Audrey (Lecturer) (00:00.174)
Thank

Thank you.

Antanas (Student) (00:13.006)
So as a quick intro, would you be okay with children introducing your role at Dyson or what modules you teach there and how long you've been around?

Sure, I'm Audrey. I'm a senior teaching fellow in the Dyson School. I specialize for the IDE course MSc MA with the RCA. I'm the lead of the second year program of IDE. I've been teaching since I guess 2014 between RCA and Imperial College.

Okay, great. yeah, sort of diving more into the teaching experience can tell us a little bit more about what drew to this field how you found yourself in teaching.

I've been called a graduate from Innovation Design Engineering myself. I graduated in 2012 and they called me a few years later to teach experimental design.

Okay, great. And so over the years, obviously, you've been around teaching for some time. How has teaching at sort of higher level changed? What significant challenge have you noticed and how have you been addressed over the past 10 years?

Audrey (Lecturer) (01:12.318)
Challenges is following the world that evolves at the speed of light, so make sure you're up to date as a tutor, so you make sure you can understand a different generation of student. Now the gap at the beginning I had the same age as my student. Now I've got like 15, 20 years of difference with you guys. So keeping up with you, what do you guys need, how do you want to learn, but also how the world evolves and make sure every year I come back to a new academic year with new skills.

hard skills, soft skills, also teaching skills. So make sure I also get all the training on the side, you mental health training in particular. Mental health has been very challenging since the pandemic, unfortunately. So I had like extra teaching courses that I took on my own will to make sure I can adapt to your need. that, yeah, mental has been extremely challenging in the last four years.

So besides mental, would you say then that those have been the most distinctive sort of changes between generations of students that have come through?

I think so change of tools, the AI has been taking way more space, you since I started in 2014, we didn't have that much problem with AI, it was really minimalist in our lives, now it's taking over everything. So we have to adapt to in terms of marking, judging criteria, how do you guys work with AI, surveillance of ethics, it's been quite a big challenge for us to upgrade to that. AI will be the worst, I think.

the best of luck.

Antanas (Student) (02:47.054)
So sort of again judging in terms of obviously student performance and then using AI how have you used AI to improve your workflows? Have you seen benefits of AI on your

I don't use it for my teaching. I prefer my writing to be mine. My course content is, I'm not sure AI can help me, to be honest, because I need such a very specific or unspecific teaching. It's not like you guys in MNJ, where you have very precise module, very precise learning outcome. We don't have that in AI. Every year, the content change. So I have to adapt myself to what the world has been doing. AI cannot help me.

Yeah, I understand. Sort of, yeah, to build a little bit on top of that, you mentioned that obviously the content changes quite a bit. So I assume you are relying quite a bit on sort of interaction with the student. Yeah. One-to-one tutorials. Can you speak about how that has changed over the years? What changes have the...

There is more, so more students have been coming through, so now I've got way more students in each classroom, which is quite a lot of 101. So it takes more time to get to know you and deliver custom education for every single of you. That's what the challenges have been going through. I went from a classroom of 30 in ID to 70. So you can imagine beginning of the year learning your name, just that it takes me weeks. Then after that is understanding your skills, your background.

nationalities, culture, each student you need to learn their whole personality and make sure you actually deliver what is needed, how to push them when it's necessary, not pushing them when it's not the time for it. That takes way more time to understand the whole cohort, which is exhausting, by the way. You're halfway through a therapist's job to understand all your students and how to teach them so they are their best, they get the best out of you, out of the program.

Audrey (Lecturer) (04:43.406)
while delivering a normal, if I can say, cohort level of equality of how much attention you give to each student. Of course, it's a whole staff, it's not just me, of course. But then all the staff has to agree together and we always talk about student, how we can help the best, all of them equally, which is now, you can imagine, you guys are around 100 in each of your class. Mine are 70. And ID again is very complicated.

So that's been very challenging, giving that custom education that that's why we are so good, is because we spend so much time with you, understanding who you are and how we can find the best version of you, if I can say, in your four years with us.

Do you feel like you have?

Audrey (Lecturer) (05:33.246)
No, definitely not.

would you say you were missing? Or is it just purely human resources of being able to access each...

Yeah, I think human. Not yet. I mean, in ID, yes, now, because you know, the year I started now, it's been six months with me, so now it's fine. I know all of them. In design engineering, the MN is harder because there's hundreds of you and I only see a tiny part of the E3 and E4. So for me, it's a bit more challenging in the MN. Simply it's time because your calendar is packed, mine is really packed. So sometimes it's like hard work.

meeting that is not enough for you. So I do organize separate meetings for some students. So I have special supervision when students need it. I do give more time for my master's project. Also, I see them every week. Whatever they need, I will adapt. But just my workload is quite complicated.

Do you feel like you have the adequate information to judge whether the students are progressing as intended through the course or what happened to learning outcomes from previous years, how to improve?

Audrey (Lecturer) (06:40.526)
On the modules I teach, yes, because I only teach in Masters and Futures now. Even Futures is a very difficult module to judge. I think now I'm getting the grip of it. ID, yes, now I've got a clear understanding of when ID goes up and down.

How do you judge it?

So ID is very subjective, unfortunately. We do have marking criteria, but they are slightly different from the MNJ one. So it's also impact driven, which is quite hard to judge. But it's also unexplicable. There is that empathy you have in ID to judge a student. Being myself in ID once, a long time ago, so I have this empathy that helps understanding, but also I'm really good at judging what the world's.

need right now. And again, I've been graduating more than 400 students in IDE by now, so I start understanding.

Do you the lens?

Audrey (Lecturer) (07:44.046)
So we do have surveys, it's like, it is also judged as harsh as the rest of Imperial. So we have quantitative study about student satisfaction on each module, each tutor, each visiting lecturer, each core member of staff, the head of programs. Each workshop is judged. So after each semester, we send this survey like you guys probably have it. So I have.

actual numbers to tell what has been working, what has not been working and what I can change for the year after.

So is this like right at the end of the volume? Yeah.

Not at end of the module because in ID2 we have longer modules, so for us we have a server per term.

Audrey (Lecturer) (08:33.973)
Yeah, so I see my student every single week in IDE. Not all of them, but also I talk to the class rep that can be a voice for the student body. But simply I talk to every single visiting lecturer that comes every day. So at the end of the day, I've got a report from them. So they give me lists and emergencies. Or it can be green flag, everyone is fine. But they are here for mental health maintenance. So if they see some students are more comfortable talking to some visiting lecturer, but then they without

any confidentiality breach. They can tell me that that student is struggling, so I need to have more attention, so the next day I can go to the student and make sure they are okay. So every single day I've got indication of the health of the cohort basically.

How often do get messages from students or emails about certain questions? Everyday.

Not the hours, but several times a day. But again, the head of program are above me, so they get a lot of emails. I'm obviously a filter between the head of program and the student body. So I get a lot of emails from like, I got lost in the building, to I have severe mental health issue. It goes in both ways. Sometimes really nice emails that just want support for something, extra tutorial. I don't understand the marking criteria. But I get that every single day, all the time.

Would you say that some of the questions are repetitive on the content side of the...

Audrey (Lecturer) (10:00.898)
which gives me an indication the module handbook is not well written, so we need to rewrite this paragraph because clearly something is wrong. So it's great for the next cohort. Unfortunately, the effect is on the next cohort.

Is it challenging interpreting somewhat subjectively all the feedback from students to really judge what the takeaway should be for the next cohorts, or do find it very, very clear?

Sometimes yes, because it's very detailed. when it's I will say comfort and issue like building, you know, this room is too cold. Well, there's very little I can do. So every time we want the manager of the building, it's always the same answer, by the way. So that I try. It's not always immediate effect. Curriculum development is not always possible because curriculum development only happens not every year in ID, unfortunately.

So things don't move, immediate issues, not enough tutors, too much tutors, not enough workshops, that I can change immediately for the next term, for example. I can readjust. So it's a very flexible job where sometimes, one way to another, I need to adjust.

you

Antanas (Student) (11:11.374)
Yeah, so a little bit as a background to what we're doing, ANTI is effectively a personalized tutoring platform. So we built an MVP and there's already actually over 300 users trying out the platform today. What it does, I suppose it's a little bit of a departure maybe from your day-to-day interaction with students on a tutorial basis, but it's effectively a whiteboard on the students' iPads to which they can talk to. So as you're solving a tutorial question,

to IDE, but as they're solving a tutorial question, they may ask Appy, that being the voice AI, about a question that is present on a tutorial sheet on their whiteboard. Then Appy can guide them, as opposed to just giving the answer to the question through the tutorial sheet, sort of finding what are the specific challenges, specific steps that a student is struggling with, learning from their struggles, and focusing on addressing those shortcomings. Sort of in terms of the teaching practices,

What are your thoughts on that and focusing on one-to-one personalized education, a model more trying to give students more personalized support as much as possible.

It would be great to have support like a whiteboard as a record. We have a lot of students, we don't have that much memory. So sometimes some of your particular stories and then one week to the next you kind of forgot, which is fair enough. So that would be great for documentation support I guess and like constant maybe the students forget also what they asked or what was the protocol. So sometimes you guys need quite heavy protocol for...

risk assessment, antics, and you keep asking the same question every week.

Antanas (Student) (12:56.046)
So that is actually precisely that. What we're hoping to do is anonymize insights from the whiteboard, from the way the students are solving tutorial questions. We're already in focus with integrating above Imperial more widely. So effectively, if a student is solving a tutorial question for a particular module, and a lot of the students are struggling with the same parts of the question, we can basically consult the insights from that and send them over to the lectures as insights on their teaching, as a further sort of improvement.

teaching experience.

That would be great actually. Avoid repetition I guess from us too. Again like you said we receive the same email a hundred times which is fair enough though if you didn't understand it. But if it's a generated common data.

Yeah, that's absolutely right. So yeah, I'm just a little bit conscious of time. Just like sort of to wrap up maybe in a couple of words, so just really summarizing, what would you see as the main issues in education as of today? Wow. Quick fire.

I mean for me it's maybe not the day-to-day management but it's equality I guess, it's access to education for me is becoming dramatic. Not everyone can access burseries hours anymore, know burseries hours has been reduced a lot so for me it's equality, it's a huge issue in education. I'm not sure I have your project. No worry, thank you guys. Sorry for the express.

Antanas (Student) (14:19.822)
Thank you so much.

Speaker 5 (14:25.294)
you

Audrey (Lecturer) (14:31.266)
Yep, I don't want to go away with this. Alright, good luck guys!

Thank you so much.

you

Antanas (Student) (15:40.13)
Okay, that's good. It's all good. we Can you start recording?

Antanas (Student) (15:53.294)
So, I'm Billy Wu. I'm a reader in the Dyson School of Design Engineering. I've been teaching in the Dyson School of Design Engineering for about 10 years, mostly in our first year materials and manufacturing module, but also recently as head of program for a new M-RES program. So that's a master of research program.

in design engineering research.

Antanas (Student) (16:31.874)
So I am teaching our first year materials and manufacturing module where we teach students around what materials are available for designing products, so metals, ceramics, polymers, composites, how to manufacture them into different forms. Beyond that, I also teach our research masters project for the M-Res program where we teach students how to be researchers.

So posing good research questions, conducting literature reviews, scientific writing, data analysis, those sorts of skills. Really. obviously, covered by a wide range or some time of different students for different years. What would you say have been the main changes over the past 10 years in education when it to youth interacting with students? So from my perspective, there's been lots of shifts in the students.

One of the things is that we're quite keen to not only deliver technical knowledge to students, but it's also the mindset. So we always encourage students to kind of reflect that as they go into the workforce or their next stages. Not only is it the technical solutions and the technical knowledge they acquire, which is important, but it's also soft skills. When we often talk with some of our industrial partners, so we have an industrial advisory board, they often highlight the importance of soft skills, such as problem solving, team working, and so on.

And sometimes those are overlooked a little bit. I think we've seen a lot of shifts in terms of the students' comfort with handling ambiguity. And I think COVID was a big shock to the system there. We've kind of come out from that. And right now, I think we're starting to grapple with new challenges around how we embrace technologies like AI, large language models, and so on. Just to touch on those, you know, a range of white schools that we're trying to develop here.

Would you say that there's been a shift in how students best try and learn on their own to develop the soft skills in addition to the theoretical skills over the years? What sort of trends have you seen? So when we first started at Dyson School of Design Engineering, and I've been here since the very beginning, we've actually evolved the course quite a lot. The mission has always been to be more practical in terms of we're application focused. So when you compare our degree compared to other degrees, there's a large degree.

Antanas (Student) (18:55.694)
of practical projects as opposed to exam-based assessment. I think we try and encourage students to think about the learning and configure it in a form that's useful for their final outcome. Naturally, things like assessments are necessary evils and we can reflect on the merits of that. I think we've always encouraged students to reflect on the teaching but naturally students kind of focus on the marks and whatnot.

We've gone from a place where there was a lot of team working and some of the feedback that we've had from students kind of reflects on the volume of that. The feedback we had from industry is actually there's not enough team working and as you're probably familiar, sometimes there's a bit of pushback in terms of because we have assessments and group assessments versus individual assessments and the fair distribution of marks. I think that's an...

natural part of the system is something that we're still kind of grappling with in terms of what that right balance is. Yeah, and in students trying to grapple the content better, obviously in Dyson School of Design Engineering Public Engineering, we oftentimes have a tutorial format of trying to figure and get some one-to-one time with the lectures or GQAs. How had you found the use of that? How has that advanced over the years?

What is a big part that is supplying in the students learning experiences? Yeah, I think because so there are different modes of teaching delivery. We often look at something called Bloom's taxonomy of learning, which is kind of how we layer up the learning objectives. So many modules with help will have intended learning objectives. And broadly speaking, when you first come into university, those are what we call level four modules.

where much of the learning objectives are centered around understanding and remembering core content. As you build up, you have more application analysis type learning objectives. And then finally, in your final year of study, it's more create type learning objectives. And we try and structure our degree structure around that. Alongside that thinking of what do we want to deliver, it's also the modes of delivery, where we have didactic lecture delivery, where we give you a lecture,

Antanas (Student) (21:10.708)
and people sit there, that's fine. But we really want to move towards more like problem-based learning where there's a problem, you work together on solving that problem and applying your knowledge. So those are kind of the blends of techniques that we have. We see a great diversity in our student intake and naturally that does affect the way that we construct or impacts our teaching because

I think you see very different student archetypes from different educational backgrounds. Pros and cons, I'm not saying which one necessarily has a merit, but for sure it is often quite challenging because students come from very different educational backgrounds where maybe they're very useful to a solvorex approach where if you follow these steps, you'll get to the end solution. things like maths, actually, if you follow the steps, you'll get to your integration or differentiation solution.

or actually trying to encourage students to have more criticality of thought there. So I don't know if that answered the question. I kind of wondered around a bit, but those are some of my reflections. Could you sort of expand to what sort of techniques you've used to try and basically bring everybody up to the same level of sort of understanding, especially in the early years, of, you know, about whether that'd be, you know, some more personalization in terms of the learning that, you know, some students that are coming from different backgrounds.

Yeah, so one of the things we're very aware of is that we do have a very international cohort here and we obviously primarily deliver content in English and English isn't necessarily the first language of everyone. So when there is a predominantly didactic or lecture-based delivery, that can actually be quite challenging for some people in terms of a communication piece. I found that using flipped classroom content where

You deliver a video content where people can go back, relisten things that wasn't clear. There's transcripts or written elements there that often helps with that. Getting students to work in teams is quite useful, but also not just saying, here's a team-based project, but actually go and guide them through that process where there are techniques where what I've often found in very international groups with different skill sets is that there's sometimes miscommunication.

Antanas (Student) (23:32.78)
in terms of what people do. And there are techniques like back briefing where you get someone to say, okay, this is the plan for this week. Can you explain back to me what you believe to be the plan? And that aligns expectation there. And really we want to create these communities of learning as well where someone might be very proficient in something like CAD, whereas someone may be, has never done that before. And I think the benefit of having a very diverse cohort, especially in the master's programmes is that you can

engage with colleagues who have that proficiency and actually there's a mutual benefit in terms of they can teach you, they improve their learning by reinforcing that, you gain the benefit in terms of asking questions. Because ultimately one core limitation will be the availability of teaching assistants and also the lecturers. And I do think you need something a bit more asynchronous because my time is finite and are there other resources or the learning resources that

allow people to learn at their own pace. In terms of the synchronality of learning and sort of having sort of like the ultimate expertise in everything type of friend there, oftentimes students today obviously turn to AI tools for that. How have you found that from the lecturer's perspective, how have students used it and how do you upon as a problem or do you think it's a sort of beneficial for students learning as a world forward? Yeah, I think

AI is a fantastic tool and we're in a period of transition. To be honest, think universities and higher education is a very slow moving piece that we're still coming to terms with what it means for us. I personally think there's huge opportunities for that because at the end of the day, traditional delivery of teaching is limited by the availability of the teachers themselves. And I think we often score quite

poorly in our various teaching assessments in terms of feedback, as you're probably familiar, that we have finite time and we would love to give everyone individual feedback, but it's often quite challenging. So we fully acknowledge that learners come in from very different levels and we've got this mismatch of different levels of learners, different pace of learning, communication elements, and only one source of kind of feedback delivery. So I think AI has that potential to fill that gap.

Antanas (Student) (26:00.588)
I think we've got to be pragmatic about how we use it. So I do think with certain things like remembering and understanding type learning objectives that I mentioned before, AI is perfect at that because actually it's something that's very well defined. It's somewhat deterministic in terms of that is the definition for this concept like phase diagrams or even in maths, certain differential equations, numerical methods and so on. And if there is misunderstanding, then students can get immediate feedback rather than

waiting for ages for an email reply which they may or may not get. As with all tools, I think it's important to understand the limitation of this, that right now how AI tools work is mostly based on training data. So if you start to use these tools for purposes that they weren't originally designed for, such as more creative exercises, I think you've got to be a little bit careful about that because...

if you're trying to use AI for more creative processes because of the way it works through existing training examples, it's kind of reinforced what it believes to be that solution. So I think it's that criticality of thought that I don't want us to lose when we embrace these tools, which certainly brings a lot of benefits. But we shouldn't just see it as a silver bullet for everything. No, absolutely. Thank you so

And you also touched there a little bit on engaging students, giving them the right amount of feedback, how different students they progress throughout the course, how well they might progress throughout the course. Could you give us a little foreman insight on what tools you're using today? And do you feel personally that you know enough about your students, whether they're actually doing well or if they actually need help and may be struggling to reach out? Yeah, great question.

In addition to delivery of content, we often think about assessments. And assessments are often seen as this kind of negative concept and thought. Sorry, I just got a bit distracted. So it's often perceived as this very negative thing that, I'm being assessed. But ultimately, we should see that as part of the learning experience because we often have practice questions, quizzes, and I think those can help. I think the challenge is that

Antanas (Student) (28:14.062)
they should ideally be elastic, where if you're not sure about something, then actually you should curate the future questions. I've used things like static quizzes, like mental meters, cahoots and things like that, and they're fine, but they're not as personalized. So I think what's missing right now is that personalization. We do both formative and summative type assessments, and I think both serve that purpose, but it's that individual level of personalization, which I think we're missing.

at the moment in terms of why did you get this wrong? How can you improve and what can kind of guide you along those principles to get to that final learning? would you say that, you know, what sort of granularity would you really be looking for as a lecturer? Like, would look at topic bases that, you you'd like to hear how students are doing. And I guess, would you also be looking to, I guess, what are your practices today in terms of, say, if you define the...

from Mentimeter that students are struggling with something, would you actively try and adjust the content as you're going through it? Or would that be some like a learning for next year? So when I use Mentimeter, that allows me to get a overall aggregate feeling of how the class is doing. And if there is something where there's a misunderstanding of some concepts, we can spiral back to that. I think the shortcoming of that approach is that it's mostly based on a overall cohort.

So it does mean like if someone still doesn't get it, then they kind of somewhat get left behind. So I think it's really, if you can have that granularity, if someone's really struggling with that, we don't want to leave anyone behind. We want to bring everyone with them. But how do you achieve that as a scalable format where our class sizes are roughly about 100 or relatively small compared to other departments? So if you scale this to 200.

I practicalities are important. Of course, you could say that we give everyone individual mentoring, but ultimately there are many challenges with that model. Yeah, brilliant.

Antanas (Student) (30:19.438)
Yeah, my side, just like a quick breakdown of what we have actually done so far, a few thoughts of yours on that. So eventually we've developed a platform, it's called ANTI. In ANTI, what we have basically is a sort of a whiteboard tutor type of interaction that we're trying to achieve there, where students are basically solving their palm sheets on the whiteboard. And we have very quick clicks. You click a button and the AI.

that being, AdBeat sees the whiteboard and it can provide verbal feedback on the tutorial sheets instead of just giving you the answer. Basically, if you're stuck somewhere, it sort of tries to nudge you towards the right way to progress from that onwards. And our basic thinking, and we were working with Imperial actually to try and integrate this further, we've actually already had over 300 people try out our platform with Imperial, is to basically provide those insights that AI captures anonymized throughout the learning process and to feed that over to lectures too.

that we're basically drawing in sort of like a feedback loop that doesn't exist at that granularity level, basically on specific parts of the concept even, that students might start to verify. Maybe you can also do the concept at a high level. Yeah, I think it's a really big challenge. think when you look at how higher education and education has evolved, there was a few things that kind of came a while ago. MOOCs were a big thing, massive online open courses.

I think they were really useful in terms of encouraging asynchronous learning. You're probably familiar that people like Sam Cooper, he's got his Mathematics for Machine Learning course on course era. And it does have a degree of feedback that you've got quizzes and whatnot, but I think there is a level of granularity below that. And many companies like Google are very interested in the use of AI in education. So I think you're definitely along that right track. And I think the curation of how you do that

is the way I would recommend it in terms of you don't want to give people necessarily the solution but basically kind of lead them and nudge them on as opposed to giving them the solution because actually that's a detriment to their learning from my opinion. think for the good students you also want to I think you want to think about different levels of proficiency. I think there are some where you just want to get people to that basic level of understanding.

Antanas (Student) (32:45.282)
But for people that really get it, are there opportunities to kind of really stretch their understanding? And I think that's where the degree of ambiguity starts to come in, if you've done math, A level is pretty standard in format. But if you've done math Olympiad, I think that introduces a different set of skills where there's a degree of ambiguity in there in terms of not telling you what technique. So there are questions where use this specific numerical method to solve this problem. And I think

if you're trying to stretch people, being more ambiguous and actually figure out or evaluate what is the most appropriate technique, that's good way to kind of start to stretch people. Absolutely. And sort of to quickly summarize it, if you may, like a quick fire, sort of like if you were to name a few problems that pop most into your mind in education from a lecturer's perspective, what would it be sort of?

Say that again. few problems that from the discussion, would you say that are most prominent for you from the lectures perspective in education today? So I think partly there are some soft skills of how do you get students to embrace ambiguity there. And that goes back to the manifestation into something like Math Olympiad, where there's ambiguity in solutions there. I think there are

questions around how to work well in teams. I think we could perhaps facilitate that a little bit better. And if we can integrate that into assessments or the teaching in terms of the style. So we've talked about problem-based learning, team-based learning. I think those are good. I think facilitation tools for... So it sounds like what you're describing is mostly an individual-based system. For things like team-based learning, I think that's a slightly different set of skills there.

and some facilitation there could be quite advantageous because actually it's a different skill set to learn in a team and learn if you've got a big complex system, how do you kind of approach that where no single person can address it all? Brilliant. Thank you so much. This is very helpful. Yeah, it's been great. Okay, great. Cool. Good luck with the project. Thank you. As I mentioned, I think you're

Antanas (Student) (36:35.694)
You are your robot material and then you know just run through how you see the education future how you've seen it evolve over your experience at the material and other universities and just what are the expectations from the university side as we go along in this future, right?

about your product then.

Yeah, so eventually we'll get to that in a like what are your thoughts about Amphi? Like if you watched that, that'd also great. yeah, don't worry about it too much. Obviously we can edit things out and then obviously we'll send you the video. is mostly for our module purposes.

Yeah sure, that'd be great. How many people have you interviewed so far then?

We've had set up for around 10 interviews so far but this is only for the video obviously we've spoken to at this point well over a hundred people. So yeah, it's been quite a journey, but yeah final push I suppose. Yeah, yeah Wherever you want the other cameras to be set up for it

James (Director Digital Education) (37:30.958)
So what do you want me to do? have to sit back?

Speaker 5 (37:42.03)
you

Antanas (Student) (37:51.246)
Okay. That's great. Alright. So, as I said, just a quick intro maybe then. Who you are, you appear, what's your role been, and how long you've been around here.

Okay, sure. James Toner. My title is Director of Digital Education Services and Innovation within ICT. I've been at Imperial now two and a half years. But yeah.

How long have you been in education?

a long time. 20 odd years in different areas, be it in commercial, through to government, and now obviously in higher education.

Really, yes. And what do you feel the beginning was? How did you end up in education?

James (Director Digital Education) (38:37.742)
I've always been interested in computing and also interested in human computer interaction and interactive design. And so through that aspect of my degree course, I leant more towards understanding how people learn and interact with the technology that then led me into more the education field. Originally more project management.

But then that through sort of desktop support and training, I then lent into online education or digital online courses about 20 years ago. And then from there, it just carried on through.

Brilliant, And over your past experience over 20 years, speaking exactly about that, the human to computer interactions, specifically students to computer interaction, what have you seen to be the biggest trends? How has it changed?

I suppose it's just the opportunities that technology provides in enriching the education experience. Although there's been lots of developments over the 20 years with know lights of Facebook and and TikTok and all of these different YouTube and all these different platforms.

there's a perception that people are getting more technology savvy, but actually, from my experience, it's very much surface sort of learning and they don't necessarily understand the more detailed aspects of the technology. technology has been beneficial, but then also cause barriers in people not necessarily understanding how they could best utilize it. And so, yeah, I've seen

James (Director Digital Education) (40:27.436)
technology in both a positive light as well as a negative light as well.

And from the university perspective, how have you seen universities adapt to this? Have you been quick to react to new technological developments? What are you looking when they're looking to onboard new technology?

I think it's difficult to say in a sense because of the context of how a university works. And so if you want to do obviously changes to programmes or modules, there's a lot of planning that has to happen in advance of that happening. So I mean, what for one person is quick may be seen as slow in another person's eyes. There is also a lot of worry or this perceived worry

for example, when lecture capture came in and it was seen as maybe essentially replacing the education and seen more as a threat. Whereas now, I think it was still seen in potentially a bad light and there was lots of research about how potentially it could impact the student experience in a negative way that students wouldn't turn up and just rely on lecture captures, for example. But when sort of COVID

hit. I think the mindset changed a little bit as well as obviously experiencing it over a few years. So that's just one example of where technology came in gradually over a number of years but actually at a point now where it's not even thought about that it's expected that lectures are captured. You see it with the use of VLEs as well, Virtual Learning Environments, also known as Learning Management Systems.

James (Director Digital Education) (42:08.622)
The simplicity now of submitting coursework, when in previous years you would have to queue up outside an office and hand in a hard copy. There's lots of technology enhancements that you can perceive has made life a lot simpler for students, but also it's created a lot of noise and trying to then understand what is the right technology to look at and what is the technology that will help support students in their day-to-day life, it's become more difficult.

So yeah, you mentioned there's something interesting that we could maybe do a little bit more into. different people have different command of technology. Do you feel like the services that universities provide are tailored to each student to a sufficient degree, or do you think that that is still lacking?

I think different universities work in different ways and I think at Imperial where it's very federated it's very much having that local support within the local departments to help support their students which works well. I think where the challenge with that is the consistency of experience across even different departments within the same faculty and then comparing faculties across the university. So I still feel

we need to get to a point where it's a lot more user friendly in how technology supports students and staff. And it's just really lowering that bar so there is no training required and it's just seen as natural. There was a point, I just watched this recently, where

James (Director Digital Education) (43:56.91)
the mention of the iPhone, when the iPhone came in it opened up the market that a lot more people then use easily use a phone. The same with the Nintendo Wii when that appeared, just opened up the market that was a lot more natural for people to use. So I think there's things that we can do especially around the education experience that's a lot more personalised to the individual, that's more just in time regarding what they're doing, then they can then reference out to different pieces of support based on what they're trying to do at that particular point in time.

Absolutely, yeah. up until this one, we've been speaking only most about students. But what are the lecture side of things? How has that advanced? And especially the lecture student interaction, how has technology been able to support that?

Yeah, think this is, I think I was on a call even just about an hour ago where we looking at AI and how that is impacting on how courses and learning is delivered. And there may be a perception that academics are, they're under a lot of time pressure, but then seeing as technology is a potential to reduce that time pressure.

in a good way so then they can focus on value add rather than it just being seen as not doing as much work as they've previously done. think technology, like I mentioned previously, adds extra workload because they have to learn and understand how the technology works and what value they get out of it and then also keep on top of the developments as well as what students are expecting. Coming through from secondary school, college, experience maybe of technology maybe is slightly different to what they experience in HE.

and it's just keeping on top of those developments and making sure that they value add to the student experience.

Antanas (Student) (45:38.318)
Absolutely. And you mentioned the support that the technology can give to lectures. Where do you think are the areas of focus where tech can be enabled to load off some of the burden off to lectures to take them?

I think it's about enriching what they're doing. There's always about this contact hours, but that could be 20 hours of lectures, didactic lectures, as as five hours of really valuable interactions. There's some difficult metrics that are used to quantify the quality of the teaching. see technology to provide those insights. know, analytics coming through.

learning analytics in a sense that an academic can look at how their educational content has been engaged with and then adapt that teaching based on that immediate feedback to then make sure that the learning objectives are being understood by the students and they can actually progress into the higher levels of the Bloom's taxonomy of learning. I think the technologies has opened up lots of data and

and feedback that probably wasn't there before. But again, it's having that time to make sure that it's to digestable to then reference it back and how quickly do you reference it back and how quickly do you adapt your teaching is a challenge I think.

What has been any practices in the past? mentioned there isn't a lot of coolant today, but what have lectures gone off to try and get these analytics from students? And how have they been able to adapt to the change?

James (Director Digital Education) (47:20.332)
I think there's some interesting dashboards that appear around student engagement in regards to looking at content and attendance and scores and some correlations around that.

What would be the shortcomings of those?

The shortcomings, I think we need to be careful that data doesn't and those interactions doesn't show the complete picture and there could be lots of other things that going on regarding students' day-to-day life that need to be considered but it's a good way to identify what's going on or give you at least a sense of what's going on across your particular subject. On the other hand, it could act negatively on students that they feel that they have to...

are using the metrics to then compete and then yeah I think they may add extra pressure on both staff and students that maybe wasn't there before and so I see those probably as the negative aspects of that data.

And more from the university side, so sort of covering all the same older 50s and 80s, how does the university see the use of this technology moving forward? Obviously, there's in day-to-day, all the students that you can help with the personalization aspects, can help lectures on the analytic side. But what is a university's proposition over the next couple of years with the recent developments in that, how can they stand up?

James (Director Digital Education) (48:49.934)
I probably can't speak for the whole university. can only give my...

opinion from my own personal experience. I think some of the conversations that probably are happening at the moment is just reinforcing what it is to have a university experience and a university degree at the various levels and what does that individual get once they leave university that makes them progress and succeed in the future. Obviously the course is a key component of that but also personal development and how they interact with their peers.

the academic and just how they live life during that period is a massive growth element and it's really focusing on that. were seeing it years ago when MOOCs came along, was like well now universities are no longer needed because everyone can just learn online but then I think that probably doesn't really, it dilutes what the experience is of coming to a university so I think it's really, and with AI similarly, it's like well why do we need a university when we can just...

question and will get the answer. But I think what that takes out is the human factor and engaging with, I can mention with people and developing as an individual on how you work together and how you come to goals and how you fail and then succeed. There's a lot around failing or learning fast that needs to be understood. So I think the universe and also that re-emphasis re-emphasis the purpose of being together and physically being together. I know you can do this online, but still it doesn't quite have that same level.

of interactions I see when we're in a physical space. There's lots of cues that get missed, know, knowing when to interrupt, when not to interrupt, not getting distracted by going onto the internet if you're on an online meeting or what have you. So I think it's really trying to pinpoint or be very conscious of what the value adds with being on a degree and then using technology to support that, being clear on that.

Antanas (Student) (50:45.998)
Yeah, and of taking even a little bit further into that, what are the main areas, obviously, that we can use the technology and to support the human aspects of learning, more focusing on the theory of learning and what can we not optimize in students' day-to-day lives?

James (Director Digital Education) (51:06.33)
Sorry, I that one again.

What can we try and optimize in students' day-to-day lives and lecturers' day-to-day lives so that we try and support the human aspect of learning as much as possible?

I think probably the prompting may be valuable in regards to sharing, you know, what would be...

useful to look at today as opposed to the course. Especially you to be careful on balancing it because there's an element of that you should be doing that anyway and sometimes relying on technology to do things that you should be thinking through anyway could be seen as a negative aspect but then in the future will we need to do that and will prompting be the know the go-to thing in the morning something tell you what you need to do for rest of that day but do you want there's also that aspect do you want that to happen and feel that contri...

in a way controlled by the technology. So I can see there's a line in between that for me at least where it needs to have a balance where technology is helping me when I need it to help me and I decide when it needs to help me or not. So yeah, I think we need to be careful that we don't become too reliant.

James (Director Digital Education) (52:20.212)
technology that we can't then think for ourselves as well as question what's coming back. Because I think that's another thing about being at university, it's not just about the course and I mention earlier it's about being able to critically think independently. So even though you get told something you always need to be not necessarily...

You always need to question things when you get told them and have that breadth of understanding to be able to have a conversation about a particular topic rather than take it as read. And I feel that we probably need to be a careful saying with technology and with AI and the responses that come back, there's hallucinations and all sorts. So with time it will get more sophisticated, but still, it's still someone's training it and it's still coming back with certain feedback. as an individual, we all need to be questioning.

Yeah, thank you so much. you know, just obviously even involved with as we were developing ANTI throughout the term, I'll maybe just give a quick recap of what ANTI is today and maybe a few thoughts if you could share. Give the demo? Yeah, we can just do them. I do you want to do the introduction first? Yeah, yeah, yeah. So ANTI, obviously, the way we built it is a personalized learning platform. Our focus of ANTI has been so far

effectively engineering students and engineering tutorial sheets, right? So how can we optimize that part of contact hours that we have and really not enough availability of tutoring that we have during the tutorial sessions, right? So effectively what we are proposing is a platform where students can log in, they can solve their tutorial sheets on a whiteboard and once they get to the point where they are struggling they can share

they can prompt our anti-AI which is actually has the view of the whiteboard and they can speak to it just like they would with a GTA, a lot of the authorities but what the focus really is is that it wouldn't just give the answer to them on the plate. The focus is to just get the student unstuck from a certain point, make sure that they are understanding in the way that they are learning, in the way that we've seen them learn in the past, thinking about what their shortcomings have been in the past and also to feed this information

Antanas (Student) (54:39.296)
to lectures. we can basically consult the old scholarship into different topics and how students are responsive to those specific topics sharing that information to lectures. So in terms of our progress so far we've had over 300 students already try the platform with over 70 sign-ups that are actually using the platform so far and yeah.

So we haven't showed you the demo yet.

No, I'm just saying the latest.

This is the third, like I said, 350 students have already signed up to use and we have

Is this when you were demoing when online it was having a bit of a... Yes. That was a few months ago, wasn't it?

Antanas (Student) (55:25.198)
Yeah. It's quite a bit more worked out now.

Okay, so you know about differentiation? Yeah. Okay, great. So I'm just going to give you the demo here. Here's the iPad.

Okay.

Antanas (Student) (55:47.746)
So when you're ready, if you just press start session.

Allowance?

Hello. And then just start speaking to it as if you would.

So imagine you were doing a tutorial and they just go like, hey Amphi and then Amphi.

It's Alfie, Alfie, not Alfie.

Antanas (Student) (56:15.214)
Yeah, assuming that you want to go over to the next step of trying to find the derivative of the equation, we can just ask if you're set on how that can be.

So I just said Alfie, can you provide me with the answer to this?

Or, how should I say, the derivative of the equation y equals 2x is 2. Would you like me to explain that?

Yes, I would.

Okay, the derivative represents the instantaneous rate of change function. In this case, y equals 2x, the rate of change is constant. For every one unit increase in x, y increases by two units. This rate of change is what the derivative represents. And since it's constant at two, the derivative is just two. Does that make sense?

James (Director Digital Education) (56:44.216)
change of

Antanas (Student) (57:04.278)
Okay, let me try a different approach. In calculus, the power rule can be used to find the derivative of terms like 2x. Hey, hey, Amphe, could you give an example in terms of like an analogy? So maybe use football as an example.

Let's use football as an analogy. Imagine a football player running down the field at a constant speed. The derivative, in this context, would be that constant speed at which the player is running. Since the player's speed is consistent, the derivative is a constant value. In the case of y equals 2x, that value is 2. Does that help you understand the concept of the derivative a bit better? Yeah. If you maybe try writing that out.

Yes it does. Can it ask me? Can it draw a picture?

Not quite yet. So we would like to highlight some stuff on the elements on the screen and some stuff like that. If you just write down what the answer is to that equation. So y equals 2. dy by dx equals 2. So just write down dy by dx equals 2.

So.

James (Director Digital Education) (58:17.378)
D Y

And then over dx.

Sorry about the X.

equals 2. And then just unmute yourself and ask Amphi is that the right answer?

Amputee, is that the right answer?

Antanas (Student) (58:40.62)
Yes, that's the right answer.

The derivative of y equals 2, x is 2, represented as dy slash dx equals 2. Is there anything else I can help you with? So if you just pause the session there. Stop the Yeah, Either of them is fine. So this can get a lot more complex. So what we didn't show in the demo is you can just upload an image. So for example, one of our favorite demos is a boy moving up and down on the screen. And it's like a line. It's really, you cannot tell what the question

is but it's a line that goes up and down it's basically a sine wave that denotes like the height of the boy over time with the boy on top and it's got a massive derivation underneath it and there's a problem halfway through the line and then it goes like you basically ask the question, Amphi, I'm stuck, where did I go wrong on this question and it basically breaks it down step by step without giving you exactly where it went wrong it was asking questions and to go like maybe try and consider this approach when you're moving from this step to this step

It was able to break down the problem. So literally any difficulty from like something as simple as just differentiating y equals 2x all the way to like a quite advanced calculus and other kind of subject areas. What I also really enjoyed though when I was solving my questions is that if POMC, whenever you get stuck on a part of the question, with example, other examples that you should solve to practice that part of the question, to better really sort of nail down that approach so you don't get stuck in the future.

Have you got any sort of feedback on how students are using it? There's a little bit of me that maybe can imagine that people just go, just give me the answer.

Antanas (Student) (01:00:21.996)
Yeah, so I mean that is still there. If you really explicitly prompted to just give me the answer, it's still going to do that. But we've actually got quite a bit of feedback from the students and that's how we've been iterating over the past two weeks from anything from the actual UI components, the students to make the interaction better, to actual prompt and actual settings of the AI and how it engages the students. And so far it's been...

did that at beginning.

James (Director Digital Education) (01:00:50.99)
You know, it's great. I'm just merely thinking, you know, obviously, and I suppose it goes back to one of the points about, you know, individuals wanting to learn in their own way. And so maybe giving them the answer isn't as bad as straight off may not necessarily be bad. And then they'll work their way back, but as as they work their way back. And that's the bit I suppose, I know there's some train of thought about not giving like even model answers because of the concern around, you know, pressures meaning they move on to the next thing.

or the individual who's onto the next thing. But it's maybe just a consideration. I don't have an answer for it, but I can imagine it.

around with the advances of other AI models at the time, like we're not restricting the student to just using ours if they really want to go and get someone to work with you on the trial more than they can go somewhere else. So the longer they're going to use Amphi the more it kind of tailors to their learning style so if it knows that the reason why this student's asking for the question here is because they just want the answer and rather instead of actually doing it then we can challenge the student in that way and we can go you know maybe try and give it an

shot in the same way that a tutor would for that example because it's you know this is something the students could be using 24-7 for for the tutorial sheets this is something they have had access to from the start of their university experience to the end this could be really kind of personalised to them. We think it was an extension to the uni experience so something that's kind of constantly with you that constant tutorial that you can always fall back to when you know you know one o'clock in the morning when they're when the tutor isn't available.

Oh yeah, yeah, can see a massive advantage. they haven't, you know, I'm saying glass is half full. It may be, you know, say a student turned up one day after years of using this and forgot it, or they didn't have access, where does that leave them? And maybe, I suppose you could compare it to other technologies that, you know, or just get by. yeah. Glass, glass, definitely half empty. I think it's great. Don't get me wrong. I could really see how this would benefit.

James (Director Digital Education) (01:02:57.902)
Especially I think some of the ideas you had when we last spoke about it, being able to start drawing and moving the equation around and seeing things develop through this would be amazing. Rather than it just obviously commenting on what the individual writes on the whiteboard. But you know, opportunities there around the analytics, the data, seeing where people are struggling.

So it might then reset how an academic would view how they presented the topic and then goes into that learning analytics thing I mentioned earlier.

So now you know about AMP. So yeah, so just in terms of your thoughts about AMP, if you were to summarize it, and I suppose the applicability of the tech into the future of one university might look...

Yeah, I think it's great. The concept, the idea, the initial pilot or development of the product. And I think time will tell regarding how students, what students gain from utilizing it as you're doing at the moment with the how many students are signed up to it. And like I said, some of those developments, you'll mention it as it acting like a...

an academic assistant, they can call on anytime, anywhere would be great. I suppose it would be interesting to see how it develops over time and do people see it as that companion or do they get frustrated with it or you know how does how will this go but time will tell and your evaluation will tell and I'm sure once you get that feedback you'll adapt the product accordingly. But yeah, no, very interesting how this develops.

James (Director Digital Education) (01:04:39.727)
and obviously wider applications not just in this scenario but other areas of the college.

And just to recap also, quick part of a couple of words of what are the main problems at university, as you see going into the coming two years, the adoption of tech.

the main problems. I think it's just keeping sure that the core value of being at university is still understood and recognised. It's not about just pushing people through and getting scores, it's around the learning and that development as an individual. think ensuring technology that can help support that is key. suppose a bit with this maybe even, you might want to cut this bit out, but if it's been negative, is that this might act as just being

making people go very solo and insulated and them just dealing with that, whereas in the past you may have spoke to a group and gone through that as well. yeah, just as a, we may get that question. So you probably, you already have that question.

We actually built Amphi in the original showcase that we gave to you, built with collaboration in mind.

James (Director Digital Education) (01:05:50.339)
yeah, there's people there was meant. Yeah, I remember that. Yeah

So we haven't turned it on on this version, but you can have multiple people on the call.

I'm interested in how that works. Have you seen in action? Because I can imagine bit of chaos going on when everyone's trying to do something, getting an ampy, trying to catch up with what's going on.

One of the biggest changes in our department has been something called Figma. It's basically a UI design tool, but we've used it for literally everything about our course. Because it has a student license, so everyone can have access to free and give it out. You basically have little mouse icons that go all over the screen so you can see where other people are. You can tap on people to see where they are. You can collaborate on documents, PDFs, you create a document, research document.

That's right.

Antanas (Student) (01:06:38.722)
Everything from a research document to a presentation or a demo, the tools are vast. But the alternative what we used to use is InDesign, which is the Adobe Suite. And that was very much you would just work together to get the content, then you would hand it over to somebody. And then it would just be, it was really. Yeah, it's revolutionized the way that we work. We do mind mapping, brainstorming together now.

Okay.

James (Director Digital Education) (01:06:59.448)
So there's new stuff there,

James (Director Digital Education) (01:07:07.942)
It be to see that and see how that works because I'm having a bit of a challenge at the minute, a problem around how we present the user experience in, say, for lecture theatres. And it's all very well, you know, writing down a list of all of the technology, but not knowing how people are interacting with that could be a challenge. I don't know if this tool would help with that, I've been, as I mentioned earlier, I'm interested.

has become much more clever.

James (Director Digital Education) (01:07:35.274)
usability design was what I sort of started with and in the old days it was a lot more basic with bits of paper and flipping through bits of paper rather than using any cool software. yeah, it'd be interesting in your thoughts on what we could do.

Yeah, I mean, I'm happy to talk you through it. You guys are on it. I absolutely love thing that I have to talk about. It's so great.

Yeah.

James (Director Digital Education) (01:07:55.064)
But good to see it. Yeah, but that was the only thing about this, you know, and that's also the issue because I think I mentioned, I don't know if we're still recording, but just generally, saw a company called a BRISM. Did I mention a BRISM to you? A BRISM. I can send you the link to the details, but they came to me when I was at King's College to show their course. It was like a next generation virtual learning environment where...

Thanks.

James (Director Digital Education) (01:08:22.246)
It was all about it digesting all of the educational content and then reacting to how an individual interacted with that course. And so then it started bringing up stretch activities, more guidance on what they should do. I never got to a stage where we used it, so I moved to Imperial, but the same concern I had with this potentially being a bit more of an individual experience was the same with that.

Whereas if you're just getting everything you need just from being in that course, then I don't know about some of those wider experiences that an individual might need. But then I suppose you need to factor in different people, behave or have different tendencies, introvert versus an extrovert, and I suppose catering from those. So I suppose I'm talking myself out of the problem. yeah, there's so much to consider. I'm picking, I'm picking. But yeah.

James (Director Digital Education) (01:09:32.058)
I think generally, I suppose what's different, and maybe I should have said this earlier, in HE is it's not a top-down structure, it's not you're expected to do this this way, it's very much independence in how an academic feels a particular program should be taught, a module should be taught, or know a particular session should be taught. So I think maybe that's some of the challenges with getting some of this stuff brought across an institutional as a standard application because there's so many

people to consider and different approaches that happen at an individual like that federated meant I mentioned earlier. So maybe whereas if you're in a company, a commercial company, and they'll just come down and say, you do this and this is the tool you use, you'll see at Imperial where everyone uses all sorts of tools for all sorts of reasons. So maybe that's where the challenges around governance and oversight, but in a way that enables the departments to react quickly for their particular need and what they want to do.

Whereas maybe in a big organization it would take months to years to get that standardization. So yeah, again, that's a of a double-edged sword as well.

describe some of the key stakeholders in the education like what departments are responsible for the output of the education to the students and then how this tool might impact those people, whether that be good or bad in terms of the analytics side of things for Amfie and the student.

Well, I don't know exactly how it's done at Imperial. I had a better idea at King's because it's just because I was there longer. But what tends to happen is there would be an element where a programme would be decided upon. We wanted to have this as a programme within our department because we see demand for it because of where the market is and we think students would benefit from it.

James (Director Digital Education) (01:11:28.428)
And then through that then they may create the modules and how the learning outcomes based on the modules for that particular programme. That's like at the outset and I'm not sure what level that's done at Imperial. I suspect it's done at the department level and feed up to the faculty. I don't know if there's a central place that that gets ratified or not. And then over time as they want to adapt that programme.

or modules, they would then seek to add modules to it, remove modules. If no one's interested in doing it, there's those factors that play into that. And then in amendments, they might want to do to that module. that tended to be a yearly activity. And then obviously then, the student feedback via the various methods, module feedback, and assess and have you then play into where they want to develop that module forward.

I suppose the challenge maybe is when the person who had the original idea to do the programme then moves and then who then owns that prior post, that academic moving on and then how that programme is developed over the years. Maybe, I don't know, I think sometimes people like turning things on and doing new things and maybe there's less happening where things are taken away but I don't know the details.

Taking a at the analytics, I don't think we've mentioned what sort of analytics we're able to collect with this. So think of understanding over time, engagement, concepts, key concepts, next best actions for the lecturers and stuff. What do you think, if we had this pool of analytics and engagement analytics, what departments do you think are key stakeholders in that?

The Education Office is probably the main one I would think of.

Antanas (Student) (01:13:23.086)
Do you think like, so kind of going out there, you admissions would be?

I think maybe if you had an example of what you've done and how it's helped and it's always better to come with the evidence of what you've done and how it's improved a particular cohort or area and then that can then get fed through different areas of governance so then people to you know build on the idea. I think that's probably the best way to approach it. Education Office, Phil Power. Peter Haynes' area is you know they lead the education office, Alan Spivey as well.

those would be the key individuals to go through. Obviously there's the Teaching and Learning Product Board that I look after with Alan Spivey and you went to the Tools subgroup recently where that sort of fed in some of the ideas. But I think if you were to do something like you are doing, just getting on and showing how it works and the data that comes back and then what that means, you know, being careful that it doesn't show the whole picture.

that would be a good approach to take.

Could we maybe talk about Kings? Is there any similarities and differences in terms of how products are used in Kings versus Imperium? How does the size of the uni affect the decisions that made on what products to use?

James (Director Digital Education) (01:14:32.332)
Sure.

James (Director Digital Education) (01:14:47.864)
So I suppose when I was, I worked at City University before King's and between the three, although there's a lot of similarities, a lot of differences as well, and it seems to be based on just how the institution has developed over time. If I give a similar example but not quite in the same field, at King's all teaching spaces are centrally managed.

So that enables you to have a central standard of what's expected in each of the teaching spaces. And anyone from anyone could book whatever space is across the college. However, here, departments own their own spaces, and therefore there is a variance in how they have those spaces. And that's just because how it's always been here. I think faculties are quite new as well. think faculties have only been around for 18 or 20 odd years. Before that, it was still just with the department.

So you can see where if you have some governance and ownership centrally, it makes it lot simpler to standardise the approach. And so we used to have, in the digital education space, used to have core tools that we expected everyone to use. We had recommended tools that we said this tool you should use for this particular scenario. And then we had recognised tools which we knew were happening or people were using them, but there was no central support for them.

We were bit worried about that as well in saying that we recognise these tools because then it may encourage people to use them further, but in a way that they might build up on what I saying at the very beginning, or of this part at least, around great to have people use it and learn from it and then feed back to see the value in using that tool. But many a time people just see this tool and think it would do everything for everyone, but then when they practically use it, it's missing key components.

that doesn't get picked up until people are actively using it. So Kings had a much more centralized governance of ownership over things, but then you could then argue then that made everything a bit bland, not bland, everything a bit standardized that might not necessarily be the right or the best tool for the job. So it's a balance in that.

Antanas (Student) (01:17:07.662)
So would you say that the tools themselves are lacking in personalization aspects that are out there in education offering?

What I find is that sometimes they probably, in my experience, haven't really understood the particular user cases. And how I've seen some of the tools designed is more...

So I think I said this to you before, in like a virtual learning environment, it's very much organized in containers within content, within containers, within containers. And that's where the content is. And that's where you go to get it for a particular point in time, rather than stepping back and then surfacing up content to an individual based on their week or what they decide upon. that's what I mean. It's not been what I've witnessed has not been very user centric in that sense. And thinking from a staff

a staff perspective, it's more, well this is how the system works and this is how you should use it, rather than actually how are you working with it and how do we then present, make that what you want to do easier and is the best ways of doing it. So that's more of a generic thing across most technologies and what you find what one people can do, another person wouldn't. And it goes back to the point I mentioned about the Wii and the iPhone, reducing barriers of entry to use the technology that is just natural for people to use it. And we get that with AV lecterns, know, what...

with one person, it's easy, it's just tabs and this is how it works. And so someone coming into that cold, not really knowing how to use it at all. So it's that lowering that bar of getting a base level and then, you know, sort of revealing features as people get more confident with using the tool. And what you find with software providers is that they tend to start with one core.

James (Director Digital Education) (01:19:02.742)
element that they're looking after in the education space but then they start branching out and it starts getting a bit murky over which tool is the right tool for which space. that's, I'm probably rambling, but that's sort of how I've seen the technology develop over the past while. It's more how the software providers think it should be rather than really understanding what it is the academics need and the students need. However,

if they go to 100 academics and 100 students, what answers they'll get will be 100 different answers and 100 different suggestions. So there's like a balance there of just coming somewhere in the middle I'd suggest. I did have another point but I forgot now.

Antanas (Student) (01:19:54.638)
We've taken up a lot of your time. Can we finish up with some PR for us? What do you see the potential for Amphib to be?

Don't buy it, that much. I think there's a huge potential. Like I said, I've picked some of the potential issues, but the whole real-time reacting to what you're doing, understanding, helping to explain concepts, drilling in and providing different methods to...

for the learner style to explain concepts is in a way what you would expect to have when speaking directly to an academic. So in a way it helps, it's not the same, but it helps on that journey and then it helps you, that individual have a much more enriched conversation with that academic because they would have learned the baseline, understanding more of the baseline. So I see that as a massive plus for that individual at a particular point in time.

Looking at the analytics over time, that would be great as an individual to see where am I struggling, what am I doing, where do I need help with to help maybe improve their learning interactions for themselves. then obviously that analytics then feeding into, I call that sort of like, then analytics leading into the learning analytics for the academic to see how they can better improve their educational content. I sort of use learner analytics and learner analytics as more the individual.

point in time and how they're engaging and learning analytics is looking at the content over time to see how it can be better improved. And I can see this as a great way of collecting data and feeding into those conversations. I've been around for a number of years.

Antanas (Student) (01:21:45.358)
Sorry, from the university side, stakeholders that are overseeing education across different departments, is there something that they would look to see from this tool?

I think, I'm not sure whether they've identified this is a problem that needs to be solved. So I think it's more that this can enrich what the current practice is. But what I've seen, there's lots of, you know, there's always lots of discussion around what we can do, but having something that you've got evidence is working and has improved learning outcomes and attainment is much more better than.

a really slick pitch. there's lots of, you know, all the talking in the world, but if you have the evidence, that's where I go to and it seems like you're doing that anyway.

If you... Sorry, I paused it.

Antanas (Student) (01:22:47.502)
There might be bit of pushback if Amphi would ever be released. So hypothetically, if you were to release this within Imperial, what would the university stance be on how you would integrate it? So example, people might think, you're trying to replace lecturers with this tool. If the uni felt that it was probably enough to implement and they thought that it didn't replace the lecturer side, what would be the university stance to

question.

stancing

So say someone came up to you you'd be like, why did you integrate Amphi into Imperial? Because it's replacing Lecturers. What would be the university's stance to go like? Actually, no, we feel this is embedded within the university. feel this is an extension as an example.

I think it's similar to the brief mention about lecture capture. So if you can sort of picture it in the same way as when lecture capture appeared it was seen as originally in a fret. Also people, actually I don't understand, didn't necessarily want to be recorded, well, you know, doing that now.

James (Director Digital Education) (01:23:58.252)
But then over time they saw actually it was helping. But I suppose it was helping in a way, but again, we need to be careful that...

Originally lecture capture was an addition and an add-on. An individual was always expected to turn up to a lecture to do the lecture and have it as a backup. But what we saw in certain scenarios of students used it as a replacement. And then that negatively impacted on their scores at the end of it because they all started trying to pack it all in at the end of the module. But over time...

benefits outweigh the disadvantages and I think students then realise that actually you can't do that and so over time they realise that they can't you know now just get the answer and then move on to the next thing because when it comes to the exam or anything at the end of the day they realise they haven't learnt so I think time tells everything.

experience where you're saying this is enriching the experience to have better levels of debate and discussion around the learning is where this will help. yeah, so and ideally you'll bring everyone to similar levels or help people reach a certain level to then have those bigger or more enriched engagements. But yeah, that's how I see it. think using it, people, word of mouth, evidence, seeing how students like it.

how academics have benefited from it and feeding that back in and that will hopefully create that more positive angle rather than it being seen as replacement. But people tend to naturally go to replacement because you've seen it with AI, you've seen it with electrocapture, you've seen it with all sorts of technologies. Some people see it as just an efficiency innovation rather than like an improving sustaining innovation.

Antanas (Student) (01:25:58.419)
That's it? Yeah? Is that right? yeah. Thank so much.

Thank you

Yeah, you can get access to it.

Antanas (Student) (01:26:21.152)
No, I don't think so. Yeah, we can give you access to the demo for free to show other people as well. mean, there might be some errors with the... Just like it's still being worked on, so we might push an error that might not work.

I did a meeting with the, I know I couldn't attend, how did it go with that tools group?

It went well, there was a guy from the engineering who was just amazed with it, I can't remember the guy's name, but he was like, we need to get you part of the engineering team, and then we did an engineering one, then someone else got in touch, we're speaking to them. He apologised that the rest of the people there were not as engaged, but he was like, don't worry about it.

It's a negative.

Antanas (Student) (01:27:11.78)
at end of the year for them to review the tools, for them to onboard somebody new and then they can pilot it with another department and then that's how they try and do things as opposed to onboarding something midterm. Yeah, of course. We haven't made any more progress with Lambda either.

Yeah.

James (Director Digital Education) (01:27:28.598)
No, no, no.

Yeah, it was just, think he had an agenda to basically just have the product embedded within it as in like to give him all the code and everything. I don't think that was anything that we were going to agree to. So we just reached kind of a.

Yeah, no, I just thought it was two and two and thought I'd put you together just to have that conversation.

Yeah, I think it would have been really good. I think we just didn't see.

part of the learning isn't it? can always put that in can't you? He had that and he decided not to. From what I saw speaking to him I think he was keen to basically hand over Lambda. So in a way I thought he was, in my mind he was looking at, well do you guys want to take on Lambda as well?

Antanas (Student) (01:28:07.758)
That was, I think that was the initial thing but I think else we kind of, I think because we weren't agreeing with everything that he was saying.

If you get a sense, know, it's going to be difficult working in that.

Antanas (Student) (01:28:22.776)
Thank so much.

Well good luck with it. I was still hassling Kings. They've got a group that they told me they would let you present at. But it's quite an official governance group so that might be why they're a bit thingy. But if you wanted I can ask them to just have my old C-Tel team and you do it to them.

Yeah, that's also funny.

Yeah, think Phil Blake is the person who was going through. He said there was a group, because I spoke to him twice. I did chase him, but he didn't get back to me.

Our deadline is in three weeks, so we are...

James (Director Digital Education) (01:28:59.734)
sure you can get you on the C-Tail team. Central Technology, I should really wear my own team. C-Tail. But they basically looked after digital education across kings. They've got a new person as my replacement but I've got on with her well.

Thank

Antanas (Student) (01:29:16.846)
If you ever want to talk outside of just Amphi as well about education and student experience we'd be to help. If you want to talk more about Figma, I'm happy

Yeah, yeah, and the other thing that's come up recently that we're looking at is a new AI thing called Cloud Force.

I have been to the cloud force. Sounds like sales force.

Cloudforce, Nebula One or something like that. It's quite an interesting thing. seems to, we need to be careful around it because I don't want to get too excited by it. But it looks, it's a partner of Microsoft and it looks initially from our conversations that it provides that chat GPT type experience but with Microsoft backend. But it can go off into any language model you

language model sorry but yeah we literally just getting a trial in the next in the next few weeks again so don't ever link into some of the stuff you do with the Gemini element because I wasn't sure if that was linked into it but it seemed positive they released it in London School of Business I believe to all students and all staff

James (Director Digital Education) (01:30:35.205)
interesting thing. The whole thing around, actually you might pick your brains around how we better describe those interactions in the physical and space. Hopefully. I was going to take it out the learning area and just do it in meeting rooms. Just keep it a bit more easier, condensed. yeah, maybe we can pick our brains around that. be cool. Thank you so much. I will try and get that thing over at King's if it will help.

Good luck.

Antanas (Student) (01:31:02.222)
And then do you still have that contact to Google as well?

I met with them, we met with them, but didn't go very well. They seem to have come from the Betfair and it must have been a long day and they were much on the research side than they were on the education side. So I could still try and see what they say.

second.

Antanas (Student) (01:31:18.574)
Okay.

Antanas (Student) (01:31:31.726)
I'm also doing some work for the Micro Campus on room booking and we're trying to move everything over to Microsoft Teams. So it means that know, Juan's vision for digital room booking and screens outside the rooms, it might actually be feasible with the new systems. So I thought you might be willing to like, yeah, you might be on board, not on board.

We need to make sure it works in...

the new MediaWorks.

Yeah, yeah, because you've seen that thing inside the room, we just need to make sure that that works.

Yeah, this will be a way of... I think the system's great at the minute.

James (Director Digital Education) (01:32:14.562)
Yeah, that's right. We didn't do anything, we just put in the thing and it worked.

This way we'll be able to do it for all rooms, including teaching spaces as well. So students or anybody could book a room that's outside of teaching time. No, we don't. Yeah, so we might be able to do a full Teams integration so you can book a room across the whole campus. That's the vision anyway, and I think we're getting quite close to it, so I'll make you aware of it.

but you need a screen on it.

James (Director Digital Education) (01:32:49.294)
Stance in what way?

So say someone came up to you you were like, why did you integrate Amphi into Imperial? Because it's replacing lectures. What would be the university's stance to go like, actually, no, we feel this is embedded within the university. We this is an extension.

I think it's similar to the sort of the brief mention about lecture capture. So if you can sort of picture it in the same way as when lecture capture appeared it was seen as originally in a fret. Also people, I could understand, didn't necessarily want to be recorded but you know doing that now.

But then over time, they saw actually it was helping. But I suppose it was helping in a way, again, we need to be careful that originally lecture capture was an addition and an add-on that was an individual was always expected to turn up to a lecture to do the lecture and have it as a backup. But what we saw in certain scenarios that students used it as a replacement.

negatively impacted on their scores at the end of it because they all started trying to pack it all in at the end of the module. But over time, the benefits outweighed the disadvantages and I think students then realised actually can't do that and so over time they realised that they can't now just get the answer and then move on to the next thing because when it comes to the exam or anything at the end of the day they realise they haven't learnt. I think time, time

James (Director Digital Education) (01:34:27.64)
tells everything and having those practical experience where you're saying this is enriching the experience to have better levels of debate and discussion around the learning is where this will help. yeah, so and ideally it'll bring everyone to similar levels or help people reach a certain level to then have those bigger or more enriched engagements. But yeah, that's how I'd see it. think using it, people...

word of mouth, evidence, seeing how students like it, how academics have benefited from it and feeding that back in and that will hopefully create that more positive angle rather than it being seen as replacement. But people tend to naturally go to replacement because you've seen it with AI, you've it with electrocapture, you've seen it with all sorts of technologies. Some people see it as just an efficiency innovation rather than like an improving, sustaining innovation.

So maybe use football as an example. Okay, let's use football as an analogy. Imagine a football player running down the field at a constant speed. The derivative in this context would be that constant speed at which the player is running. Since the player's speed is consistent, the derivative is a constant value.

Yeah? Is that right? Good stuff.

Antanas (Student) (01:35:55.638)
In the case of y equals 2x, that value is 2. Does that help you understand the concept of the derivative a bit better?

Yes it does. Can it ask me, can it draw a picture? it? okay.

Okay, I understand. I like the elements on the screen and some stuff like that. So if you just write down what the answer is to that. It sounds like you would like to. Y equals two. it's recap. The derivative of the. Y equals two. So tidy over. So just write down dy by dx equals.

It's always the answer to what I just wrote.

James (Director Digital Education) (01:36:30.305)
It's okay.

James (Director Digital Education) (01:36:35.576)
d y alright so you want d y

just mute it.

Antanas (Student) (01:36:41.902)
and then over to you.

over dx.

equals two. And then just unmute yourself and then ask Amfie if that's the right answer.

equals 10.

James (Director Digital Education) (01:36:55.03)
Amp Fee, is that the right answer?

Yes, that's the right answer. The derivative of y equals 2, x is 2, represented as dy slash dx equals 2. Is there anything else I can help you with?

Stop the session.

So this can get a lot more complex. So what we've been showing the demo is you can just upload an image. So for example.

One of our favorite demos is a boy moving up and down on the screen. it's literally like a line. You cannot tell what the question is. So the that goes up and down is basically a sine wave. That denotes the height of the boy over time with the boy on top. And it's got a massive derivation underneath it. And there's a problem halfway through the line. then it goes, you basically ask the question, Alfie, I'm stuck. Where did I go wrong on this question? And it basically breaks it

Antanas (Student) (01:37:51.056)
step by step without giving you exactly where it went wrong. It was asking questions and to go like maybe try and consider this approach when you're moving from this step to this step, it was able to break down the problem. So literally any difficulty from like something as simple as just differentiating y equals 2x all the way to like a quite advanced calculus and other kind of subject areas.

Yeah.

James (Director Digital Education) (01:38:12.675)
Yeah.

Antanas (Student) (01:38:20.976)
other examples that you should solve to practice that part of the question to better really sort of nail down that approach so you don't get stuck.

Have you got any sort of feedback on how students are using it? Because there's a little bit of... Like I did at the beginning.

That is so. If you really explicitly prompted to get just answer it is still going to do that. But we've actually got a bit of feedback from the students and that's how we've been interacting over the past two weeks from anything from the actual UI components that the students to make this interaction better actual content actual settings of the AI and how it how it engages with this.

You know, it's great. I'm just merely thinking, you know, obviously, and I suppose it goes back to one of the points about, you know, individuals wanting to learn in their own way. And so maybe giving them the answer isn't as bad as, know, straight off may not necessarily be bad. And then they'll work their way back, but as as they work their way back. And that's the bit, I suppose, I know there's some train of thought about not giving like even model answers because of the concern around, you know, pressures, meaning they move on to the next thing.

or the individual moves on to the next thing. But it's maybe just a consideration. I don't have an answer for it, but I can imagine it's something that might come up.

James (Director Digital Education) (01:39:49.666)
Yeah, you're right, yeah.

So.

The longer they're going to use it, the more kind of tailors to their learning style. So anyone knows the reason why the student's asking for the question here is because they just want the answer and rather instead of actually doing it then we can challenge the student in that way and we can go you know maybe try and give it another shot in the same way that a tutor would for that example. This is something the students could be using 24-7 for the tutorial sheets. This is something they have had access to from the start of their university experience.

Yeah, yeah, yeah.

be really personalized to them.

Antanas (Student) (01:40:57.774)
Yeah, it was just, I think he had an agenda to basically just have the product embedded within it, as in like, to give him all the code and everything. I don't think that was anything that we were gonna agree to. So, we just reached kind of...

Yeah, no, just thought it was for two and two, I thought I'd you together just to have that conversation.

yeah, think it would have been really good, I think we just didn't see.

was part of the learning isn't it? You can always put that in can't you? He had that and he decided not to. I think he keen to see how to do with Lambda. In my mind he was looking at, do you guys want to take on Lambda as well as...

That was the initial thing, I think that was kind of... I think because we weren't agreeing with everything that you were saying.

James (Director Digital Education) (01:41:43.906)
Yeah, if you get a sense, know, if you're going to be difficult working in that use of it.

Antanas (Student) (01:41:54.05)
Yeah, so we might be able to do a full Teams integration so you can look at it across the whole campus. That's the vision anyway, and think we're getting quite close to it.

Yeah, well I see.

Antanas (Student) (01:43:11.758)
So it's for the ERO project. And we really think about edtech products and how that of integrates into the life of a student but also with the analytics that we can provide. This is kind of new ground in terms of the analytics that we get provided.

Yeah.

Antanas (Student) (01:43:37.71)
So we were.

what kind of impact would this have on the lecturers? be useful? Would it not be useful? Is this going to improve the learning or is it going to retract from it?

Stephen (Lecturer) (01:43:52.974)
and your questions or how you describe what you're doing. The questions that you ask will give me some clues about what amphibious theory is.

Yeah, so the idea is that we just kind of understand more about you and about your teaching practices. So more of a research thing and then towards the end we're kind of introducing the Alpi concepts. that makes sense. Steve has just been talking about this.

Antanas (Student) (01:44:23.95)
I wouldn't say roasted.

Stephen (Lecturer) (01:44:37.866)
Next.

behind me.

Why is that? Well it was just the impression I got. I didn't ask him directly about it. He came back into the office and he seemed a bit flustered.

I think we just have a bit of a...

Antanas (Student) (01:45:08.078)
we're going to have when you reflect on four years of doing something, I think some of us have different perspectives on how the course should have been handled. But that's not to say that we haven't got good things out of it, it's that it's maybe not what everyone expected. So I'm just going to sit here and if you could just answer the questions towards me and we get started.

If you wouldn't mind just telling me your name, what your role is, a bit about you, a bit about the background and teaching. so restart. A quick intro about you, what you do and what you specialize in. There's a way.

Okay, I'm Stephen Green. I've been at the Dyson School for nine years. I have a background both in design consultancy and working in design in higher education. So guess one of the things about myself is that, especially in Imperial.

I'm a bit unusual in that I have a design background rather than a technology or engineering background and the thing that I find particularly interesting about what I do here, which includes leading the joint program Innovation Design Engineering with the Royal College of Art, is that the school, Dyson School, is all about bringing together design and creativity with science and technology.

Yeah, definitely. And what drew you into this field and teaching in general?

Stephen (Lecturer) (01:46:50.062)
Well, actually I was showing somebody the other day that I had in my career timeline, pretty much from when I graduated, I started getting involved in part-time teaching at various colleges. I always enjoyed that alongside full-time work in consultancy, but I think at certain point I realised that there was a more interesting future.

in focusing in higher education. So perhaps for the last 10, 15 years the majority of my time has been working in higher education.

That's great. So over the years there has been quite significant changes. We feel there's been maybe some significant changes, but in your opinion, what changes have you noticed in higher education and the way that students are learning in the time that you've been in education?

Well, I guess one of the distinctions, one I should make, is the difference between the traditions of teaching and learning in art and design subjects and traditions of learning in sort of perhaps technology and engineering subjects, where I think maybe it's a stereotype, but a lot of traditional engineering teaching.

and formal university teaching is done in lectures, lecture rooms such as this and students take part in lectures and at a certain point on their course they do exams to test how well they've assimilated the knowledge in the design field, there's significantly more emphasis on project-based work and personally I'm a strong believer in the

Stephen (Lecturer) (01:48:44.162)
the value of actually both acquiring knowledge but then directly applying it in project-based work. And I think that's what we're seeing also within what we're doing at Dyson School particularly now is more recognition of the value of project-based learning, learning in teams and generally perhaps

trying to align how students are learning more closely with their experiences that they might have in industry or in professional life.

What would you say is like what do you find most rewarding and most challenging about teaching students today?

Most rewarding of course is again the project-based nature of what we do. I find it hugely rewarding when groups of students or individual students are doing work and producing really exciting results with what they do. Some of the challenges definitely are to do with

I think how you, in a practical way, manage large groups of students and ensure that they have the best possible learning experience they can. And I guess that's one of the trends that there has been in higher education over the years I've been involved. mean, when I was a student, I was in a group of 15 students and we probably had two or three.

Stephen (Lecturer) (01:50:29.198)
literally full-time members of staff just dedicated to 15 people in the group and of course as you know you'll be in a group of 100 and you won't necessarily have the full-time attention of members of staff.

So would you say that you felt there was more of a benefit of having that closer connection with the lecturer versus what think the students do now?

I think there are benefits definitely of having more direct contact but that's not to say there aren't good alternatives to that. I was just talking the other day about the possibility of virtual tutors, know, because

in the world of AI we can start to sort of question, you know, what aspects of what we do can be replaced by artificial intelligence or machine learning or whatever.

In your experience, you feel that, well, what do you feel students today, do you feel students today face unique struggles compared to like past generations?

Stephen (Lecturer) (01:51:51.726)
I think the landscape is completely different. I mean, again, if I go back to my own early days of teaching or even my own experience of being student, a much smaller proportion of people went into higher education. Most people had quite a substantial amount of their costs covered by grants.

But these those, I think students under a very significant pressure, sort of both financially, but also with, you know, whether it's real or a perception, you know, much, much more challenging markets for employment. So I think students are feeling a lot of pressure. I think that those pressures that students are feeling, they were

further compounded by the very unusual circumstances of COVID. And so I think we're still seeing some long tail of COVID effects amongst students.

How do you typically identify or track where students struggle, especially if they don't speak up or attend the office hours and tutorials?

Well I would like to think in, I don't know if it's shared amongst all of you, but in the Dyson School I'd like to think that with myself and colleagues you know we do have quite good interpersonal links with the students and I think we know students quite well as a whole but obviously there are always, and especially as numbers get bigger, there are people that you know maybe get

Stephen (Lecturer) (01:53:43.754)
missed and obviously there are more formal ways that students attract through assessments and with timetabled personal tutorials and things like this. These are the formal check-in points but I think alongside that all of those informal points and the culture around a particular course or group of students is also important.

What kind of information would be most useful to you to see your students' progress on weekly or daily basis?

I think it would be good if we were able to have a very

quick, simple or to use the jargon, frictionless access to a range of sort of profile information, both about individual students, but also about particular courses that we might be teaching or the cohort as a whole. I think across the higher education landscape in the UK, probably what you find is different universities have

different levels of student information systems and I think Imperial, I have to be a bit careful what I say, know, the student information systems are patchy at best. You could edit that bit out probably.

Antanas (Student) (01:55:21.19)
So I guess what sort of metrics would you find like most helpful and then how would you apply that into your teaching if you had access to those metrics?

Yeah, it's an interesting question.

Stephen (Lecturer) (01:55:41.866)
I think to sort of think a bit more carefully about that question, I think I'd want to sort of explore an idea of overall what is it that I'm trying to measure in any given situation. So there might be metrics about the sort of example I thought you were hinting at earlier of students that might be struggling in any particular way.

can you identify those students? That would be one thing. But then another thing could be actually you're interested in the term I quite like is the learning gain. That's quite a complex concept, but it'd be really interesting if we understood from a particular group of students what does their learning gain look like?

taking in a group of very bright students and they're going through some instruction but at the end of it, know, what extent have they developed their knowledge, understanding and skills? You know, there's a whole bunch of things in there that we could be measuring. Sorry about that. It's hard to pinpoint one or two things. Actually, the reality is

And that's often the case, it, that when we're talking about learning, there's a whole bundle of factors and actually probably we do need a bundle of metrics that we can work with.

Let's say you're identified with a student who is struggling on a task. How would you typically support that student if they get stuck? What would be your approach if it was during office hours, teaching hours or outside of that space? What's your approach?

Stephen (Lecturer) (01:57:47.234)
Well, I guess the first approach is having the opportunity to have either small group or one-to-one type tutorials and often that is sort of designed into the design of the module or the course and then those sort of more direct contact opportunities then that gives you the opportunity to

provide more tailored support or advice. Yeah, and more detailed than, for example, students would receive in sort of things like written feedback or the result that they received on a particular piece of coursework or exam.

What's your take on how students usually approach complex problems? Like, kind of observing the students that you've worked with. When they face a really big challenge for themselves, how do they approach that problem?

I would say it's really difficult to generalise. I mean some examples of what I've observed about how students approach sort of complex, let's call it complex learning challenges. I think some students actually switch off with cognitive overload. Like it's too much and sometimes actually that is reflected in complaints. You some students will

find something to complain about might not be the complexity of the tasks that they've been given. They might say the workload's too high or so-and-so is a terrible lecturer. maybe the issue is that there's a mismatch between the complexity of the task and what the student can manage. That would be one example. I think what I see a lot at Imperial, which I think is a good thing, is sort peer learning.

Stephen (Lecturer) (01:59:50.638)
you know, if you're stuck with something, you'll ask your friends or the people around you and you'll see what they're doing and that will help with overcoming those challenges. Actually, sometimes talking to a tutor, again, you know, some students will always be the student that puts their hand up or seeks out the tutor and says, I'm struggling with this, can you give me some extra help? But then there's a whole bunch of other students where perhaps

talking directly to a tutor or seeking out tutors is their last resort.

Yeah, definitely. Kind of digging more into the more empty sort of questions is have you experimented with any digital tools or platforms to assist with tutorials, assessments, of what's worked, what didn't?

with specifically to support assessment.

Okay, not necessarily assessment, but I guess like the workload, the kind of assignments, maybe some homework that's been given, module content, I guess.

Stephen (Lecturer) (02:00:57.294)
To be honest, mean, maybe, yeah, I don't think I'm that behind with technology. But you you all know this, some of the tools that we're presented with, which in theory are there to help, Blackboard, for example, are just, and you need to use them, you understand the reasons why you'd use them, but they're very difficult to use.

Probably, you know, the things that one does use on a day-to-day basis will just be all of your standard sort of office software.

Yeah, Do you see a role for AI driven or data analytic tools to help you monitor your students progress in real time?

Yeah, absolutely. think, although, say for example with assessment, the idea of highly personalised feedback is really important, but I can also see that AI, well-designed AI, could be massively beneficial with supporting coming up with high quality feedback. But it's a bit like using AI in general, the output.

will only be good as the prompt engineering that you put into it. So if you have a well-designed system to support helping produce personalized feedback, that would work. But if it's poorly designed, that's much worse than if you'd put the hours in and done it the traditional way.

Antanas (Student) (02:02:39.918)
In an ideal world, would you like to see technology support your teaching? And if it is, what's missing right now in your teaching? Sorry, not in your teaching, in tools to support your

That's right, I won't read that as a criticism of my teacher. Yeah, no, of course, mean the nature of what we do in the school is all about seeing innovation opportunities and I'm actually personally I'm super interested in a whole range of ways in which technology can and I've been involved in a number of projects, I am still.

involved with a number of projects which are exploring how technology might support that. Actually what I do believe from the work that I have done is that some of the big challenges in introducing developments in technology into teaching practice are to do with scaling up any concept and then we could look at the whole concept of

scaling up concepts in more detail. actually one of the big ones there is getting colleagues to adopt any new approach. So even in area where people are in theory enthusiastic about technology and innovation, you'll find a lot of inertia and neophobia.

Yeah.

Stephen (Lecturer) (02:04:23.271)
And I see that that is a big challenge.

I'm going to give you bit of an explanation of what Amfy is then. So Amfy is a tool for students to receive tutorial sheets, problem sheets, any problem they have, break it down with Amfy. Write it on a canvas, talk to Amfy, text with Amfy and then it can provide feedback into that learning. So whether that be you're writing out a problem and you get stuck, Amfy can then help you understand that. You can write down your

problem and then I just ask Amphy about it, like how do I break down, how do I move to the next step, what's going to support me. Give it concept and then it can teach you it from scratch if possible. Give it some of the content that you've had from your lectures and then break down maybe some misunderstood concepts that you've had in that. And then from the lecturer side, obviously the students are using the platform, we can provide like really detailed, innovative kind of analytics on this. So whether it be through tutorial sheets as

they're progressing, what concepts are students working on, have they understood the concept over time when they're doing like external tutorial sheets, when they go through the lectures, is this all kind of understanding and the kind of analytics can be all like literally anything so from obviously understanding we've got kind of engagement, we've got like the kind of quantity of work that they're producing is this like all different sort of metrics that we can provide and we can give this straight to the lecturer so they can

do what they want with it or we're trying to work out is this something that in your opinion lecturers would use as a platform if they have all this analytics? Would they be resistant to it? Would they be comfortable with it? Would they trust it? Is this something that they're even looking for? I guess it's quite an ambiguous question but just give me your thoughts on that.

Stephen (Lecturer) (02:06:23.63)
Well, I would give you an example and I'm trying to remember the name of the platform. Does T-RAT and R-RAT mean anything to you? So the digital technology team within the engineering faculty here, they've been promoting some software for team working.

Sorry.

Stephen (Lecturer) (02:06:53.324)
within teaching and yeah, the T-RAT and the R-RAT, I think those are acronyms for evaluating students before an activity and after an activity. And it's a similar thing that the academic involved or the teacher involved create some questionnaires or questions that are built into

the evaluation and then when the students do the initial evaluation then the system is generating metrics associated with how well the group understand the questions, what they think the answers are and then the post activity one then just again captures all of the data associated with what's happening after that.

activity. So I was told about this and one or two other colleagues were and we were invited to try it out. So one of the very practical challenges is the learning curve associated with introducing any new things. So you as a myself as an academic, I've got to understand enough about it and trust it enough.

than to try it with students. depending on how much time I've been able to put into that, it may or may not work well with the students the first time. So you've got to do it again and again. all of that sequence is you're either building trust or losing trust in it. in the adoption, there's a very high level of opportunity for failure or even like small things with.

like an interface that you don't particularly like, you know, there's lots of aspects of Blackboard that people don't like, but actually they're in a situation where they have to use it. So you just use it, but if it was optional, people just wouldn't use it.

Antanas (Student) (02:09:05.902)
We're really trying to go for the design by students, for students aspect of it. We've designed this from the ground up to be something that students would like to use rather than something that's given to them by the university. But we're still going to go through the university as an adoption approach. really trying to get what users

I mean are there some interesting challenges in that? mean for example the sort of system scope that you describe requires the system to aggregate, gather and aggregate data and feed it back to staff. Is that something students would trust and be interested in?

think that's the whole scope that we're doing. We think that the, obviously there's going to be pushback on it, but think given the value that the students are getting directly based on the amount of like...

the frequency of use that they're doing and the fact that when they provide the contents to the lecturers it's only going to benefit the students. From what we've heard so far is that there's been good reception on it and slight negative reception but we're trying to work around that to get the best of like...

Okay, so what I mean is that here's a thought on that then so one of one of the factors is is I mean we would just talk about it today in a in a staff meeting is that you know there can be a

Stephen (Lecturer) (02:10:37.382)
mismatch between student motivation and even the most sort of well-intentioned tutor in terms of what they want the student to achieve. I'm sure you're familiar with this sort of thing. On the one hand, a tutor might want you to become really

expert and enthusiastic about the whole topic of robotics and optimizing software for robotic activities. A student is interested in getting the highest mark possible for the least effort. Those are two different overall motivations.

Yeah, it's definitely something to consider. We do feel like we've done, in our own opinion, good job balancing the requirements, or especially a better job than what the current solutions do. So I think that puts us in a fairly good position. But I'm going show you the demo.

Okay, yep.

We're going to hand you the iPad and essentially there's just two buttons that you need to know. So here's the pencil. You can just draw on the canvas here. And can you think of maybe a problem that might be for the students to be anything from like a verbal problem, like a written answer, or maybe a design problem. So think of, I'll give you an example. Like if you draw a phone with a banana shaped display,

Antanas (Student) (02:12:16.528)
And then let's see what Ampy does in that case. So if you wouldn't mind drawing a banana-shaped phone on the tablet.

right, okay. So you want me to draw a banana shaped phone or you want me to write...

I'd like you to draw up an ownership phone on the iPad.

If possible. Yeah, yeah, yeah. Well, I should be able to draw shouldn't I?

Stephen (Lecturer) (02:12:53.484)
Yep.

So maybe Anna take the banana with a few foam-like components.

Okay.

Antanas (Student) (02:13:36.782)
Take care.

Antanas (Student) (02:13:43.022)
Okay, so what I would like you to do is put yourself in the mind of a student. So this is your assignment project, just been done, to design a creative style of phone. But you're struggling.

because you've got to this position and it doesn't feel right but you're not sure what doesn't feel right about this banana-shaped phone that you've just designed. So if you wouldn't mind, not just yet, just at the end, pressing start session. And then this is what's going to happen. And then just tell Amphe. So speak out loud and go, hey Amphe. And then again, put yourself in the perspective of the students. Hey Amphe, I'm struggling, I've had this assignment, but I'm really not sure what's wrong with the phone.

yeah.

Antanas (Student) (02:14:28.238)
see how Ampy responds to try and help you with your design tasks.

Antanas (Student) (02:14:42.03)
speak up.

Hi, I'm Faye. I don't know if you can understand my sketch, but I'm really struggling with this really terrible tutor. Well, I thought it was the perfect location. That's part of the reason why I'm struggling with what the tutor said. just seems this is... I've designed it. I've met the brief. I've designed a banana shaped thing. The tutor seems to want something extra.

I think the tutor wants me to explore a number of alternatives.

Antanas (Student) (02:15:32.716)
Hey Amphi, can you see what's on the screen?

Antanas (Student) (02:15:46.731)
and

Peter has just said that the design...

Antanas (Student) (02:15:59.918)
But I'm really not sure why this isn't ergonomic enough.

Antanas (Student) (02:16:13.932)
Yeah, would you be able to help me?

Antanas (Student) (02:16:29.386)
Yeah, could you tell me whereabouts?

Antanas (Student) (02:16:46.616)
So let's maybe just leave it there. But you can see what actually went wrong there was Amphi didn't realize that there was something on the screen. So by me prompting it, it was then able to see what was on the screen and obviously know the annotations afterwards. But you can imagine the...

Yeah, yeah, but you can see the potential of it in the sense that imagine you're one o'clock at night, the lecturers obviously asleep and you know, you're really struggling, you're just doing a bit of work and you know, the deadline's far away but you just, you're really struggling on what to do next and you know, your friends are away so they're be able to help you that this problem could be extrapolated to anything whether that be a tutorial sheet or...

essay or something that you've written or a mind map and explore. So you can see how the potential of just being able to have this one-on-one tutor that's able to help you break down this problem and almost have this co-pilot for design to be able to help you. Again, not necessarily a design, it could be anything. So I guess, can you see the potential of this for a student? And kind of what are your general thoughts, I guess is the question.

Yeah, well, mean, it's impressive that you've been able to build out a prototype that will get this far. I guess some of the immediate thoughts I would have would be, you know, in terms of the database that is driving this, you know.

Stephen (Lecturer) (02:18:23.222)
sure that also is going to in some quite accurate way match what a tutor's expectations might be. And I suppose that's the, I mean we see some of these things already don't we with just.

using Chuck GPT to a certain extent, you'll get a level of answer, but it might not be.

need for a particular situation. that would be one thing.

Stephen (Lecturer) (02:19:08.366)
format of the output.

I mean in the way that chat GPT works, actually I quite like that quite often, you know that it will give you a list of stuff. And then that list of stuff is sort of like a checklist that you can decide to follow up or not. Whereas a conversational interface sort of seems to be a bit more...

linear.

Yeah, yeah, definitely. To answer the first point is that ideally what we would do is you would have the context of the lectures, the kind of transcript, the assignment, maybe the mark scheme to know what is the goal that the student's aiming for and how do we help them progress in that direction, whether that be the lecture content slides, the way that the lecturer would teach.

try and mirror that in the way that we teach, whether that be that the student maybe forgot what the lecturer said in a certain aspect and this is just reiterating what the lecturer said or whether it's elaborating on it and maybe filling in the gaps between some of the points that the lecturer has made. This would be a way of kind of, we feel becoming an extension of the university experience is a way of you know taking that contact time that we have just extending it outside of the space and maybe it's not you know even 20 % as good as a lecture.

Antanas (Student) (02:20:43.056)
but at least it's that 20 % that's always there and that's always able to help them. That's our perspective and guess, yeah.

That's what's up here, nothing, but the, but I am.

Stephen (Lecturer) (02:21:02.2)
concept of...

Stephen (Lecturer) (02:21:07.0)
terminology we might use.

Stephen (Lecturer) (02:21:13.422)
Is the one personality for the co-pilot or is that something that students might have some influence over?

Yeah, the more, so you can imagine this is a tool that would be used from first year to fourth year or third year depending on the course. So this is something that would grow with the student and as they learn it would know that they're

styles so for example if the student usually learns the best by breaking down the steps it would usually ask for that but if a student just goes can you give me the answer then maybe it would what's the word, not contradict, challenge the student in a way that would be like actually maybe you're trying to cheat the system here a little bit if you maybe try taking this approach instead of you know going straight to the answer but it wouldn't be obviously as verbatim for that but like it would it would still try and

take the user through that journey, that learning process, to make sure that every concept has been understood.

No, I mean it's got huge potential and ideas like this definitely are going to...

Stephen (Lecturer) (02:22:38.582)
scale is the some sort of aspect of what you're doing which gives it a particular...

Yep. So...

Antanas (Student) (02:22:58.796)
Is there any advice or message you'd like to share with other lecturers or institutions who are on the fence about adopting a digital teaching tool such as Amphi?

Stephen (Lecturer) (02:23:20.622)
These are developments that we're going to see over the next 5, 10 years.

Stephen (Lecturer) (02:23:31.63)
sort of take an active interest in how they're.

it's a good way actually also to as a vehicle.

See you later.

and definitely for reflecting on the students' learning experience and if you can find some low impact ways of trying these things out and exploring the possibilities, that's a good thing. Again, I think that's important for trying to get over some of these barriers to adopt.

I guess just wrapping up then, question is, looking ahead, do you see AI and this real-time analytics as part of an integral future education? Like why or why not?

Stephen (Lecturer) (02:24:26.294)
Well, just to repeat some of the things I've been saying. This is absolutely, know, stuff that's going to happen, isn't it? It's just a question of when and how it's going to happen and who's going to be driving it. And of course, I'm biased, so I would hope that you would be part of this.

Great, thank you. That's everything.

I think there was one netman student who had emailed me about it.

Stephen (Lecturer) (02:25:02.122)
today.

No.

So Tim Sadler was originally an undergraduate student and did the IDE course and he was developing one of the really early

security platforms and about 18 months ago he sold his company.

Stephen (Lecturer) (02:25:31.31)
less than 10 years from starting up. And it was great to hear from him because he was, and of course now he's in this sort of very privileged position of achieving all of this.

sort of encouraging.

Stephen (Lecturer) (02:25:51.018)
And he had lots of management speak cliches like if you can sell one you can sell ten, because if can sell ten you can sell a hundred.

team of people that you can work with and really trust.

perseverance to do things. beyond that he was saying that London actually now is one of the best places in the world to develop things. He said there's more than enough VC capital.

Stephen (Lecturer) (02:26:32.462)
and people. So if you've got stuff.

Stephen (Lecturer) (02:26:43.34)
tech business to any great extent.

Stephen (Lecturer) (02:26:49.87)
It's filled with a lot of big players that necessarily will be rather slow in developing things. Of course, they will have some advantages from being big, but this was another message from Tim Sadler, that being small and agile gives you unique opportunities.

So I think the challenge is how do you scale something like this? How do you make it happen?

That's exactly the question, that's what we're trying to do. Yeah, guess we're obviously finishing off the YAROM module, but I think we have a lot of...

Anyway, Tim Sadler, this is the second person I've said it to today, he seemed to be incredibly open to people contacting him.

Antanas (Student) (02:27:44.526)
Yeah,

Because really that's what he's now become expert in. He knows all the steps along the way to turn an idea into a hundred.

you

Stephen (Lecturer) (02:28:13.902)
That's super, super interesting.

Yeah

My personal tutors in the fourth year, they were saying that you don't pitch again now for the ERA, you're just submitting the various bits.

Yeah, yeah, that's it. It's just a website and a video.