this really existential question about the purpose of school and we've kind of broken it down that like now, one of the fundamental things is like explaining to kids why difficult things are good. Um, because you know otherwise.
Julian Fisher:You get that response all the time. We're like well, won't ai do this? Or, you know, won't ai do this in the future? Or or can't I just use AI to do this? And it's like, okay, so now schools in some ways then a kind of factories for developing resilience rather than the kind of transference of skills, because ultimately you know AI, can, you know, create a kind of piece of electric music? Or you know rock music tomorrow, you know, does that mean no one's going to bother learning the electric guitar, cause it's actually really slow and difficult. So I think that was a really interesting reflection on this nature of kind of. Maybe it changes the tone of schools and kind of the message of schools themselves.
Matt Cartwright:Yeah, and I do wonder whether in that, in that example the guitar example is, you know, if creating AI ai me sorry, ai creating music means that less people are inclined to produce their own music, perhaps actually that puts a price premium on being able to produce your own music and actually, and you know, I wish everything wasn't about the economics of it and and whether or not you can make money out of it.
Matt Cartwright:But actually you kind of reach a kind of equilibrium point where actually it goes the other way and actually so few people are able to do some of these skills that actually there becomes a a premium on that skill and then it drives more people back to it. So you know, you see no one playing electric guitar to a certain point and then electric guitar becomes this really sought after skill and then everybody learns electric guitar again. I, I don't know that necessarily happens and and that that's necessarily the right example. But I do wonder about that, whether you know the music industry I think is a good example of one where you know we, something like suno is really fun to play around with and I think for background music and sort of functional music, yes, it's absolutely fine. But there comes a point where why do you pay to go to a concert? You know you go to pay to a concert because you go to see that performance, not just because you want to listen to the sound.
Julian Fisher:So I think there are many things in education that will be like that.
Julian Fisher:We still buy handmade furniture, right, and we've been able to kind of, you know, factory make furniture for hundreds of years.
Julian Fisher:So, yeah, I'm not completely convinced and I have to say I have been pondering that it feels that we're getting to a point where AI is going to be able to produce pretty phenomenal novels and that maybe we should have 2022 as a cutoff point for, like, pre-ai novels and post-AI novels.
Julian Fisher:But then, you know, I do wonder if, in kind of 50 years, 100 years, you're going to have this kind of Japanese artisan and people are going to be like, what's that guy doing? And it's like he's writing a novel. You know he's doing it by himself, like as opposed to just getting it written with an idea in 10 seconds. So I kind of yeah, I do like the idea that a lot of the stuff we do now kind of becomes craft and becomes valued for such in the same way that, yeah, you know, you can get, um, you know, factory made beer, but I still think there's a real passion for craft beer that's made by kind of artisans in small batches. So, yeah, maybe that's going to kind of span across a lot of the things that we do so does that mean that education of you know more craft skills?
Matt Cartwright:you know everyone wants to study woodwork again because they want to be able to be an artisan and create stuff. I mean, I kind of joke a little bit but I think if you maybe not in the long, long term future, but certainly in the the sort of medium term future I would think that if you, for example, are able to do a trade, there's still going to be a job for you because you know, a plumber, for example, may well use an ai robotic assistant to do certain things, but you're not going to have a plumber for a long time where it's just a robot comes in and does it. There's going to be supervision. So, providing that those kind of hand manual things still have a market, wouldn't the same thing apply to learning you know crafts and learning to be able to create stuff? Maybe that's a really good thing that we're able to have those kinds of careers rather than everybody just, you know, going and working in an office.
Julian Fisher:Maybe in the future, you know, you'll hire a boutique accounting firm where they do it all by hand, just because you kind of appreciate the craft.
Julian Fisher:With a quill yeah, exactly A quill and an abacus Quillen and Abacus, yeah. So I think you know, on that kind of question of short term, long term, I had an amazing conversation with a guy from New York State in the US who had been tasked in the late 80s with getting everyone up to speed with technology. Obviously, you know, computers had started to become more common and he was talking to me. You know, we were at this ed tech conference in Shenzhen or something and there was loads of kind of futuristic stuff and he was looking around and he was like you know, the truth is it took us 20 years to get to a point where we felt like we've got most educators in New York State to a basic level where they could use like Microsoft Word, ppt, basic Excel, like that was actually how long it took to get an entire state up to a really basic level of using what is now kind of common technology. So I do think maybe we shouldn't overestimate how quickly this is all going to move.
Julian Fisher:You know, and I think a good example of that too is like you know, um, during covid, you know, around the world you had a year, let's say, um, where students have to study from home. Um, and all this ed tech that was out there right, that's been talked about kind of game changing ed tech, all this online stuff. In the end, what was the tool that actually got used? It was zoom, right. So. So that was kind of like if your ed tech didn't work in a year when everyone was stuck at home and you're kind of proselytizing, saying that online education is the future, I mean, how's it going to work now?
Julian Fisher:So I think, sort of combining those two tails, I think we probably shouldn't overestimate the impact this is going to make because things move very slowly, underestimate the impact this is going to make because things move very slowly, and I think that's true in education perhaps more than any other industry. So it might not be this generation of teachers that is kind of changing things. It may be the kids they're teaching right now. You know, it's a bit of an oil tanker when it comes to moving. We still haven't really progressed much from kind of you know, hundreds of years ago ago, right with teachers at the front of a classroom talking to a bunch of kids.
Matt Cartwright:so yeah, I I think we should probably be quite careful about how dramatic we think this is going to be yeah, I, I think that's true for a lot of sectors, but I think you're right, particularly for education, it's slow moving and and also I think you know parents don't necessarily want this change. I think that's an important thing. You know, there's a lot of stuff with ai where, where we ask this question is who's asked for this? Does anyone want this? Is it being done to us? Are we actually, you know, are we in any way? Do we have a choice in this? And I think the answer is no.
Matt Cartwright:With a lot of it, um, obviously, but with education, you know, parents have a say. Uh, we did a another episode on with a lot of it, um, obviously, but with education, you know, parents have a say. Uh, we did a another episode on education a couple of weeks ago and and eddie that we interviewed was talking about his own experience in a school of it was carrying out I think it was carrying out calculations with a, with a calculator, and the parents saying, well, we want our kids to be doing it on a piece of paper with a pen and doing it manually, and the teacher saying, well, we're not not teaching the maths, but these particular things do you ever use your? You know, do you ever type or do you ever write this out, or do you just type it out and the parent well, yes, we type it out. Well, okay, so why do we need to teach that?
Matt Cartwright:Now I'm not giving my opinion on whether or not we should be teaching maths in that way, but it's. You can see that there is very strongly a view from parents that they still want people to learn the skills and the things that they used to learn and probably want them to learn from a person, which, I mean, is a good thing. But it does show that it's not just about the kind of capacity in the industry to embrace the change, but also about whether your customers actually want that change. I think in education you're right, it's particularly pertinent in that industry.
Julian Fisher:I mean, I mean, I think one of the things that I do now when I do a lot of training and kind of working with organizations is we need to understand that this word education does not have the same meaning, you know, between you and I, but especially between cultures and between different industries, and you know different stakeholders. You know what is. Education is pretty fundamental to this discussion. I thought you know we would sort of get to this at some point. It's like is it conveying information or conveying knowledge? Is it a kind of preparing for the workforce, control, national identity? Is it about well-being, kind of individuals flourishing? You know, I think addressing that is quite important. And you know you sort of mentioned parents there and I think, as a parent myself, that's really really fundamental. Right? I'm 42 years old.
Julian Fisher:What is it that I want from my life now? Because surely if I can understand where I've got to, I can sort of want that for my kid as well. And you know it's things like having positive relationships, making things happen, kind of knowing what makes me happy. You know that's what I want for my son too, and I'm not entirely sure that AI has a role to play in all those different elements. Positive relationships probably, not so much Making things happen yeah, probably, you know. I think that's something where AI can kind of empower stuff and make it more efficient, knowing what makes him happy.
Julian Fisher:I don't know. I don't know what role AI I mean identifying things perhaps, but I'm not sure what role it has to play there. I think probably that's the question here for parents is what is it you do you want for your kids long term? Because if you focus on the short term, which I think is very common in East Asia where we live, you know you are going to focus on test results, exams, kind of university destinations, and probably AI does have more of a role to play there as a kind of expedient. But I worry that if you just use it for jumping through hoops, ultimately you might reach the age of 42 and be quite unfulfilled and not have that many positive relationships and kind of wonder what, what real positive value it had for you in the beginning the contradiction here as well, I think, is I think most of us see that that kind of you know happiness, in kind of inverted commas, is in many ways is about how much time can we get away from screens and away from technology.
Matt Cartwright:And so the contradiction is, you know, we need to educate people to be able to use this, this technology, to embrace it and to get ahead on it.
Matt Cartwright:But actually, if we're looking at their you know, their well-being and their future happiness, we need to ensure that they have as much time where they're not involved in technology and where they are, you know, with the other world, the kind of real world, and interacting with people.
Matt Cartwright:So I I wonder, in a way, if ai is going to do all this stuff and let's be honest, it is whether it's in 10 years or 20 years, in the kind of longer term it's going to do all this stuff. Actually, how much of it do we need to teach, because we're not going to be involved in that process at that point. Ai is just going to do that. So maybe there's an opportunity, like you say, to spend more time on on learning life skills. Things are going to make us happy, things are going to make us kind of grounded and and better people. And I don't know how the sector kind of takes that on, but but maybe, trying to look at a positive, there is an opportunity to say, okay, well, if these things going to be done by AI, we can cut out these kind of less interesting things and we can focus on things that are better for our long term well-being.
Julian Fisher:Yeah, I mean what you said there at technology. That's probably my biggest concern here is that we have unleashed some things on the world in the last kind of 20, 30 years and kind of just, I guess, just because we could and because there was a kind of commercial imperative, and I'm not sure that a lot of them have worked out really positively. I think social media is a really obvious example. We're all completely bound to it. You know, I spend hours a day, kind of you know, sending funny messages and looking at memes and all this kind of stuff, but I don't think it brings me much value. And I also think that we've seen with teenagers, particularly especially teenage girls, just the impact social media has had on anxiety and depression and all that kind of stuff.
Julian Fisher:I think things like mobile phones again, I'm really unsure about what to do with my own child and that technology Because you know, in the country I live in it's like you look around in restaurants and every single person is on their phone, you know, and it just feels a bit worrying.
Julian Fisher:So I feel a little bit the same about AI, like you know. You know I'm just I'm willing to hold back a little bit. You know this is with my own family, but also with with with schools too, because I think we need to kind of see where this all pans out. And maybe just because we can, it doesn't mean it's good. You know, I think you're certainly seeing that with movements now. You know in the UK there's discussion about banning mobile phones for under 16s, and you know there's a huge amount of parents who are really supportive of this because they feel like this technology has just been unleashed. It's kind of ruined their childhood and kind of ruined their development and has made them kind of really uninterested in doing anything other than playing Candy Crush. So I kind of I do have a wariness about this and a worry too that we are unleashing something without fully understanding the repercussions so what can parents and and students be doing to kind of get ready?
Matt Cartwright:I mean, we you know it's difficult, we don't know what get ready for what. But I mean, are the things that at the moment that you think that that students and parents should be doing to get their kids ready for the the kind of new future of education that they're going to be entering, or that they're already, I guess, for a lot of them are already in?
Julian Fisher:I mean, there's a lot of buzzwords in education and you know there's a lot of different ways you could take this, I think. Probably a word and I say this just because I had a good friend with a good lunch the other day with a friend and we were talking about this for our own kids I think that concept of resilience is quite a big one because ultimately, what phones and the internet and um ai offer is kind of convenience, speed, efficiency. You know, if you're constantly just seeking that at all levels, you know delivery, food and kind of whatever it might be, you know television on the go whenever you want. Possibly that stops you from doing more difficult things, because difficult things it requires a lot of time, it requires failure. You know, and I think that's probably something to consider now, how you develop resilience is really, really difficult, you know, because it's it's something that schools will talk about. Okay, you know we develop resilience, but it's like how do you do that? You know it's not, it's not something you can aim to do because by its nature it requires some resistance and it requires some challenge.
Julian Fisher:Um, I do think it's about being persistent. Um, you know, from a school side, it's things like doing regular sports and kind of making sure that kids are engaged with it, even if they're not that sporty, that they're kind of going a little bit out the outside of their comfort zone, and I think likewise with parents. You know we're up at the great wall at the moment, um in beijing, and we've been here for the last 10 years, 10 days or. My kid has with with the family and you know he's climbed up to the great wall 10 days in a row every morning for about you know two to three hours.
Julian Fisher:And you know that's no screens, no technology, um, you know moaning at the beginning but now something that he kind of skips up the wall and lots of storytelling, coming up with ideas, kind of playing with sticks and stuff. So yeah, maybe it is that kind of reversion to nature and kind of, you know, away from technology. I'm always kind of reminded of, um, the ordless huxley book. You know the brave new world where essentially they kind of break out of the, of the kind of strictures they're in, just go off and let let out a primal scream in nature.
Matt Cartwright:I mean that's kind of where we've got to a little bit as a culture there's a school of thought, isn't there, that the skill that I mean this is kind of, I think, what you're trying to say actually but that the skill that is going to be the most useful for people is an ability to be able to sit and do something for three, four, five, six, eight, ten hours in a row without being distracted. And I, when I first read about this, I kind of did a bit of an experiment on myself to put my phone at my side and sit and read something and see how long it was, until I felt that the twitch where you go to check the phone or check have a quick look at something else, and it wasn't long and I was able to fight against it. But I was. You know, I was having to fight against it and I definitely noticed that. You know the films that I watched, the books that I read, the amount of the articles that I read, you know becoming. Oh well, I'll just read that quicker one, because I can't. I can't sit for 20 minutes and read one thing. You know I need to flick between things, and so the ability to do that I don't know how we educate people to do it.
Matt Cartwright:I think that is resilience. It is persistence that's going to be a skill that, if everybody is, you know, and let's be honest, our brains have been rewired by social media. That that's you know. It's not just that we have been influenced, they literally have rewired the way our brains work. We need that kind of dopamine hit of a, a fast tiktok video or a, you know, a quick article, something that you can read in a few seconds, being able to not rebel against it, but being able to, when you need to, to sit down and focus on something for a long period of time. While ai will do a lot of things, there's definitely going to be a space for people that can do that, and I think that if you can get your kid to do that, that's maybe one way that you can.
Julian Fisher:You can get them ahead in the world I mean, we, we went to vietnam on holiday over spring festival this year in china, and for nine days we decided not to use our mobile phones. And I will say something that was reassuring, which was like after the sort of third or fourth day I was able to sit down with a book and read for like two hours straight without kind of constantly reaching for my phone, which I honestly don't think is something I've been able to do for kind of 10 years. So I think it is reassuring that we can kind of get this quality back. I do think, you know, on that kind of persistence and sort of resilience piece, it's sort of easy to sound like a grumpy old man, you know, and I'm sure every generation has done this, you know, throughout history. But you know, you do see, I hire a lot of graduates in my company and you do see this trait, that kind of when encountering something difficult or when having a difficult conversation, the kind of maybe something I've sensed more and more is this kind of idea well then I'll just stop trying or then I'll give up. You know, and again, I don't know whether that's something that kind of generationally has been said for a long time, but I kind of sense that it's come from this kind of quick wins idea that you know when you play a phone game you're getting that dopamine hit every five minutes and you know if you're constantly at a point where it's like you know.
Julian Fisher:And I have to say I'm slightly worried about AI in the workplace where there are times when, especially in the industry, we're in doing consultancy and market intelligence, where you really really need to have attention to detail because you know businesses and governments and universities are going to make strategic decisions based on the information you provide them with. I do worry sometimes, you know, if I've got staff members who are using AI. You know it's a shortcut, it allows them to do this thing quicker. But like, is the information you're getting really really high quality or not? And if you're just going to use it without thinking, you know, sort of to bring it back to AI? That's one of my concerns. Is that kind of bad data in, bad data out problem, and I'm certainly quite wary about using it kind of completely freely in the office it's not, but I think you know that that's one of those examples where we're looking at the now you know, and it's not.
Matt Cartwright:It's not at that point, but it will be, and I think at some point you'll be using a tool that is, you know, let's, let's, let's forget kind of Asia, I think.
Matt Cartwright:I think the all the folks around artificial general intelligence is very problematic because it's not. It's not actually what we need. What we need is specific tools that are really really good at their job. So, if you're looking to do market research and you have a tool which is better because, like we say, you know people make mistakes, the machine makes mistakes we use a person at the moment to check the work of the machine, well, at some point a machine will check the work of a machine. So I think that quality will get better in time and hopefully, what we see is less of a function on this idea of you know, an agi or an ai that can do everything, and more of a development of specialized tools that are able to do one particular task really really well. I think that that that would be a you know, that would be a good thing and and some of that work, I mean it will affect jobs, but it's sort of an inevitability.
Julian Fisher:I mean that that's gonna happen anyway and that's been really noticeable with the tool that we're building that utilizes AI, you know. So, as I said, it's a sort of tool for comparing schools. You can ask chat GPT to, but of course ChatGPT is basing it on a kind of corpus of data from the existing internet and actually, you know, even just getting a list of every single international school that's currently operating in a country, that in itself is a challenge, and there is no good one online. So even building that, that feels to us now that's really core to what we do is that has to be really good and really accurate. Do is that has to be really good and really accurate. But then what we've noticed then is, you know, sort of scraping the data and kind of analyzing stuff, getting that accuracy level up to, you know, really really high 90s. That's taken months and months of training. You know huge amounts of work using our team to kind of train data, and still you know it kind of.
Julian Fisher:You know misunderstand certain things. You know, maybe, the address of a school. You know there'll be two addresses on the website. There'll be the kind of office address and then there'll be the school address. You know they're different and picking up that subtlety.
Julian Fisher:So I do think still we have to be wary that no matter how effective AI gets, it can only work off of the inputs that it's given if those inputs are human or kind of real world in nature for example, where is a school situated and you know, or you're talking about kind of parents and their thoughts. If everything's just being sucked up from the internet, um, that doesn't necessarily mean it's going to get any better. In fact, I worry that it's going to get to a point where 50 of the data on the internet is going to be AI generated and then it's just going to be this kind of horrific circle where things just get deteriorated. So I still have faith that there will be value for all of us in being able to kind of accumulate and sort of build real world information that gets then fed into the kind of digital infrastructure that still has to be accurate and good, otherwise everything else fails above it.
Matt Cartwright:I don't know if you know the sort of theory around this, this dead internet theory, which is essentially that the I think it's that sort of 50 as I think that was example you just used or 51 of the information on the internet is actually. It's not real. It's either you know, bot or ai generated. And so therefore, because you know more of the data is being, you know it's like 50 of the data was in the world was generated in the last two years, 90 in the last four. You know, as we get more and the data is being, you know it's like 50% of data was in the world was generated in the last two years, 90% in the last four. You know, as we get more and more data, the percentage creeps up.
Matt Cartwright:I actually did a bit of research into this and apparently and this is again you know this is one source, but around 10 to 15% of the data on the internet at the moment is either you know AI or kind of bot farm generated. But 30% of the current traffic on the internet is AI or bot. So you can see from that, although it's not necessarily quite as high as we maybe worry at the moment, the percentage of traffic is 30%, as you you know, exponentially kind of scale that out over time. We're probably not far from the point where it is 50%. It may well be within the next year or two.
Julian Fisher:So there's this amazing. It's literally just one paragraph. I'm going to recommend everyone to read it. It's called On the Exactitude of Science and it's by Borges, that famous Spanish writer.
Matt Cartwright:So we'll link this in the notes for anyone listening. We'll link this in the show the notes for anyone listening. We'll link this in the show notes it's phenomenal.
Julian Fisher:You can read it in two minutes. It's basically a short story about this empire where they demand bigger and bigger maps and they get to a point where they build this map that's one-to-one in scale, and then, as kind of time passes, you know the map tatters, and then underneath, when they kind of, you know, look underneath the map tatters and then underneath, when they look underneath the map, there's just desert. And it's a really classic piece of writing for describing postmodernism, but I think it's very applicable here too. Which is like, at a certain point, yeah, if we look under the internet and realise that underneath is just desert or just more AI, you're still going to need people out there gathering real world information and kind of. Yeah, that's my worry at the moment for students actually, and just in general, is that this you know we've always struggled with that in education. So you know I gave that example earlier of kind of training.
Julian Fisher:You know people to use basic technology taking 20 years in New York State. You know, the truth is, I've worked in education for 20 years and you know you have to train kids how to use the internet. You would think that they would be more computer literate. Right, they're digital natives? They're not at all. That's a complete myth. You know, you watch them. Go to Google and they'll type in you know the question they want the answer to and then they'll just take the first link. So that's why so much of school training around internet has to be like how credible is your source? You know what? What might be the reason that somebody's written this? You know, essentially getting them to understand that. You know there's a big difference between what fox news might say about something and the guardian might say about something. Kind of understanding that nuance.
Julian Fisher:And I worry that you that you know, if you know, again, you've got to be training kids at the moment, not on on how to use AI, but on how AI works. Because if they understand, oh OK. So it's just being trained on the corpus of the Internet with certain weighting and blah, blah, blah. That's going to help to inform them that, like you know, there isn't necessarily AI is not going to give them the right answer on something. It's going to give them a right answer that's kind of informed, with the same level of biases as everything else that's produced in human history.
Julian Fisher:So I do think that's probably where we have to lean to at this stage is setting parameters. You know when can you use it. Can you use it to help write this essay at school or not? That's probably the most fundamental thing. And then I think the second thing, too, is just information and training about how ai works, because I can say with my ai guy, he's really good at informing me exactly how it works and I have a lot of light bulb moments where I'm like ah okay, I misunderstood, I thought it was doing this. Now I understand it's doing this.
Matt Cartwright:That actually helps me to use it more effectively I bought my my five year old daughter, a book called neural networks Networks for Babies and it shows you how a neural network works. And it was quite funny because we'd been reading a book about how the brain works and it was showing neurons. And then the book about the neural network kind of showed neurons and then a kind of digital comparison and she was like, oh, that's like the thing in the brain and I thought this is brilliant for her to understand it, because she knows what chat GPT is, she knows what Claude is, but you know to her it's kind of magical and I've tried to explain that it's a massive computer and and trying to get her to understand what it is. Because, like you say, understanding how something works, you've got very young kids who know how to swipe things on a screen and they know how the kind of physical side of a phone works and they can. You know they can find stuff and you know I've seen examples of really young kids buying stuff on amazon. You know because they're able to do it, but they don't know how it works. They're just able to carry out a kind of process and without understanding how it works, like you say, you don't understand the limitations of it.
Matt Cartwright:A friend of mine, who is, you know, pretty well educated, asked me recently if I could ask a large language model when a certain war in uh, that that may have been particular interest to us was going to happen. And I said it's not a kind of super intelligent life form, it's just trained on things that have already happened and you know they. They just were not thinking about it. To them it was really, really clever. So let's just ask it a question about when something is going to happen, and so you know.
Matt Cartwright:If adults have got that misunderstanding, you can't expect kids to understand how it works. And there is definitely a strong case for overestimating how clever a large language model is because it appears clever in the way that it answers you, but when you understand that it is basically, you know, finding combinations of the most likely thing and putting them after each other, then you kind of say, oh, okay, right, well, you know, maybe it's not as intelligent as I thought. And I've said you know, I think super intelligence will happen at some point, but it's not going to be a large language model. A language model is is, you know, able to do certain things? Well, but it's not. It's not the answer to this kind of super intelligence question yeah.
Julian Fisher:So just one thing on that that I, you know, as I mentioned, with the platform that we're building you know, we've had those discussions about where this could go in the future and you know what could it predict? Because obviously, if we're looking at, you know, a thousand or five thousand schools across asia and let's say five of them close, could we then start to actually understand, you know, what were they saying in social media and what were they language were they using before they closed? That might hint at schools that are struggling in the future. You know, could we use it from patterns in the past to kind of predict the future? You know, and this is a kind of interesting thought experiment. And then what would we do with that information? You know these schools are doing really well and these schools are doing poorly. But I think also, then what it leans to is, as soon as you kind of identify those traits, of course, then people can just modify their language to kind of mimic. You know what the ai picks up a success.
Julian Fisher:So I do think that's also something that is really interesting here is is are we going to enter into a phase where there's going to be an entire industry and kind of countering AI and that there will be a kind of thing is like you know, you see that already with students that they now deploy.
Julian Fisher:They will tell you know, chat, gpt, write me an essay, but make sure that it contains 10% mistakes, because they don't want to be found out by their teacher. You know, and you could, I think there's also apps online allow you to put in an ai essay and then it will kind of mess it up enough that it won't be checkable by an ai checker, which is what some schools are now using to see if the work is produced by students. And that's kind of funny, kind of considering we might enter into a future where, you know, we all end up kind of you know sort of with contrived acts of kind of um, you know, uh, sort of absurdity or just kind of difference, just to make ourselves not fit into that cv. You know, checking ai thing and it's going to. You know, this whole cat and mouse game is going to be really interesting.
Matt Cartwright:I mean I think you know just to finish that. So yeah, we're going to have to change probably the way that we assess um people's performance. So you know the traditional way of writing essays. Maybe that's going to have to change and maybe more kind of you know, in-person exams, things where you can actually see people producing the work. Maybe that's the direction that we move in.
Julian Fisher:Yeah, the big push is portfolios. It seems that that's where it's going to go. A lot of US universities are talking about that now, saying we want to see you know a portfolio of two years work or three years work, because this notion of you know a portfolio of two years work or three years work, because this notion of you know a personal statement or a final exam, is just increasingly absurd. So, yeah, it does feel that that might be where we're heading is portfolios rather than you know high stakes examinations we have talked about this a bit.
Matt Cartwright:I mean, we've kind of touched on jobs and the impact, but I'm wondering if you can look at this from the perspective of your own opinion, but also people that you speak to within the sector. Are people worried about their jobs? Are people in the education sector worried that AI is going to take away their jobs, or do they really see it as you know?
Julian Fisher:this is going to make our jobs more interesting or better so you touched on this earlier but um pearson released uh released a document I think it was 2017, so this is sort of pre-ai era, but it was called the future of skills where they basically looked at what are the skills that different jobs require and which ones are going to kind of be made redundant and which ones will remain. And essentially, the number one job out of every job that they could identify that would exist I think it was by 2030, was teacher. So I have to say I think if you're a teacher, you feel pretty bulletproof at the moment because it feels like we're a long way from schools being replaced with robots and screens. Because, I mean, let's be clear, you know, adaptive learning already exists. If you want, you can go online and find a maths platform where a kid can sit down. They do a question. If they get it right or wrong, it adapts. So in practice, why do you have school? You could just sit a kid in front of that and they could kind of get better at maths.
Julian Fisher:So I do think that teachers do feel that AI is actually more of a benefit to them because it can help with reports and homework and all that kind of stuff. At the moment. I think they probably have doubts or concerns about how the students are using it, but, yeah, they're not worried that they're going to lose their job anytime soon. I think where you're going to see probably opportunity for AI in education, rather than a challenge, is just where there's scarcity. So you know, an example that I can think of.
Julian Fisher:I don't know what's happening in this space, but you know, if you look at a lot of schools mental well-being school counsellors you know they're hard to train, it's expensive. Most schools don't have that resource. You know, do we get to a point where kind of counselling chatbots and I never would have said this five, five years ago but do we get to a point where kind of counseling chatbots actually have quite a lot of value? Because if you look at the counseling and coaching process, a lot of it's about questions. Actually, it's about kind of, you know, asking the right questions and and and and.
Julian Fisher:The part of the process that is transformative for people is being able to talk through their problems and kind of understand where things are coming from you know, is that something where, you know, would I have a problem if my kid kind of you know, had a kind of counseling chat bot that he that he talked about problems? Probably not. So I think that's that might be an area. And I also think, looking at kind of big data and kind of analyzing things and getting trying to kind of problem spot, you know, if you're looking at kids who are maybe a bit troubled, you know, are there patterns that they follow where we can kind of make interventions earlier? So I think there are opportunities here actually for AI and the kind of invention of new roles around that, as opposed to teachers losing their job. So I guess I'm here as a bit of a positive one compared to maybe some sectors where they look at mass redundancies.
Matt Cartwright:Yeah, I mean some of the research I did on this. So teachers was one of them. The actual top two jobs that were the least threatened social worker, I think, was number one and counseling was number two. I think teaching was a little bit further down on where I saw it. But the counseling example is really good.
Matt Cartwright:I'm really torn because I've done counseling in, in person, I've done counseling online with a person, and they are very different. But the one thing that I found with the doing it online was actually, if this is effective, then why do I need a person? Because I don't have that one-to-one interaction where I'm in a room with a human, with the kind of you know, the warmth and the kind of not fuzziness, but you know that, that feeling of being there with a, with a person. I'm speaking to a screen. Do I care if it's an avatar, do I care if it's a, a human being, as long as they're good at asking the right questions. Because, like you say, the process is not really about. A lot of people, when they start counseling, are disappointed because it doesn't give them the answers. But actually what you realise is it's just you talking and going. Oh, actually, yeah, I hadn't realised I thought like that and going through those processes and thinking about it as kind of companions and as not necessarily kind of formal counseling, but to come home from a day's school or a day's work and to talk to it and talk through what you've been doing that day, and you know it'll just say, oh, yeah, that's, oh, that's really terrible. Yeah, how do you feel about that? And they find a lot of, you know, a lot of peace and a lot of help in that and that's going to get better.
Matt Cartwright:You know, especially if you get your own tailored, you can use character. Ai is a platform that allows you to choose different kind of ai characters that have different personalities that will respond in different ways. So you can find that kind of thing. Once you tailor one to your way of thinking as a parent, I kind of feel like, oh god, do I want my pet, my child, to be? You know, having their online fake friend, but actually, you know, if it solves a loneliness problem, is it a bad thing? I think you know there's a balance there, like we said before about, you know, doing outdoor activity, interacting with real people, but if we accept that ai is going to be here, it feels like that is a potentially a fairly positive use case for ai yeah, yeah, it is really interesting, I think that for all of us, that when we thought about AI five years ago, I don't think we understood the sectors that it could disrupt fully.
Julian Fisher:We thought about robots didn't we?
Julian Fisher:let's be honest, we were imagining robots and automation, and I think it's interesting that things with counseling, where, when you actually understand the process, you realize it is, it is quite replicable and it's's really really based on patterns and so in that sense, ai is actually really, really effective at it.
Julian Fisher:I mean, I do have a sort of thought experiment, which is I'm pretty convinced that within 10 years we're going to see holographic stand-up comedians. Now, my logic for this is that you would say comedy is the kind of last bastion of humans, but actually I'm pretty sure if you just plug in you know 10,000 comedy performances get to understand timing, what it's doing with language. Actually it's quite replicable. So it is interesting that those areas that previously we thought were completely off limits for AI and technology now, in a way, you realise well, if they're based on pattern and they're based on kind of repetition, and that's actually sort of something that you can acquire through kind of absorbing, you know, a huge body of similar material, then yeah, things like comedy, music, film, literature, you know, those do seem eminently pregnant well, anthropics Claude, which is the kind of large language model that we we prefer on this podcast, already has a kind of sense of humor.
Matt Cartwright:Now, it's not hilarious, but it's a sense of humor that's translatable. So, you know, if that's where we are now, I, I definitely think within 10 years. Whether people want to see it or not, I guess, is a different question. But, yeah, I, I don't see any reason why it's not replicable. Replicatable, okay, we always try and finish off now with, uh, just a chance for a kind of introduction to any, you know, tools that you recommend or anything that you recommend that people would take on. So if you're a doomer, like me, maybe you want to increase people's awareness or get them involved in advocating. If you're an optimist, maybe you want to give them a really interesting use case, but just an opportunity, I guess, just for listeners, to give them any advice on anything that that you would recommend that they uh, that they get involved in or use um, I'm I'm not going to talk about ai tools.
Julian Fisher:I think what I'll do is I'll just recommend, uh, that that short, short thing it really is just, you know, two minute read on the exact, on the exactitude of science, by borges. Um, I'm also going to say that if, if anyone hasn't re-watched the steven spielberg movie ai from 2001, um, I think it's worth re-watching, I think it's actually a great film and was probably a bit underrated at the time. But beyond that, yeah, I'm afraid I don't have any any, any kind of recommendations. I mean, I can tell you that a lot of people in the education sector use something called magic AI, which essentially, I really just think is mapped onto chat GPT but allows them to do the common, common things that they want building presentations, you know, building lesson plans, you know setting up tests and that kind of stuff, so that that seems to be the sort of most used tool, I think, in the education sector k 12, at the moment is magic AI and then chat GPT as well.
Matt Cartwright:But yeah, outside of that, I would just stick to literature, movies, and let's all wait and see what happens, stay away maybe your advice can be no, stay away from ai and spend time with nature and reading and and listening to real music yeah, there we go, that's.
Julian Fisher:My final recommendation is uh, come to the great wall, stay away from the touristy parts and go for some hikes brilliant.
Matt Cartwright:That has been absolutely fascinating, genuinely one of the best interviews we've had um. I hope you'll come on again. Maybe in six months time we can do this again and see where we are and see how much has changed and whether any of your predictions have come true cool.
Julian Fisher:Thank you so much, super interesting brilliant.
Matt Cartwright:Thank you, julian, and so that's it for this week. We will leave you, as usual, with our song and keep listening, keep subscribing and, yep, see you all soon schools are transforming.
Speaker 3:It's time for evolution. No more just sitting. It's time to stand tall, building resilience, answering the call. It's time to stand tall, building resilience, answering the call. We're a factory of resilience, forging strength with deep fires. Love's a plan, but never break. That's the future we'll make.
Speaker 3:Wise Julian says let them play in the grass. Nature's classroom, where real lessons pass Scrapes and laughter. That's how we grow. Resilience blooms where the wildflowers show. We're a factory of resilience, forging strength with defiance. Learn to bend, but never break. That's the future we'll make. Reforge is worse. A map and the land when knowledge overwhelms we must understand. Life's not exact, it's messy, and real Resilience teaches us how to deal. Read Borges' words A map and the land when knowledge overwhelms we must understand. Life's not exact, it's messy, and real Resilience teaches us how to deal. We're a factory of resilience, forging strength with defiance. Learn to bend, but never break. That's the future we'll make. From school halls to wide open spaces. We're building resilience in all places. The future's bright, just wait and see. Factories of Resilience is what we'll be. From school halls to wide open spaces, we're building resilience in all places. The future's bright, just wait and see. Factories of resilience is what we'll be. Thank you.
