Due to difficulties in our audio filming of the podcast we were unable to upload an audio form video podcast for you, however, we took the time of writing a transcript of the podcast to share with you.

Podcast: The significance of the rise of Artificial Intelligence (AI)

Guests: Steven O’Rourke (Economics expert at Haileybury Almaty), Joshua Liversidge (Humanities expert at Haileybury Almaty), Jon Lawrence (Humanities expert at Haileybury Almaty).

Hosts: Arsen Ashlyayev, Yasmin Uzykanova

Transcript:

Arsen: Gentlemen, thank you all very much for joining us today for an interesting conversation revolving around the topic of Artificial Intelligence and what does a rise in such technology mean for the rest of the world and education in particular.

Yasmin: Our first question to you is, what are some of the potential problems that teachers face following the increase of AI usage in the classroom?

Liversidge: I think the issues don’t lie particularly in the teachers’ use of AI, it is with the students preceding and going ahead, or moving faster with technology than the teachers themselves. The teachers are playing “catch up” at the moment until we fully know how to use it (AI), embrace it, and actually monitor and check for the students using it too.

Lawrence: This is what I actually said to the school we should do - an interview panel with students. One should be anonymous, one should be open to see what people are actually doing with AI, because there are people I know in my Debates class who are using it to rephrase things, and there are also people who are using it to just generate stuff which is in some bizarre realm of plagiarism that no one is really sure what it is anymore. Before we know the full extent of the problem we actually need to know what’s happening, and currently we don’t. I think the big “red flag” that comes from the top of the school is that people will use AI to “cheat” in their essays or homework tasks.

Liversidge: I think there isn’t an issue with people using it in lessons, I had a Year 10 student this morning using it to formulate notes which is absolutely fine, but homeworks somewhat become a little obsolete, because students can now access whatever they need to access at home, copy it and it is done in two minutes. That is probably the biggest issue.

Lawrence: In an infinite number of permutations, because you could just copy things from Google before, but now you can keep rephrasing until it never flags, and then it becomes very difficult (to monitor use). I think another issue with younger students in years 7, 8, 9, I had a student the other day, and she used it to generate some portion of her research report, but what happened was when she asked Chat GPT, it spits out an answer that isn’t at her level. So there is also, just like with teaching students how to Google search, you now need to teach students how to write the correct the right prompt into Chat GPT in order to generate something you are fairly sure is right, and it is also at your level of reading, and that is also probably not going to generate false information, because there are issues with that in subjects like History depending to what Chat GPT is reading to generate the response.

Liversidge: As a teacher, I can understand what it is spitting out, but I don’t necessarily at the moment know how to tell the students how to get what they want at the right level. They can get anything they want from there, but to actually target it at the right style, right length, right detail, I think is a little bit challenging.

O’Rourke: I think another thing as well is in terms of citations. I’ve played around with it myself, and asked Chat GPT to create an article and give citations for that article. Sometimes those citations are fake. They are not actually proper journals, so that’s another issue as well in terms of how we fact-check what is coming out.

Lawrence: I think that really links to the understanding of how Chat GPT works because, if you ask for citations it might not actually give you them. What it is going, is predicting what a citation should look like in a sentence that looks like the sentence it’s read before. It’s just being aware that you need to be skeptical, and moving back to some of the later questions you’re asking, when you are teaching with AI now that’s a thing. You are teaching students how to use it in a way that is beneficial to them. In a way in which they actually understand how to critically analyze or evaluate, and I think that’s really important what (Liversidge) said, where you’re talking about understanding it as a teacher, but if you are a very low level person in the academic discipline you won’t. I think that’s now the future. It’s like: you need to have the skills to use AI, and know when AI is maybe giving you something that’s not right.

Liversidge: My deficit as a teacher is that I don’t know how to teach the student how to get what they need.

Lawrence: It is really interesting because I was talking to Alex (Science teacher at Haileybury Almaty), and he was saying how you can basically get Chat GPT to write code now. So there is a huge amount of grunt work that might be eliminated, because it can literally just write the code, but what you need to know is how to read the code to debug it, so that skill of checking is going to become more important than the actual production of the thing, which is a weird shift.

Liversidge: A shift in skill, I guess.

Arsen: It is interesting that you have touched on the issues regarding AI, but I would just like to know if you could offer any solutions to the problems that you have been facing?

Liversidge: I have not restricted student access to it. I have encouraged students to use it. It was probably last term really, when Key Stage 4,5 were using prolifically. I was encouraging them to use it for essay preparation, kind of like essay detail, content, structure. It was spitting out some good ideas, and I think it is something we should not be discouraging as a school. The world is going towards AI, and therefore we should be embracing it. I think making mistakes with AI is going to happen for all of us, and we should be encouraging people to take risks with it.

O’Rourke: I think for simple things like creating notes it is really good, and not necessarily for recreating an essay, for example. For example, for compiling the key words for key topics you need to revise for whether it be Economics, Psychology or Geography, it is a time saver and I think it is really useful for that. In terms of the challenges and the solutions, how we as teachers are going to try and solve these issues, in terms of the homeworks for example, it’s possibly going to result in, if you create an essay at home, you will need to be questioned on it in class in terms of the presentation. You present an essay, and then there is a “Q&A” afterwards to check if you actually understand what you have written.

Lawrence: Like a defense, a homework defense.

Liversidge: Knowing what you must have computed to give you. You can ask it to give you something greater than notes, greater than homework, but do you actually know what it is written down?

Lawrence: I think there are two prongs, the immediate prong of like, how are we going to stop the “red flags” of people using it to get around homeworks, so there are discussions of ways you can give homeworks which are not necessarily easy for AI to create. Also giving homeworks which may use skills which aren’t necessarily prone to AI, which is where Debates is becoming really good, because even if you did generate your arguments from AI, Debates is a live event where it’s not about reading something, it is about interacting, thinking quickly, and doing lots of things like that. So I think the immediate prong is: how can you taper off the negative effects of people using it in the current system? The other prong is: going forward, what are we going to do with it? I think we have had some conversations about the school, about moving towards integrated projects where students can actually use everything at their disposal to solve some task. Again mixing it with more digital, Google Classroom, online learning, Chat GPT, how do you make a modern 21st century classroom? When you are saying this thing about using integrated revision notes, you guys create your digital notes on a computer, right? You could feed that through Chat GPT and say, “make flash cards”, “make revision questions”. Then, it has literally created it for you, and now you can use Chat GPT, you’ve made the notes, and it will now give you space for repetition, it will now create multiple-choice questions for you that the teachers don’t have to give you, and it’s only going to help you guys. I think there is lots of attention if it’s embraced, I kind of fear that there is going to be a big, “no no no, everything on paper. Everything right now in the classroom,” because that’s the other way to avoid it, and I would say that’s wrong.

Liversidge: The big shift, I would say, would be in terms of A-Level Geography. What the future looks like, kind of consultation. In 2027, our exams are going digital, so therefore we have four years to really embrace the digital world, to prepare students for writing them (exams). I don’t know what the context is or anything, the format, but they will be online.

Lawrence: That’s no longer to do with AI, but what’s really good about that is, what I noticed was that the students who struggled to write by hand, the reason they get better grades typing isn’t only because it is quicker to type, but it is also because you can reorganize. Whereas, if you write something by hand, you are going to have to keep a mental track of everything you’ve written, and loop back to it. If you’re typing, you can reorganize, shuffle things around, copy paste, delete. I think that level of organization is going to enhance. I think that is really cool.

O’Rourke: That makes sense , as you go on to third level, to University afterwards and with coursework and everything there is going to be digital.

Yasmin: I’m curious to know whether you think there are any negative aspects of this transition to digital examinations, as in my opinion I believe that difficulties will certainly arise in students' abilities to draw graphs for example. Which are particularly vital and necessary in subjects such as Economies and Geography.

Liversidge: I don’t know. I know that they are introducing a skills paper for Geography at A-Level which is looking at data analysis, investigations, and more scientific approaches, kind of like the paper Psychology has at the moment, they are reintroducing it for Geography, but I don’t know how that will work with an online exam.

O’Rourke: Potentially, it is looking at using devices like iPad where you have a stylist. How that would work in terms of monitoring, I am not too sure. Going back to Covid times, the AP exams were written online for Economics, and the questions were changed so that there was no drawing of graphs, instead it was focused on how students interpret the graphs. I do not necessarily agree with that, because it removes the skill of graph drawing which is important. So I would say one solution is using stylisism and iPad.

Lawrence: We were talking the other day about how we make this particular graph on Excel. In the modern day, we wouldn’t hand-draw graphs, so actually if you have a graph-drawing programme, is it more important to teach pencil and ruler, or how to teach the programme? I think we need a bit of both. I remember thinking “how do we make that graph now”, and I wish I knew.

O’Rourke: There is a level of data which we need in order for Excel to formulate a graph.

Liversidge: And they can’t do all the graphs. The problem is with Geography, Henry (Geography teacher at Haileybury Almaty) and I looked at some graphs and we go, “wow, we have not seen that ever, or for years”. We looked online, and you can’t do it on Google Sheets, can’t do it on Excel, because they are just some quirky, Geographic-specific triangular graphs, and you can’t get that on Excel. You can’t get the data for that, can’t get a structure for that. I think we are probably at the moment in education with the greatest shift in how we will teach and how students will learn in the next 5-10 years. It’s a big curve, and there is a lot to change. It's quite exciting.

Lawrence: Just to jump into the negatives that I think I can foresee. One, especially with online exams and AI, creates an accessibility divide and a class divide as well. Who has computers versus who doesn’t, who has internet versus who doesn’t. That will be a big issue moving forward, especially with everything being done online. I worked with a South African colleague in Korea, and he applied for the job in Korea with the one internet connection in his town. So there are going to be issues like that.

I think the other one is, the more you get with AI, the more it eliminates mundane human labor like checking, generating things. You do get this kind of quiet violence of the computer, where the computer makes the decisions. I think where I see this happening here (education) is in a shift towards an AI-marker in exams. Exams will become a lot more mark scheme focused, so that the computer can read your answers and generate a response. If there is no human checking, there is a real capacity for the computer to be like, “computer says no”.

Liversidge: Mark schemes are there as a guide, but are not always necessarily explicit. There are some mark schemes, especially with Cambridge that give “etc.” at the end of the mark scheme. They give you a suggested list of answers, and I think if we do move towards an AI-marking system, we remove that human aspect and actually think that will massively impact attainment for a number of years.

Arsen: With the rise of AI involvement in the everyday lives of people. A rise in structural unemployment can also be expected to be on the horizon. To what extent do you see that rise in unemployment to be, and which jobs in particular will be impacted most? And do you as teachers in any way fear for your jobs, given the prospect of AI potentially filling in for you?

O’Rourke: I guess, with technology comes creative destruction. Yes, some jobs may be obsolete, but that has happened throughout history. When something new comes in ultimately, that will create other types of jobs. Jobs that I don’t know, you don’t know. It’s something for the future. Ultimately with any challenge, come loads of different opportunities. Yes, maybe in the short-term an increase in structural unemployment, but what is going to happen essentially is people are going to develop different skills, different skills will be required, as has happened throughout history. Even recently, the internet for example. I remember when the internet took off, a lot of people looked at it as a bad thing, as a challenge, and yes there are challenges with the internet, you got dark web, etc. With AI I think it’s going to be similar. We’ll learn how to use it, learn how to get the best of it. In terms of teachers, can’t remember when it was, but calculators were seen as a massive threat. Math teachers in the US went on strike and said, “well there’s no point in us being in the classroom anymore”. Nowadays, can you imagine taking a math exam without a calculator?

Lawrence: For me on this point, to sort of go into a philosophical deep-end of this conversation, I think it is good to remove jobs. What is the point of all human civilization if it’s not to remove as many tedious jobs as possible. We are kind of locked in a system which values money, and values like, “do you have a thing that you do?”, which I think is a worse system than we could have. I kind of wish to live in a society where we go, “we’ve deleted this job! Nobody has to do it anymore”, and everyone sort of goes “brilliant. We are going to paint. We are going to go traveling. We are going to literally do whatever it is we want to do.” For me, within the current social framework, yes, we are trapped into this very narrow economic thinking of, “you must have a job,” but I think again society is facing this issue where people are going to be unemployed not because they are unskilled, it’s because there just aren’t any jobs. I think this is going to shift how we start evaluating what a human is in society. I think humans should be the creative producers. I read this off Reddit or Twitter, but it says that we are currently on the verge of living in a hell-scape where AI writes movies and music, and humans do mundane tasks on data entries. This is the wrong way around. We want it this way around: where humans create art and have fun, do theater and do whatever they want, and AI just chugs along in the background doing data entry, analysis, things like that.

Liversidge: I think to link it back to education, there will be chaos before there is calm. People will lose jobs. People will have to switch careers, and society will have to adapt. Education will come to fill that void, and reeducate people in different ways, on a different skill, and different targets for society. First, we’ll have to have that large-gap societal shift with AI influencing what the world will look like in the future, and then education will play catch up. Education will not jump forward to predict AI, because we can’t do that. There will be a big change. A change in education, people will lose jobs, mass unemployment, but then education will feed into AI, and realize how much it can settle in a way.

Lawrence: There is a real potential in education for AI to delete really mundane, tedious stuff. In terms of resource creation, teachers come up with new resources whether it is PowerPoint, booklets, or something for students to interact with. That is technically dead time. If AI can just generate those resources, then you have more time to interface with students, to give feedback, to build relationships, which is the real core of teaching. The less time we can spend on just generating something, the better. With things like report writing, and other things you’re going to start to see AI creeping in, whether people want it or not, it exists now. People will use it. I don’t think teachers will go. Things like calculators, TVs, VHS I’m sure was the end of teachers. Interactive video games as well. A lot of stuff has been the “end of teachers”.

O’Rourke: I think you (Lawrence) touched on a really important point there. The most important part of being a teacher is the relationships you have with your students, the emotional support you can give. I’m pretty confident students can read a textbook, understand the majority of what’s in the textbook. Loads of other things, students will need some support throughout that journey.

Liversidge: A machine can’t replace that. Also the extra things that happen beyond the classroom within schools, all schools have such a wide variety of extra curricular activities such as sports, or events that take place that a machine can’t replace. The school building itself won’t not exist in whatever form it existed. Classrooms will change with AI and what not, but actually building character and holistic education, building young people to become good adults - that can’t be done with a machine.

Arsen: I’ve stumbled across a really interesting piece online, and what it basically said is, from the point of view of world leaders and governments of major countries, they don’t actually see AI as a human-assistance machine, but rather as something so powerful that allows them to develop their resources and become more powerful. Kind of like the Arms Race of the Cold War. Do you see any validity to this, from your point of view

Lawrence: Yeah. I think the kind of threats that AI poses to society are slightly more mundane than how from 2001. AI can dispassionately process facial recognition, that’s one of the major issues. Any bias that a programmer has, AI has too. What you are seeing flag up right now in places that are using this, is bias systems that are identifying people wrongly. I think Israel, Palestine and China have been using this to identify people, and this is very bad. This is not the type of society you want to live in. In terms of going into the general AI that might do all kinds of crazy stuff, I don’t think we’re quite there yet, but that will also be bad.

Liversidge: In education, there will be schools that embrace it much faster, as well as countries and institutions that embrace it much faster than other countries. They will see themselves as having advantage. It’s also a marketing tool. What we will be doing is marketing the fact that we are setting up students for 21st century skills by embracing AI. So if I were to compare a school system to a political system, countries wanting to adopt AI as a force for good or a force for bad, it can still be the same in a micro scale down the school. We’ll be competing more and more with the schools that embrace or do not embrace AI.

Yasmin: Those are some incredibly valuable insights that you have all shared today. Due to time constraints, we will have to conclude our podcast filming on this note. However, I'd like to thank everyone for devoting your time to voice your thoughts on such an important topic at this time in our world. Thank you very much!

Regarding our beloved OhMyEcon listeners, I hope you enjoyed this episode of our podcast. If you have any more questions or would like to express your thoughts on this subject, feel free to email us at ohmyecon@gmail.com or direct message us on Instagram. I'll talk to each of you soon. The end!