#280 - What are Student Expectations for AI in Education?

Welcome to the EdTech Podcast. With roads lucky. is Professor of Learner Centre Design at UCL and founder and CEO of Educate Ventures Research, The Artisans of AI. We help many organizers. To leverage AI to be more impactful and effective. We provide training and professional development for educators and companies and undertake bespoke consultancy to enable organizations to use data and AI.

In today's rapidly evolving educational landscape, artificial intelligence is emerging as a transformative force, offering both opportunities and challenges. As AI technologies continue to advance, it's crucial to examine their impact on student expectations, learning experiences, and institutional strategies. One pressing question is, what do students truly want from AI in education? Are we reflecting on the value of assessment when AI tools have the ability to complete assignments and exams?

This begs the deeper question of what we mean by student success in education and the purpose of knowledge in an AI driven economy. I'll hand over to Rose for today's episode. Now, before we get started, a word from our sponsor, Studiosity. With a mission to increase the life chances of every student in the world, universities partner with Studiosity to support millions of students, ensuring ethical, trusted feedback for success.

Now, with uniquely ethical AI-powered student support for institutions, Studiosity's AI for learning is distinct for its evidence-driven approach to student support. Transparency for educators, including evidence of learning gain, and permanent humans in the loop. With more than 200 institutions choosing Studiosity and seeing firsthand the effectiveness of ethical, scaled, transparent student support, review it for yourself at Studiosity.com.

Today, as we explore AI's potential and its limitations for education, I'm joined by three wonderful guests. I have Michael Larson, Sally Wheeler, and Anne Bagshaw, and I'm going to get each of them to introduce themselves to you as they answer the first question. So let's get going with the first round of questions. And I really want to get down to knowing. What it is you feel is happening with AI in education?

I'm hearing from lots of colleagues, particularly in higher education, that students Expect. AI to be part of their working lives and they therefore anticipate and demand that it's part of their studies. But obviously there are also some tensions and some aspects of AI use that you want to make off-limit. So I'd be really keen to understand a little bit more from each of you about what areas you feel AI can be particularly useful with.

And are there areas where you feel AI might be a bit off limits for students? And I'd love to start with Sally, please, if that's okay. And if you could just introduce yourself a little bit first, Sally, and then tell us your perspective on AI, that would be brilliant. So I'm uh Vice Chancellor of Birkbeck, uh University of London. That's a role I've held since January. Prior to that I had a senior role at the Australian National University.

So I don't think any AI should really be totally off limits for for students. I think that is akin to saying they shouldn't use the internet and they shouldn't use a calculator. It's here to stay. We have to move with it. I also think that there is a huge digital divide and knowledge gap between students. And some of my students are actually frightened of AI. They're frightened of using it in the wrong way. They're f they're frightened of being accused of academic misconduct.

And it's incumbent on us as educators, I think, to try and bridge that divide. I think it has huge possibilities for students who are working in a language that isn't their first language. For students who are trying to swap disciplines and build up a a disciplinary knowledge. And I also think there are knock-on effects of trying to push AI out of the education workspace, which means possibly retreating to forms of assessment.

That prior to AI, a lot of us were beginning to think were not suitable for the modern world. So I hear horror stories about people just saying, well, okay, it's going to be exams from from now on. That is completely inappropriate and we just should not be doing that and not thinking about that. That's music to my ears, Sunny. I I couldn't agree with you more. I think it's really interesting the challenge.

to assessment because there's been a desire to revert back, as you say, to forms of assessment that many of us feel really are not fit for purpose anymore. I think we need to really try and encourage something much more innovative. How do you think one might do that because some people are really quite passionate in their um desire to return to some very old school forms of assessment.

So I think there are other pressures for that, returning to old school form of assessment, which is around student numbers, it's around the ability to get feedback quickly, and actually AI clearly has a role in in feedback to students. But I also think it's part of a bigger pattern where certainly in HE we just over-assess students. So we are looking to replicate a pattern of assessment load that simply is not.

Not feasible, not suitable, and has a huge knock-on effect to students who are in education with any form of learning, uh, physical, social disability. And that's something that we need to bear in mind as well. Education is supposed to be about. uh social mobility and and access, returning to old form assessment and assessment overload, it's just not going to deliver that.

So true. Thank you, Sally. And that really cues me in to ask Mike because you talk about feedback at studiosity and and Sally you know mentioned the importance of feedback there and I do think it is a really interesting use case for AI. But before we get into that could you do a little bit of an introduction and give us a a sort of a general So thoughts on positives, negatives, AI, what's happening?

Sure. And uh we uh at Studiosity provide uh you know online support for for around two million stu uh university students across five continents. So we we have a very large audience. Uh and and look I uh look like Sally, I'm a a a devotee of of AI. in certain applications and contexts. So when ChatGPT first burst onto the scene, the the the most powerful use case was just improving productivity and improving and and for

for workers, for staff, for for people in their daily lives, and and I think that's wonderful. And I think it can be useful in some instances for students. What it doesn't necessarily translate into, though, is AI for learning. Uh and so I think there are still plenty of um powerful applications of AI where students can be better supported, um, where they can get feedback and and and we're seeing this now. We have

you know, uh ten universities in the UK that are using our uh AI powered writing feedback tool and they used to wait six, seven, eight hours to get personalized feedback. Well now it's less than two minutes. And so that actually means that rather than asking just being able to engage with students who were organized enough to have written their draft the day before, the support is available f for for anyone even up until sort of a few minutes before the essays due. So so that means that

personalized feedback you know can be uh can be incredibly powerful. And it also The technology allows us to go beyond what we used to see just in in essay writing structure and choice of language and and referencing into critical thinking. Uh and I think as AI gets more and more powerful for for us as citizens and and you know for employ future employers, what what they're really going to be looking for is is is

Graduates can think and think for themselves. So we'll we'll sort of provide a student with feedback on a piece of text according to Bloom's Taxonomy of Higher Order Thinking and say, hey, you've done a great job of understanding. What you could do is elevate that to analysis. And here's how you might think about doing that. So so look I think there are plenty of wonderful applications. Um I think we should charge ahead but with with with AI for learning as the uh as the primary goal.

Again, very wise words. You're right. The hype around AI makes us think that it can turn us into superhuman. Speed, very productive, et cetera, et cetera. But it doesn't always translate into learning. You know, I've worked in this area for 30 years. And a lot of that time we talked about educational technology rather than AI because people weren't interested in AI. But we had the same problem.

Actually, technology that had been designed for the business world made it into the classroom, but that doesn't necessarily translate into a tool that supports teaching or supports learning. So you you're you're absolutely right. I I I agree with you.

You know, and we do need to make sure that the way AI is used in education is driven by Learning You uh of course it could be used in the back office, but in terms of teaching and learning, it does need to be purpose driven and that purpose needs to be driven by something sound pedagogically like Bloom's Taxonomy or or what edit whatever it is, it really needs to be

integrated into a learning experience. So it's interesting to hear what you're saying there. And I remember, you know, through teaching in schools, further education and higher education, the desire to get fast feedback to students, but the reality of not being able to do it because you've got too many scripts to mark. I think it's a wonderful use.

I think there might be some challenges and I'd be interested to know if you've come across this at all with perhaps some academics feeling that it's part of their role, it's part of their identity to do that feedback. I mean, have you had any any kind of Pushback against it, or most people recognize that actually this is really useful for the student and it saves me time and all of the positives.

Much more the latter. I've never met an academic who wouldn't love to be able to have the time and the bandwidth to provide that personalised feedback. But I mean th as much as there's a student well being crisis on campus

Or we know that staff are really overworked as well. What they'd much rather do is uh or see is students get high quality ethical feedback so that the work that gets submitted is ultimately well written and they can focus on teaching the content rather than kind of needing to say, well

Gee, you your facts are right, but yet you haven't really made a compelling argument here. So yeah, a ac academics are are actually really enthused about it. And it also starts to normalize the help seeking process. when a student can get that first little bit of feedback and look, our our system will recognise perhaps that well, gee, this student needs a lot more. uh a lot more support and and then we can refer the student to on campus support or or another

uh another means of support that the university might have available. That's a really important point, isn't it? It's about breadth of support. I I can completely understand that.

Thank you, Mike. Interesting. I'll come back to some of those issues uh in in a little bit later. But Ant, I'd love to come to you next and especially I'd like you to tell us a bit about your new role. because that's exciting. And then, you know, your perceptions of the the good, bad, the ugly of AI and what it should, shouldn't be doing.

Thanks, Rose. I recently became the Executive Director of the Australian Technology Network of Universities, or ATN, and I represent half a dozen of Australia's most innovative and and enterprising universities in uh

in and advocating for for policy settings and and outcomes that are that are great for those universities. And and obviously this topic is hugely of interest to to that network and we've got a whole slew of examples from across the universities of implementation of of AI tools and so on.

But I want to pick up on a couple of the points that have been raised. Actually really building on that staff point. Because I'm I'm interested and and prior to ATN I've been in consulting and in regulation and in other parts of the higher ed sector. And I'm really interested in in what are the kind of enabling conditions for universities actually changing practice at the coal phase?

of uh of of of learning and teaching. And when you say, you know, what are the you know, should anything be off limits to to students? I I'm I'm completely with Sally, I don't think anything should be off limits, unthinkingly. But we do need to say what are the how are we creating the conditions in which our particularly our academic staff feel confident, able, uh they're it they are deploying those tools within their own practice. in a way that is enabling those learning outcomes for students.

But in and in a way that is not threatening but is enriching of their own academic practice, is nuanced to the discipline. And also conforms to institutional policy. Because I think this is a class, you know, universities all over the world have have you you mentioned it yourself, Rose. We're going to have a twenty one day feedback turnaround or a however whatever we were gonna go with. And trying to roll that out across these large complex institutions.

has just always always kind of come apart somewhere. And I think this is this is the same challenge, the same challenge of how we change hundreds and thousands of people's individual practice. And there are some things that we can probably be done centrally, where there's, you know, and studiosity is a great service that can be applied across, you know, every discipline, all students. But there are equally

Things that need to be nuanced to the individual disciplines. And it's not obvious to me that that every institution is capable. of of operating at all those levels. The top line, strategy, setting the ethical conditions, getting it translated through the organizational structure so that it actually delivered. two students in the classroom, physical or virtual, wherever that may be. It's such an important point and thank you because

You can't bring about change unless the individual people in an organization change their behavior. It's just not possible. Particularly, I think in any business, but you know, I know education uh better than others, and you know, unless a lecturer, a teacher changes their behaviour, then there isn't organisational change. So it's a really interesting point and a really important one. But I'm interested in this enabling conditions for change of practice.

Do you think that the enabling conditions for AI are different to the enabling conditions for other, I mean, it could be other technology comparisons, because I think technology does bring specific um demands for those enabling conditions. But what's your thinking about perhaps any differences that AI brings?

I think there are some peculiar differences. One is I think about some of the unknowns, this idea that within many of, say, the large language models as as this this obviously this where this kind of current enthusiasm is for, is that we do we don't know exactly how they work or what they're trained on and

um and and how they're operating. And I think that that creates uncertainty and then introduces a whole series of other risks and and kind of governance concerns that universities are absolutely rightly concerned about. There's also a speed of change issue. So we can't just wait a year, wait two years, wait three years, see what someone else has done and then pick up the practice, which is very common within the higher ed sector. But to your fundamental point about students,

Our students today need to be prepared for the jobs of today and tomorrow. And so we can't say, Oh no, no, just just hang on, we're going to um not bother implementing this because We can't work it out even though you'll need it. And so I think there is a real urgency to accelerate this learning and implementation loop and the that that organizational learning piece, which I know is something very close to your heart, Rose, you know, including evaluation, including that that that rigor.

I think there's also another dimension here which we you may want to explore. It's actually also the commercial side of this, because there's a lot of Yeah, it's vast amounts of money, Silicon Valley money going behind AI tools, pursuing the education sector as a as a market. And I think one of the risks to institutions is is being supplier led in in some of those implementations and not necessarily having the right sort of

dialogic relationship with a supplier where universities get to shape the tools as well as to um as to implement them. Really important point. I'm very struck by there was a publication recently from the Clima Anthropic.

who are behind the dark language model used for Claude. And it was a really interesting publication. There's been other similar publications that basically do demonstrate that even the people who make these large language models don't really know how they work.

And this was one where, very, very generalized uh summary, by tweaking particular parts of the model, you could produce a result where In answer to questions that didn't mention anything about California or the Golden Gate Bridge, Claude would give an answer that related to the Golden Gate Bridge.

So for example, well what's a really good way to spend$20? Oh well, going across the Gold Cape Bridge or whatever it is, which just shows that the the people who are making these models are still exploring how these models work. And I think that does leave us with a dilemma. And it marks me to come back to, which I think is a really important point. But we have to find ways of helping students to deal with that.

I don't think we can use it as a reason not to get involved because students are going to need to be able to understand, well what does that mean? What does that mean in terms of how I use these models? Because I'm going to be using them in my career most likely. But it makes me want to come back to Sally and something that you said right at the start about fear of AI and the fact that you're finding some students.

are fearful, you know, fearful of getting it wrong. And I know when Ant and I were visiting a university in Australia last year, there were some students who just felt this was unethical and they didn't want to be part of it. So coming back to you, Sally, how how do we deal with these tensions and and and recognise that students might have quite legitimate concerns? And yet at the same time as educators we want to make sure we are preparing them for the world they're going into.

Well, I think we do need clear policy settings. One of the things my students tell me is they just want to know which side of the line they are. I think we also need to to think about how we teach students all the time about ambiguity. But how t how they can tolerate ambiguity, how they deal with uncertainty and ambiguity. That's always been a core part of critical thinking. And we need to absorb AI and the potential of AI into that.

I just wanted to pick up on one of Ant's points about what what it is we're being sold and the I think In some ways, the AI applications give us a chance as universities to be co-producers and co-creators. in a way that we have never had with other things that are that we might have bought, you know, uh tools to help us enroll more students, tools to help us find more students.

online delivery platforms, etc. We've never really, whatever we might think, have been co-creators there. It was this is our model, and if you don't like this, then go and find one that does suit you. Whereas now I think there there's a real possibility and I also think that AI can be very democratic as well. So I see huge roles here for subject communities.

So so AI in history might be very different from AI in one of the science-based subjects or in social science, for example. And so how do we involve grassroots teaching and research staff in AI and AI's potential. And part of it I think is overcoming the fear that this is going to create more work.

The fear that we will do something as an institution and not be able to get it to work and it will just create far more hassle than it than it than it delivers. I think we have to be in a position to say to people, this is a positive. Again, very wise words. And I I'll come back to the critical thinking piece because I want to come back to Mike on that because I know you mentioned it as well. But

Something in in what you were saying there about I agree about the ambiguity, uncertainty and if we can't do that in a university, where on earth can we do that? It seems like the right home for that. Tying back to something Ann was saying, which is also something that concerns me a lot. I mean Last week we had the AI P C. I wonder what price tag that's gonna come.

This week we've had Apple Intelligence on the latest iPhones, another hefty price tag coming with that one. There are lots of little devices, most of them really don't work, that have action models rather than language models. There's a lot of consumer push, there's a lot of commercial push on these products.

And I really worry that it is AI is is coming at education. Education is having AI done to it, if you know what I mean. And we really must avoid that. So your point about co-creation and seeing that opportunity is so important. But isn't it is it not the same with every product? So I remember as a post grad student investing in a one of those Clive Sinclair, MS DOS, what I can't remember, Amstrads.

And people thinking that I I was crazy to do this, and this was a ridiculous amount of money. But what we actually saw was the price and accessibility point of those. coming down rapidly. And I suspect although AI is still evolving, it will be exactly the same with that. I suspect you're right and I I think your points are very well made.

Uh it's just a personal worry, perhaps aimed more at the school sector, because I know that so many schools don't even have an ordinary laptop for their children and now there's this push to have these ever more expensive tools. And that is a point about diversity, which I also want to come back to again, something you mentioned earlier, Sally. But first, I just want to pick up that critical thinking point.

Um Mike, because you mentioned it and said how important it was. Now it seems to me that it's even more important now, isn't it, with AI, because we're having AI being offered as something that can do so many things for us.

But actually sometimes what it does isn't right and it does make mistakes and and there's lots of issues. So Tell me a little bit more about your your why you bring critical thinking in and I agree with you, but I'd love to know a little bit more about where you're coming from with that. Well, i i it really comes down to um being able to operate in a world of increasing ambiguity. A as well as

being able to to thoughtfully question and challenge outputs from from AI. And that's even for for all of us on the productivity side, when you're experienced and educated and you you get something from AI and you go, Oh gee, that doesn't seem that doesn't seem quite right. Um, but students, of course, don't have the benefit of that experience. And and it actually leads me into uh back into to Ant's sort of observation around enabling conditions.

But I think one of the things that uh academics are are appropriately concerned about is when they hear, well, yeah, AI already knows economics. Well, AI does know something about enough economics. Um, but if I'm teaching microeconomics one oh one I've got very specific ideas about what it is, w what are the foundations of what my course should be.

And so I think what will enable uh an increase confidence for academics is when they are then they're when they're able to use AI that has a constrained knowledge base. and they are in the driver's seat as to what um that knowledge base is comprised of. And that I think will also support the continued div you know, richness of diversity that we see, particularly in the in the UK with

uh universities large and small with different styles and and different sort of teaching approaches. I think that's a wonderful thing. And and I think that actually adds uh adds colour and richness to the sector. Yeah, your your point about agency being in the driving seat is very important here because I think it's quite

tempting if you open up any of the large language models rather than an application that uses it. You say, well, what can I do for you? How can I help you? And and it kind of you you could feel like This thing could almost take over. And if you don't know enough, you can easily end up absorbing inaccurate as we've seen, and then use it in a legal case and actually have a real problem as a result.

So i there is a real need to help that agency, isn't there? It it is. And uh and I think it's in the in the higher or higher levels of critical thinking that the real that the real capability building occurs for students and the preparedness for for for a world that i is obviously evolving uh you know much more quickly. now than it was ten years ago and will be even faster ten you know uh ten years uh hence.

So so yes, uh having students be able to think independently and critically as active citizens I think is more important now than ever. It should be, shouldn't it? I agree with you. This episode of the EdTech Podcast is generously sponsored by Studiosity.

Studiosity is uniquely ethical, AI-powered student support for universities for transparent learning outcomes, cost sustainable Equitable access and ethical academic Studiosity partners with universities to deliver trusted, academically relevant and equitable AI. Studios the TS. Pioneer of AI for learning. Ensuring help not on. For learning before Gen AI shortly. and permanently embeds humans in the loop for quality assurance. With more than two hundred. Review it for yourself at studiocity.com.

And coming to you, I I have the whole load of things I'd love to talk to you about. Um, because I think your point about enabling conditions is quite an inst it it's very much at the kind of institution level.

Obviously recognizing that whatever the institution does should be in the best interest of the student, but there is something about that institutional piece, isn't there? And how do we get a greater voice for education in what is happening with AI, you know, a sort of purpose led dialogue dialogue so that we get the right things for education. How do we as a and maybe you know you you are in charge of a network of universities, maybe it's even at that level, how do we get education?

in the right conversations with the people who are developing AI, because at the moment I feel they're just seeing it as some way to make money and and and what's being developed is not optimal. I I I take Sally's point that as time goes by things will will wash out and things will improve. But but how do we get education more actively in the conversation about what AI is available. So I think there are some really good examples of universities and

providers of AI tools working collaboratively together. So within within my network, you've got RMIT in Melbourne working with Adobe on on applied AI tools in in their suite of software and and RMIT has a lot of creative uh discipline so that it's very you know, makes sense so they would work with Adobe in that way.

uh University of Technology Sydney is is doing some co pilot uh pilot uh type type work and so on. So I think those are those make sense. I mean I think to the to the point I made earlier about the sort of the big tech interest, I I think I think institutions need to be to manage those relationships really carefully, so as not to just be giving away IP or just be providing an environment which can then

be exploited, but actually genuinely to make it a a partnership. And I totally understand Sally's point about, you know, in the longer term things would evolve and become more accessible. I think there's just a risk given the sort of economics of this area potentially meaning some consolidation or s you know, some big suppliers sort of having dominance. And I think that's that's always a risk, obviously, in in

in any industry. And then and then you have, you know, and I I know Studiosity well and Studiosity is a very, you know, values led and and mission driven organisation and and I think you know, and and and more focused than some of the you know than than obviously than Microsoft, right? So so you could say the the sort of the the tech available. So how does does a company like Studiosity make sure that that spec the the specificity of that offer is

uh is maintained within the context of what are other things that are happening in the in the AI environment is obviously a a kind of ongoing question. To your to the heart of your question on the primacy of the sort of the what matters, the educational piece. I I think what gives me confidence is that the real enthusiasts, the champions, the people so excited. A lot of Deputy Vice Chancellor's academic.

I haven't met one who's who's who's afraid. You know, there there actually is a group my my experience is the sort of the selection bias of people who go into those roles tends to be the the enthusiast, the the entrepreneur, the the experimenter. And I think the c my question for the back to the sort of conditions piece is how well they're influencing all the other people they need to do within the the senior team. So, you know, the CIO, the

uh the finance director clearly, you know, crucially. And and then obviously, as I mentioned before, within the individual discipline. Thanks. That makes a lot of sense. And and just one point. Thinking back to something Sally said about clear policy setting. What are you seeing in your institutions around policy for AI? Because that can be quite helpful in setting parameters and helping people feel they know.

the space within which it's okay to explore, to get excited and and where the boundaries are. I I think the the best work there is the recognition that you're going to have a policy with probably quite a short shelf life that the tools are developing and so on. And but you've got to have something out there.

So better to say something, have something in place, but then to revise and uh an update and and I think that Often quite difficult in universities where where often sort of policy development and the sort of committee cycle takes about a million years to get something to get anything done. So that that's gonna have to work work quicker. But but I think again back to a sort of point point made earlier, you you can't wait on this one.

No, you can't and it's an interesting point. As a previous pro vice chancellor, I chaired lots of committees and I know how long the cycle took and I know how long it took. And now we've got to be positive. But we have to recognise that we have to be it's be dynamic, isn't it? It's kinda like dynamic policy making because You will need to be revising these things because things are changing and we're learning more and regulation change also has an impact on that. Thank you, Anne.

Two last questions that I want to tackle and and uh look for volunteers to start me off on these. The first one is about evidence and it's come up in the conversation already. There's a lot of technology out there and now there's even more with AI and huge investment being made and lots of things being rolled out at education, in education. And yet at this early stage. There's very little evidence about

Because even if you're using an AI tool that may not cost you money, it's still going to cost you time. And there's obviously the data issue in terms of data rather than money. But my point here is it's if there's a resource implication. beyond the financial cost of any technologies, there's the the time it takes people to learn how it works, to learn what infrastructure needs to be placed, the policy implications.

So how can good decisions be made to make sure that you do get a return on that investment? So I think just generally universities should pull knowledge around things like that very much more than they do. and not necessarily see it as a competitive playing field where they're competing with each other. Because then I think you do get decisions made that aren't always The best?

And and I think within an institution you need broad com you need to facilitate broad conversations about what might be useful and what might not be useful. And then there are all sorts of certainly within the the UK there are things like subject benchmarks. uh for different disciplines. And they can be very useful in terms of getting constituencies based across universities thinking about what interventions should be.

So I I I I know it runs against the grain for some people, but I don't think competition in every space is useful. It actually costs you money. I couldn't agree with you more. And I think you mentioned subject communities before, and it does feel like there's quite a role for subject communities here. You know, maybe the collaboration in that space is perhaps a little more straightforward than at the institution space.

But Ant, you have a network of universities. So presumably that's an area where you can really excel because there is an agreement to share. That that's exactly right. And our particularly our D V C's academic are really already working on that knowledge sharing um and and and good practice because you know, to to Sally's point, it's absolutely it's the same it's the same challenge everywhere.

we're all port for purpose, mission driven institutions where there is not this is not an area of competitive advantage. This is actually and where we ought to learn. And actually I think, being honest. And I think my experience is universities are not always honest when when things that have implemented have have not gone well, you know, not always the kind of a

not necessarily accountability, but just even the the willingness to do the lessons learned. And I think this is an area again with this idea of sort of faster learning cycles where we should understand what what didn't work. I mean I think one of the uh we're still a long way from universities as as the uh the the learning organisation to borrow a phrase from a former colleague of yours, Diana Luriard, that that I think

But but universities are fantastic natural experiments. You know, you've got huge variation in practice, so we don't have to do, you know, A B testing every day. We've actually got that natural experiment, but can we apply the data and the AI tools to the data to to to get that to get that loop? And that and that's where

you know, clearly that's already existed in the sort of learning analytics space for a number of years. But I think that we can really accelerate. And I know Rose, you've done a lot of it really interesting work in in that space. And I think that's something that

um I'd love to see to see more of. That would be great. And it's a real reminder and thank you that It's easy to get hooked up on conversations about generative AI, but there's a whole world of AI beyond generative AI, and some of that is really incredibly useful.

So it's important to remember it. The last question I want to deal with, because sadly we've run out of time too quickly on these podcasts, is one about diversity. Sally, you mentioned very early on about huge opportunities. Students for whom English is a second language, using these tools is amazing, they are good at translating. There's a huge possibility through personalization.

Mike, you've mentioned that, to meet the needs of a wide range of student requirements. So I've always believed for twenty nine of the thirty years I've worked in this space that AI was a wonderful tool. for increasing equality, for meeting the needs of a diverse population of learners. The last 18 months I've been a little concerned because I think at the moment it's rolling out in a way that

Gives people with money a huge advantage. These things always do. I hope we we move away from that. But let's look at the positive side of this. Oh, who'd like to kick me off on talking about well you can talk about the negative if you like, but but the aspects of the possibilities for AI and diversity and you're quick in there. I guess I'm gonna go big picture and speak about a a kind of current policy issue in the Australian context. So

you may know we've had what's sort of universities accord and and within that central to that is this ambition for ri for significant growth in participation in the sector. And it actually comes back to a point Mike raised earlier about productivity.

We are going to have to grow the sector, but we're not going to get the same level of resource. And so we're going to grow and we're going to have a more diverse student cohort because Rich kids have always gone to university and so if we're going to grow w we're going to be more diverse.

And so we need to get better at meeting the diversity meeting every student where they are and giving them a fantastic experience. And we're gonna have to get more financially efficient. I think these two are just two facts. that institutions are going to have to deal with. It's true in the UK as well. We've seen a you know the eroding unit of resource there and and the need that. So I think this is what unlocks the productivity gain and potentially gives us both enhanced quality

and a more efficient model. That's a really interesting take on that question. I think you're right. And and again, that's where the data analytics can be super helpful because if you get your data infrastructure right, you can learn so much about the way your organization is or isn't growing effectively, efficiently, you know, and then you can use the tools in the back office, so to speak, to increase productivity. Um, and that enables

so much more for that greater variety of students. So that's a very interesting perspective. Mike, you sort of brought up the productivity point earlier and you also engaged in a a conversation around diversity and the personalization of feedback to meet the needs of a diverse population of students. Where do you sit on the belief about whether AIB are good thing for diversity and equality or not? And I'm gonna go even uh a little broader than than and from a global perspective.

I mean one of you know I I think a natural concern in developing countries is that they you know countries in in in Africa and South Asia. uh their education systems will fall further. behind um w without access to to quality um AI. And the approach that we're taking is is actually building uh an evidence base. So we have two university studies running in South Asia right now.

one in uh one in Africa, because uh done well and sort of with the right configuration, uh AI can actually be delivered at scale really quite economically. So so I think there's a huge opportunity uh across developing countries to actually raise up not only the education systems, but as a result, the the productivity and and economies and living standards of entire countries. So you can tell I'm incredibly bullish about the uh about the potential.

But but as always we you know, we're reluctant to sort of charge in and say, Hey, we've got the answer. Uh uh our approach is, hey, let's work with you and provide you with some grant funding to examine this and and let's see if we can prove together what the outcomes might be.

I like the bullishness. I really want to be bullish myself. I I'm not feeling very bullish particularly at the moment, but I'm very keen to get back to that space because it where is where I always sat. It's just I I hope that those developing countries do get access to the tools to enable them I'd love you to be right. That would be a really good outcome. Thank you. And and Sally, to use the last word on the diversity question, because it was one that you prompted me with.

uh right early on. Obviously Birkbeck is an interesting institution, isn't it? You must have a very diverse student population. We do. We do w everything from accomplished public servants returning to do almost leisure learning or to credentialize something or career change. to students coming to us who have essentially opted out of the school system a long time ago.

And we are moving them from sometimes through our FY foundation year, sometimes straight into university year one to graduating with a University of London degree. at at the end of three years, four years if they take extra time, six years if if they start part time. So We really need the opportunities that AI gives us to support those those students.

And interestingly, I like the democracy of things like studiosity because quite often the students we're getting are students who are very reluctant to ask what they think of as being official channels for help. So things that they can do as self-help, either through something like the Studiosity platform or just through getting onto Chat GPT and other other sources of of AI themselves is so important for them.

Some of them would rather cease their university education than approach anyone who looks official for help and support. And so some of our role is going to be explaining this. telling them what they can do for themselves and channeling and front-ending them into this. I think that is that is hugely important for universities to be able to do that. And for us to be able to do that, we need bottom up buy-in and top down policies that allow this to happen.

That's such an important point, the self-help, that not feeling able to approach I I really have seen that myself in students are talking Particularly when I was teaching in further education, where you are often dealing with students from very poor backgrounds.

in the area I was working in and and it was very hard for them to to approach anybody official looking for anything. So I think it's a point well made and it really maps to um some fascinating work that we heard about on a previous podcast from Paul LeBlanc at Southern New Hampshire College in in the US, who's who's passionate about the possibilities of AI.

Um, for increasing diversity and really helping to change the system. It sounds like if you don't know him already, Sally, I think you'd get on like a house on fire.

Sadly we've reached the the end of our podcast today, which is always a sad moment, but I'm gonna give each of my lovely guests a a final word. I'm gonna ask each of you What would your tip be for anybody who's listening? Most of our listeners are educators. Not all, but most of them are. Some of them are uh developers, ed tech developers.

What would your tip be for what's a good thing to do at the moment to explore artificial intelligence? So I'm gonna start with you, Ant. Sorry, I didn't give you warning. No warning obviously the answer is to uh listen to you Rose and follow that podcast and all of the things that you you produce, uh you're always the the font of all things. I I and actually I think but I also think that's true. It's it's it's actually it's be inquisitive. It's

This is not a flash in the pan. This is something we all have to have an interest in. And I think it's making it part of your own personal reflexive practice. Yeah, for you might get productivity gains, you might just be interested I had a good laugh making some uh pictures. of of the Australian Tertiary Commission. I mean I'll I'll send them to you later. But the you know, you can you can have fun as well. It doesn't all have to be too serious. I love it. Thanks and inquisitive. Good one. Mike.

Uh look, I I I think if you haven't already experimented with some of the models out there, do so and as Anne said you you c you can have a lot of fun. They can be really, really quite entertaining. I in terms of the the uh the education applications, I I would say Don't leap to judgment. Um, perhaps r reserve uh reserve a sort of final decision until you've thought about.

how you can overcome obstacles because there certainly will be obstacles, but I I think there really is a um Uh th there's a prize. for for for working through that and and I think most folks will find it's very much worth the effort. Sounds good to me. Thank you. Sally, final one for you. So Ans and Mike have both I think talked about universities facing outwards on that. Talk to us and let's see what we can do together.

spot on. And that's wonderful. I love this. So inquisitive experimentation, a bit of patience to make sure you don't make a judgment too quickly. But if you're a tech developer, speak to educators, speak to universities, speak to the people who are going to be using your tools. Very wise words. Thank you. Thank you all of you for joining me on the podcast today. I think this is going to be a very popular episode. So brilliant. Thank you.

I'm hugely appreciative at having your great contributions to this episode of the podcast. Many thanks, Sally, Mike and Ant. I know our listeners are going to get a lot out of the discussion that we've had today. If you want more information on the series and our wonderful guests, visit the EdTech Podcast website at theedtechpodcast.com and connect with us via social media.

To see how Educate Ventures Research is helping any organisations operating in the education and training sector to leverage AI to be more impactful and effective, go to educateventures.com or join the conversation on LinkedIn. Thank you so much to Studiosity for their sponsorship of today's podcast.

Studiosity's education partners, universities around the world are ensuring their students have ethical AI-powered feedback that they can trust to uphold higher education standards and the integrity of student learning. What do students think? Let's listen in. Doing the same thing I've done with Studiosity, but just getting my feedback a lot quicker. If you hadn't told me it's yeah, I wouldn't figure out it's yeah, I'm a huge.

procrastinator. Now I can just upload, get my review back, and I won't have to wait that five hours. Your idols might show that you're pretty it's that mark. I think this is going to be really, really helpful for a lot of people. to give people confidence to to submit. I would describe it. اشتركوا في القناة I feel like I'm not sure. That's an eye open, I like some life same way. I would get it.

With more than 200 institutions already choosing Studiosity and seeing firsthand the effectiveness of ethical, scale, transparent student support. Read about all the universities investing in their students' sustainability for defensible evidence and return. You've been listening to the EdTech Podcast presented by Profether Rose Luckin. Have a great week. A quick message now from previous the EdTech Podcast host and founder, Sophie Bailey.

Sophie's latest venture work trip is all about connecting global teams in the context of distributed work and social learning. If you're a global learning and development leader for a scale-up, corporate Sophie would love to hear from you. Work Trip have a new product manager who is taking on user feedback and you'll get perks as part of the beta community. like you. Hello at work trip. That's hello at work trip W-O-R-K T R R W dot work trip.com