 my name is Jim Turner I'm a white kind of nearly late 50s got gray hair I'm wearing a purple hooded top in front of some shelves and I'm currently chair of Ellie Sigg and Ellie Sigg is a community as I was saying before of tell technologists and researchers and academics and practitioners who want to come together to try and learn more about the learners experience in the in that world of learning technology so this session is going to focus on AI and neurodiversity and we've got a couple of different speakers and the community is really trying to connect up people and share those experiences and you can join us by joining gisk mail and finding us on there where we post out about our sessions but you'll also if you're members of communities like cedar and alt you'll also find us advertising on there and in a second we'll hear from Dominic who's going to give us because this is a roundtable event a virtual roundtable event what we're going to have is a little bit of an introduction from Dominic and then also move on to Sarah and Ross from City University so if Dominic while I set up you wanted to run a questionnaire so while I'm setting that up I wondered if you wanted to come on share your slides and I'll get your question up and running yes thank you hello good afternoon everybody I'm hoping I can share my slides and they should be coming up shortly there they are with a beautiful AI generated image of a robot holding a book in an ancient library with some nice patterns in the background as well to show us that we are in new and different times now I wanted to start off by maybe asking you all a question what is your experience with Chagy BT because we keep saying AI but we really sort of mean Chagy BT and and the others as well so here are the five options or four options I use it almost every day I've played around with it use it sometimes heard a lot about it or I've heard something about it not something that I know so I'm just seeing some responses I saw some responses coming up but they seem to have disappeared now okay I see I'll release it in a sec everybody's having a little go all right okay good good okay so so we're setting some responses and while you while we're waiting for the response to the poll I want to ask you next question to answer in the chat what do you think about Chagy BT what is it what are your impressions if you use it or not okay so we have a small percentage of people nine people who use it almost every day 47 have people have played around with it that's the majority and then some also people haven't heard a lot about it and some people have just heard a few things about it so we're seeing that they're in the in the numbers and rather than the percentages so in the chat what are your impressions what do you think about this so far just want to see what some responses are chad generated possibly true okay so that's that's a good one and biggest challenge my sense of ethical practice of amazing opportunities like the potential great mixed views some people yeah some things are good some experience is good some less useful when used on certain applications context very good okay a lot of promise wonderful okay that's cause academic integrity concerns loss of skills potential danger has used somebody okay so I guess I should have a dangerous or plagiarism are wonderful okay yeah so but I like what Brian says I saw the fear of the empty page right that's that's an interesting one that's was one of the first reactions I've heard from any people who are working on your diversity space somebody saying Google bar is been worse in Chagy BT okay well let's have a look about that useful primer for critical thinking wonderful because those are all great and interesting responses I'm not going to read through them all but you feel free to continue as you as you go through these so let me just kind of start just quickly in five minutes into a quick introduction to what I think is perhaps worse keeping in mind as we continue that discussion for the next hour so just to introduce myself I work at the Oxford University I'm the assistant technology officer and I also run the reading and writing innovation lab where I look at e-readers tablets stylists is reading apps writing tools but also many different strategies and people can visit me and feel free if you ever in Oxford let me know I'll be happy to show you around as well so I think really what I want to answer that question I want to ask ourselves the question what is your diversity really what is what are the barriers that we're worrying about right and so it's things with language money even though doesn't seem like your diversity but that's has definitely dimension and complexity is perhaps the one that's the more the most relevant and there's different things that are happening when it comes to your diverse there's cognition but there's organization both sort of mental physical socialization emotion but also perception that those all things kind of relate to each other so those are all things that we might be thinking about relate reacting to when we think about your diversity what is what is involved in that so in cognition that's those are the big ones right dyslexia issues with decoding attention ADHD understanding SLI speech and language impairment then we have organization again that those different cognitive organization issues task-based organization issues that have dyslexia ADHD share but also physical like dyspraxia socialization we're talking about people struggling with social cues or fitting in as a consequence of other issues as the emotional aspect there's directly anxiety and depression you know or source of issues but also there can be a result or there could be sensitive to stimuli there's lots of things that are happening there and we're interested in perception sight hearing but also touch and so on so those are all things that are happening there but also it's not very straightforward there are issues that are permanent temporary and situations so it's not always clear what what what you know what that means talk somebody has an issue or something and here's a quite well-known chart from Microsoft's design inclusive design principles so how does AI fit in right then so the first question you know we have the big three that I started off with and so this language money complexity and so does it help with language it's incredibly good at languages money despite people's concerned about all the costs it's actually much cheaper than many other things that are out there and complexity I want to show you is in a way the most powerful aspect of where AI can help in those sort of big three excluding vectors but of course we have to ask which AI are we talking about so I already said we mostly mean chat GPT nowadays right so there's this big circle of the so called good old artificial intelligence you know go go-fi and then machine learning is a subset of it so that's not all of it and then neural networks which is what we have is even a smaller subset so you know the old style AI was the old things this expert systems old style chat bots they're really out of the game span filters used to be called artificial intelligence Google search also things like that but also just basic spell check also that really isn't much what we're talking about nowadays what we're talking about is things like image labeling the really good spell check the grammar checking the fluent speech recognition and chat GPT and of course making images so those are the things that we really mean so nowadays there's the other AI only one development distinction and then there's the generated AI which is what we're talking about chat GPT images really natural Texas speech all these things and so what comes to your diversity you might be interested in solving some of these problems like helping with the with the cognition creating bullets outlined summaries organization plans outlines chart socialization providing people with script scenarios decreasing demand on time helping with emotional aspects of conversation motivational stories perception we can translate between different modalities and so how can AI help any of these well yes it can before chat GPT we were using things like speech recognition and Texas speech or even spell check those are incredibly helpful but nowadays chat GPT can help with all these other things in different interesting ways that we still have to discover how to actually make them fit into this support structure that can help with what I'm calling cognitive scaffolding and provide emotional support can create learning tools for us so let's just see a quick example of that so it's a really a new paradigm for how a computer can assist so for example I gave it a long article and asking me who are all the people mentioned in this article right so they gave me a list but that list was quite hard to read and you may be not as useful so I had to make it into a table and add to that table well how would it work were there from what is their role what they actually said and all of a sudden I have a way of approaching that article that may be really difficult for me to read otherwise because of various issues that I may have I have a structure I have this cognitive scaffolding to approach this but that can take very interesting unexpected forms as well so for example I've talked to a student who studies medicine so he uses Chajipit to create little poems to help memorize these long long lists with long long lists of terms and so on and here again we can use it to create examples and counter examples here's another example of something that I used recently used starting to read your reading process you're not helping with exact readings starting thinking about for example working memory in this case so I ask it make me a list of working memory terms but that was a long list hard to really deal with I say okay make me a sub list break into categories here it is but I was one of the categories was too long so can you make subcategories of that again it made them it's all that important here for me as a student that the categories are right it's important that it's helping me start in thinking building mental structures that I can approach complex concept and of course I can do even better I can create into a tape I can make it on the table with flashcards build relationships in that and again when I work with students I always tell them this is just a hypothesis you're testing these are not the answers but this is a way for you to approach complex issues and of course you can also take it further and apply with other tools so for example I ask it make me list of propositions then I asked it to make it into a table and then I took that table and pasted into C map is concept mapping tool and that showed me how those concepts relate to each other so Chagipi didn't make this for me but it made it possible for me to make it easier to start thinking about this in context that's really that's that's really the you know some examples of where I think this is incredibly powerful and my sort of last statement where I think it's just to get us thinking about we're probably not using it enough to help ourselves as we work and also our students so that's kind of where I wanted to leave my introduction where I'm coming from what I'm thinking about and and pass over to the other panelists so thank you. Thanks very much Dominic yeah that was that was a great start kind of I love this idea of cognitive scaffolding really a useful way of thinking about that. I'd like to hand over to Sarah Hopp from Citi to perhaps introduce yourself get on the mic Sarah. Hello can you can you hear me? Yeah, yeah thanks sorry I haven't used this platform before Ross has been trying to get in for the last 10 minutes but unfortunately he can't get into the meeting so hopefully it's just trying to to help him get in at the moment so apologies. So hello everyone I'm Dr Sarah Hopp from the Citi University of London and I'm the student disability and neurodiversity manager and I manage a team of neurodiversity study skills tutors and I also manage a team of D&ND advisors and the main change sorry excuse me the main change that we have sorry I've got really bad hay fever so apologies for the cough and the main change that we've made in the university in the department this year has been a switch over to a person centred support so rather than before someone would come in they'd give their diagnosis and they'd say these are the reason for adjustments that you can have and that's it and thank you very much off you go and what we've done now is the first two sessions of the student telling their story so they can have a holistic approach and they choose what support they want now what the pattern that's coming out from us at the moment is that students are not but not they're not encouraged at secondary school at sixth form college to be using assistive technology in AI as part of an integral part of their support and so what we're the challenge that we've had and that we've been working with Sandra's team in Leeds with is to try and create an environment where students are encouraged to make that leap that into to developing skills with assistive technology and lots of different things like that and so that they can they can be more autonomous and they can be more flexible in their learning etc and it's this is the real challenge that we've faced so many students have before they've had students coming in that are they've got learning support assistants that are human note takers that are human and they are just they just refuse point blank to say no we're not going to do it we're not going to use AI and there is so much as I've just seen in the previous presentation which is great there's so much out there in it that it could it's absolutely life changing for students and so for the students that are taking making use of it is really good so what we're doing is we've started to to create an HE transition program and what we're doing in September is we're going to be introducing we're doing workshops introducing students to assistive technologies different types of assistive technologies and AI as part of their integral support so that they can they can develop these skills over the time and and be more open-minded about accepting support I think it's it's a real travesty that in schools and colleges when you have a student has an education health and care plan that it's still assistive technology is still not in there assistive technology is still about speech just tech software and things like that and dragon dictate and they all have their place or it'll be have an extra laptop but there is no real focus it's almost as if you've got IT as a subject but there's nothing overarching there's nothing that's sort of embedded within and I think that really from primary to secondary level that there really needs to be a push with education health and care plans that these skills that are so important to students are developed I'm supporting now PhD students who would quite easily be able to complete their dissertations and they don't need human note takers to human transcribers and yet they are assisting on having them it costs the university money it costs it you know it takes up so much time and yet there is so much sophisticated software out there so the challenge for us really is over time it's building in workshops building up skills so that students have more autonomy and so that's and that that's where we are at the moment we we really are at sort of the the embryonic stages of that I don't know whether Ross has managed to get into the meeting yet or not yeah not not sure about Ross but as soon as he joins we can add him to the conversation but that was really interesting let's get and we're all at this embryonic stage you know I'd I think I'd like to start with Emma and Jess if you can catch up with any questions that have come through and and please anybody contribute this is your chance now to ask or raise pertinent questions and and points but I'd like to start with obviously this is moving at breakneck speed and you know we're struggling as Sarah and Dominic pointed out you know struggling to catch up are we jumping too fast for instance this idea of replacement of note takers these sorts of things is this going to take a while to bed in or it's it's so obvious of the benefits that we just you know where might be the risks I guess and I think I think the risks are I think there's things that I think it's second primary secondary level everything is so funding driven that and not student-centered so I come from my previous job was as a Senko in a sixth form college and I mentored Senko's trainee Senko's across primary schools and secondary schools and Senko the Senko job is really one that is it's um you know they're the unsung heroes really because it's never ever ending and the volume of it is so much but when you have and there are a lot of colleagues in local authorities who are also working very hard but they're not seeing every the paperwork is this is centered by this is a team around the student a team around the child but the team really isn't really focusing on the skills that the child will need as an adult that the child will need to be autonomous and AI is a massive one there so I think it's difficult because the risks are that schools don't have enough funding for AI when when the government's saying well there should be a focus on AI there's focus on coding and everything else like this so that's that's one risk so financial risk I think the other risk is is trust so the problem is is that by the time students come to us at university they're adults and and they've learned one way of learning and that is relying on another human so the the risk then is is gaining that trust now we have students in for like one or two but but years or three or four years that's not enough time really to be able to to turn someone around and get them to use assistive technology as possible and we do do it with some students but I think it's it really needs a sort of right from primary age all the way through to second and unfortunately it's something that's only going to take time and I think it's going to take a lot of lobbying to government as well I think I just passed that over to Dominic and then I think we've got some questions coming in go on Dominic yeah so this there's two there's two aspects to this uh and I think some of the things that people are mentioning uh mentioning in the chat as well is is the the speed of the change right and the and the difficulty of knowing what's coming what's coming next and what can be done and what cannot be done so we can and the biggest danger here is extrapolating in the wrong direction from what we know AI can do so I for example you know we can we're quite amazed by what speech dictation can do you know when we dictate nowadays with the the new improved technologies but then we're kind of appalled how bad lecture transcription is still automated lecture transcription is because sometimes it's hard to extrapolate for that and in a way I've had that from the very beginning of AI people of speech recognition people are always coming to me I can dictate into my computer I have a recording I'd like to have transcribed and and now it's at least decently usable but five years ago it was completely unusable so it was a huge disconnect between on the one hand seemed like it can do it but on the other hand the inference you make from it was completely wrong right it just you know that was not your fault but essentially that was not you wouldn't know that that you cannot take the next step and so a lot of the things that we're seeing right now is of that nature we're seeing these amazing things that are happening and we're also seeing some sort of silly errors happening as well and we're trying to find a way of thinking about the future how will we extrapolate from that into the next step what can it do and we're in a stage of discovery but it's a very strange kind of discovery in the old days you know when we had a vle or a new piece of software it just kind of played around with it and you figured out what it can do and then you learn from other people how they can use it but with these new tools we don't we have another layer which is people thinking about a completely different way of thinking about them and so things that were true in in February about what chat gpt can do for you are not true anymore not just because of advances in technology but because of advances in people's thinking about what can be done with it but also in in terms of other tools that are happening that are integrating with it and the problem is that there is in one place that you can go and learn all of this these updates are happening in on reddit on twitter and blog posts youtube videos there's newsletters there's there's sas tag and so all of a sudden if you want to keep on top of it which is what i'm trying to do this is not something you can just rely on the news reports here and there or some sort of an easy checklist of things you can do you cannot cannot do so so so the the the journey from knowing it exists to knowing how to use it in a useful way is quite is quite difficult and complex one yeah so if yeah it does sort of summarize in those two points that we're still in that discovery phase but there's i guess expanding on sarah on your idea of trust i guess you know we're all trying to develop that in many different levels aren't we at the moment emma i wonder if i could lean on you i wonder if there's anything happening in the chat space that you wanted to bring forward if you want to jump on the mic emma and i see we have raised hand as well jim i don't know if we oh okay yeah sorry while we're waiting for emma might do you want to come yeah i mean i put it in the chat but um ai literacy i mean at the moment we seem to have prompt engineering literacy you know chat gtp3 literacy chat you know ai bars i mean where are we going with this um you know that the developers of open ai so they don't quite understand how it works even though they built it um you know how much do people actually have to understand i mean if you take it's very hard to come up with an analogy for this for example if you hire a car um you can get any car how long do you spend working out how to drive it before you get in and drive away because they all work in the same way and they're designed to work in the same way um so where do we go here how much do people actually have to understand how the ai works to be able to use it sarah do you want to start with that one could you repeat the question i couldn't use my reception was going in and out could you repeat the question please yeah sorry do you want to how much do people have to understand how the particular ai works because it's dominant pointed out i mean you know there really isn't a definition of ai it's you know changing all the time so how much do people have to understand because you know i've been on so many webinars about chat gpt and even though people are saying it's just generating text it's not searching you know for information uh most people don't seem to even understand that even though they've read it and been told it so many times um so how much do people actually have to understand how the ai works and its limitations i think it's quite important for people to understand how it works um and how um and the limitations of it as well and i think for for us in our services i can only speak for us in our services and i know that we use um like those types of chats in a well-being app that we have for that in that role it is useful in a very sort of initial very initial type of of sort of dialogue so to speak um but the unfortunate thing is a lot of our students think and especially our autistic students think that this is the most effective way of communicating for them because of their anxieties and what happens is what's missing with human dialogue is that when we um as specialist practitioners when we discuss things with autistic students we're actually um reading in between the lines we're looking at um we're analyzing facial expression body language and things like that so um so the the use of those sorts of chats it it does have its use but it has its place as well and it can't really replace human contact and it's um you know i noticed in one of the chats one of the comments was about it taking taking away creativity i think that ai has its place and it's you as long as it's used as a tool it won't take a creativity away i think if anything it will enhance it um it's but it's the way in which it's used and because for that there needs to be understanding of what it is how it's used and how it's used specifically and for that i think there needs to be a lot of training both for staff and for students and again like i say in much in much earlier generations and not journey in much earlier in the i'm much earlier journey sorry sensei so it's sort of like primal school and secondary school Dominic have you got anything to add yeah i was was several things so i think uh the question um you know from mike has sort of many dimensions that are worth that i'm dealing with every day i'm thinking about every day and and we're at the moment starting to think about how can we support as the center for the channel learning at the university of oxford the rest of the university what is our place in this but also what is our place as individuals who are doing the support in learning and what what you know because obviously uh there is no one easy access to expertise that's as i mentioned earlier so there is no one person who you could say this person is an ai expert uh by dint of what they did two years ago so for example as again mike mentioned the makers of chad gpt didn't i sort of dispute the fact that they don't know how it works they know how it works but they can't quite imagine all the things it would be able to do once different people try it and that was their intention they didn't quite realize at the beginning once people see the the early power of it what they what else they'll be able to do with it and and and so uh for example we didn't think it would be good at labeling attacks but now people have figured out how to use it to label things so those are all sort of interesting things that are coming coming from it and the but the other question that mike asks is what is this like what is a good metaphor what is this like what is the other thing chad gpt is like and i in my other life actually work on metaphors this is something that i study so i think a lot about metaphors if you want to go to metaphor hacker to see me write about some of these things you can go you can do that and i the main thing about metaphors you need always more than one to help these things through some of these difficult issues and and with with chad gpt and the generative ai in general we need loads of them we need to kind of switch between them all the time and think about in some ways it is like this in other ways it is like like something else so you know in some ways people mentioned in the chat the the worries about the skilling so we can think about is it is it kind of like calculators in the old days we had really strong worries about calculators and they they haven't come true in in many ways as a as a society decline in numeracy uh quite the opposite in in some ways and you know saying goes for spreadsheets and things like that but they're also very much unlike calculators in the fact that they're open-ended they they can be doing different things and one of the things that i would like to uh bring up is the comparison with the previous ai hypes so we heard you know many of you may remember ibn watson that was being touted as the future of of medicine and i just last month the business school inside released a report uh actually let me share my screen just for a second just to show you what that what it says uh and and in that report what they what they said is well ibn watson prom made a lot of promises about uh about what uh it can do what this ai is and there was a huge gap between what happens in the lab and what happens in the field and it turns out the lab was just completely not predictive of what happened in the field but it was a product that the sales force of micro of ibn was selling really hard they were making all these claims that sales people made and that the management of the hospitals that tried to use it was really not taking into account all the stakeholders there so if you compare this to the chat gpt hype right this is different all of a sudden it's the utility of chat gpt is really obvious to its users they are the ones who made the hype really it didn't come from the companies it came from people trying it and seeing i can do useful things with it and also the labs are actually quite honest about the limitations you know open ai every time they release a new model they say this is where it fails these are the scores here's the benchmarks and everybody else is doing it and you know so we have relative amount of honesty here compared to the previous hype and but it's also very much grassroots driven it's not the management the universities and schools and and institutions who's thinking we must use this it's the people who you know who won't help who think they it can help them so that's the big so i think that's kind of a useful metaphor there as well that it's we need to think about this as this in the in the context of other hype technologies but also also think about it as this as a quite a different thing that we're seeing people and look at what people are doing with it and and help them you know and and learn from that but that's not easy because that takes time it takes engaging with the right communities finding the right people to follow think about it uh in in in complex ways so it's almost more like introduction of computers themselves or introduction of the internet where it was new to everybody everybody was trying to figure it out and it took a lot of effort remember the 90s and and then the early 2000s how much effort we put into training and upskilling in all these technologies and we just may have to do that again because this is so new and we may and maybe even harder this time because it's changing so fast so this is not just like one new thing like VLEs or interactive white boards this is a completely new paradigm possibly you know well in that that's one of the ways that we can look at it yeah that's brilliant thank you Dominic um Dustin's been waiting oh do you want to come on the mic hello can you hear me yep okay thanks very much um i have a lot of concern and this reminds me of COVID-19 there's a sense of paralysis at least among some of you and what you're saying in the comments so i have a question because i'm only 43 years young the bells are much older than me what did you do when computers were developing and getting more and more powerful what did you do back in the kind of 80s and 90s and i asked that because that changed a lot now i grew up with that i grew up with VHS all these gaming systems all these computer systems so i see chat gpt as an annoyance because it's making you guys anxious it's making me because you are anxious and i'm like why don't you just get on with it why don't you try it why don't you play around with it and martin Compton is a great source of some good examples but i guess the question is what are you doing to get to know it to put your anxieties to the side and to see that by the way as a person of ADHD i have ADHD it can be really helpful for tackling certain things for getting a different perspective you know when doing various things the tone thing was right on the mark someone mentioned tone checking an email with tone it's great at that it's not perfect but i can't go on a course to teach me that because work's not going to pay for that okay so what are you doing yeah so sarah uh in terms of trying to find the connection between AI and neurodiversity what what sort of everyday things are you doing so like i said before we're sort of we're running workshops so first of all we had um at the beginning of the year we worked in collaboration with the sandra partinson's team um lead and we worked with um that they gave us a lot of training and we gave them training on neurodiversity so interdepartmental collaboration um and that then was the beginning of working together um sort of and thinking about things like possible workshops for students workshops for staff um and then creating sort of thinking about different spaces so for example he transition days welcome days where both departments could have a presence um at all those sorts of functions where students could come up and ask questions as well um and sort of um trying to to build on sort of starting off just trying to sort of build packages of support together so for example our students have what's called a student support plan an ssp so the ssp would put in about um using the at room the assistive technology rooms um and they're having an induction so we would make sure that that induction was part of that student support plan um that an induction to using captioned for example was part of that plan as well so it's making so it's not just putting down the assistive technology or the ai on the student support plan and saying this is this is what the student is entitled to but it's actually putting down on on those support plans the um the induction for it the workshops for them and things like that things that will enhance for the doctoral students we've come up with a program of a series of workshops um and one of those workshops is focusing on ai and transcription services um that are available um and things like that so so that's that's where we are at the moment but there's there's so much more to do um really on that bit but really like I said before it's at its embryonic stages Dominic do you want to expand upon that yeah I well in a way sort of let me just put up my screen again for quickly so Dustin really kind of uh reinforce the point that I was I was trying to make right and and and the I really think this is the important challenge what are you doing and one of the things I'm a bit worried about we keep saying oh we must support our students in using ai correctly but I don't think uh the we we don't quite know yet how to do it so uh so I want to so this is the point that I ended on you know we are probably not using it enough to help ourselves or our students so somebody mentioned the chat that they find it to distract annoying unreliable and those are all true things but maybe but also somebody said people are not using it because they've tried it and it gave them a stupid answer and that's true but but the problem with it is that sometimes you need to start thinking and developing ways of interacting with it so there's sort of two different initial reactions people have to one is it's inaccurate unreliable probably more trouble than it's worth and the other one is well actually it's helping me with the things that were too difficult and the checking of it we're going from a place where checking things was more difficult uh or fixing things was more difficult than uh than doing it yourself but now where many people are finding that actually all the work it's doing is helping them so let me just show you an example of a conversation that happened yesterday uh that sort of showed me how I'm underusing chat gpt myself so I was talking with a staff member about we're having this project about screener use and she asked me can you give me a quick link so that I can start learning how to use a screen reader and I said you know what I haven't really found any good links any good descriptions for for that kind of usage right there are lots of guides for people doing professionally but nothing that's a good introduction so what I did after we hung up I said you know what let me ask chat gpt and I used the more advanced model uh for that and it gave me sort of a fairly vacuous response to the first bit you know so it was the best way to get started with a screen reader but I said you know what I now have one of the tools recommended how to use it and what it gave me is this wonderful step-by-step and I've and I happen to know nvda enough that I know it's all correct at a glance and then and then uh it gave me more information when I asked for it so I keep on asking expanding and then after a while it also gave me when I asked it a really nice outline of how I could do it step-by-step in six 30 minute sessions so so I can have that curriculum and I can create a table of the shortcuts that I needed for that just like that and what I could use then is new feature of chat gpt where I can share a link of my chat so I took that link copied it and I sent it today to the person and what they can now do at the bottom is this new button that says continue this conversation and so so she can then now go and use chat gpt if she has an account she can just use it as a as a guide that's a good guide or she can go and use it to ask further questions and elaborate on it and keep continue learning from it so all of a sudden this is me underusing chat gpt when I could have used it to create a tool that I've always wanted to create a guide like that I just never had the time so all of a sudden immediately I I have a good starting point and and something that I could use immediately so it's an example of of a really used but it's really because it I'm constantly trying to look at others and learning from others how are you using it what are you finding the strong points and that's really kind of the necessary thing so it's not so one of the dangers in there is to just say we need to go try it because just trying it is probably not going to get you there it's you need to kind of learn from others as well there's just so much happening so my one plea for everybody is to if you if you'd like to there's the one piece of advice I can give everybody take a prompt engineering course right if asking chat gpt things is called prompt engineering that on Coursera there's a really good beginner prompt engineering course for chat gpt takes you through the steps and and that's a good place to start right it doesn't do away with all the problems chat gpt has about hallucinations and so on but it will it will let you learn thing you help you learn things like how to how to prompt it better so that you can go from a from a result isn't very good like I can tell you more about this where you prompt it give it examples and all of a sudden can give you a result that's much more like what you wanted so that's something that I would I would recommend for people to take this seriously this is a learning process but even if you take that course that is not going to be the end of it it's just the beginning of a journey and going back to what Dustin said this is a lot right I know I'm asking you a lot but also this is a lot of new stuff so that's kind of we're back in the 90s where all of this because I was there in the 90s and I I was taking courses and teaching courses on like introducing people to computers and so I know this is not a small ask but that is the that is really the the demand of the moment I think and perhaps that relates Rob asked in the chat has anyone completed an equality impact analysis or data protection guidance for using AI tools do we inform staff and students what happens to the data they submit are all tools accessible for all do they increase the digital divide a lot to unpack there let's start with Sarah I don't know if you can see that one in there do you want me to repeat the question Sarah or you might yeah sorry please I've got this a storm my end and it's going in and out all the time sorry here we go has anyone completed an equality impact analysis or data protection guidance for using AI tools do we inform staff and students of what happens to the data they submit is this something that you're worried about Sarah in the work that you're doing at the moment I think it is I think it's a concern we've we're just moving at the moment to something called Supported City so it's in it is a completely new platform and one of the things that we're concerned about the staff are concerned about my teams are about sort of the the security of it and also about sort of collecting data and how that data is collected etc and I think that and it's partly because we the creators of this program because they are well informed about IT and AI etc and as you can tell I'm not of that ilk so it's it's very easy for them to trust the system so to speak but for us it's it's much more difficult and I think as well going back to the training Sandra's team are fantastic and the way that they told the way that they informed us the way that they trained us on the the use of the AI was it it was very much because we're creative thinkers and we think all over the place and things like that I'm not putting I'm generalising here but for me speaking personally as a neurodivergent person who's dyspraxic it's very it was very easy for me to understand because it was coming from someone else who was thinking creatively I've been in other training sessions where the person has been very knowledgeable but it's come from a very systematised way of thinking and that for me would be very difficult to understand and you know that person blessed them has sat there with me and broken it down so much and they've said to me you know and you can see on the face it's like what don't you get but for some reason for me if it's too systematised and it's too sort of if it goes by steps and the flowcharts and things like that it just will not compute with my brain I don't know why so I think anything sort of designing training needs to have in mind bear in mind the different types of thinking so you know a lot of our students autistic students are very good at systematisation they're very good at finding patterns and things like that and coding and that sort of thing but there are others of us like some of us with ADHD or dyspraxics or you know who are very visual all over the place thinkers you know we're very messy people sort of thing isn't and so a very sort of creative way of getting the message across to us is it works really well and I think that's you know my very sort of like clunky way of explaining it but I think that's one of the difficulties that we've come across with our students we're part of the reluctance as well for staff as well as students is it's making that training accessible to everybody and different types of learning and understanding if that makes sense yeah and Dominic are you going back to the question are you concerned about data protection and these sorts of things no I mean not in any other way that I would be anything else right people sending emails you know I think my sort of my sort of the standard of comparison I always want to you know be aware of the status quo bias right so if we're asking people to send emails are we are they really are they any more secure than putting stuff into chat gpt and the answer is no so as long as we're happy to communicate with people by email we're kind of doing those very same things I don't think there's any extra dangers I'm not saying there are no dangers there are no so concerns to be had but as I said many of these mainstream tools are quite open about what they use the data for and what they don't you know they open AI now lets you delete all your old data they let you they're happy to exclude it they're not actually training live on your responses they say they may use it for fine tuning but they're not doing that so obviously many but the other concern about the reality of it is that many companies have banned the use of chat gpt because of this right it's it's entirely possible that if you're running a bit lots of trade secrets that may be a problem but what they're finding a lot of their users are using their own machines to do it because they're finding it so helpful right so that I think that that is just the reality of it so in some ways we're seeing you know we're it's a slightly new space but I'm not necessarily seeing anything that's vastly new in terms of privacy or data protection issues than the things we're using right now brilliant thank you so uh Nikki you've just put your hand up do you want to come on the mic we're getting into our last 10 minutes here so if people have got something burning uh issue that oh go on yeah hello just because I am a Zenco in a secondary school and you said you want us to be using more AI to prepare our students for further education I just want to know what I should be doing brilliant question Sarah I think that's a good question and I think having been a Zenco myself it's different I think it's it's really difficult because it comes from a sort of a it's systemically from from the school so the the principal or the head teacher should really be making it part of the the the curriculum and as a Zenco you've only got so much power there's only so much that you can do so I think if I was back in my old post what I would be doing is I would be contacting um sort of lots of um different um associations for example um diversity and ability um and looking into the different types of AI that that they put that they have for students and they have sort of like really good demos and things like that and I think sort of and then making that part of any sort of support plans as well so um sort of helping the students to access three demos and stuff like that and then seeing if in their one-to-one support sessions they can use some of that assistive technology the speech tech software um the organisational skills type software and stuff like that and then so it becomes sort of like part of almost like their daily practice but it is difficult when you know you've got different schools with different amounts of funding you don't always have the hardware or the software available and you don't always have the budget so it's really difficult to try to try and do that but I think what I would be doing is having a conversation with the head teachers um and even you know taking it to sort of like um governor's level sort of thing and it's for the for the importance of this because what I can see happening being now at university a manager at university what I can see happening is whether we like it or not AI is coming and the students are going to have to have these skills in in order to be able to cope at university um the problem is is that the government and and schools and colleges haven't put this infrastructure in which makes our our jobs as Senco is really really difficult I think that's pretty helpful Dominic have you got other things to add yeah and in a way to kind of also respond to what other people were saying in the chat because there's a bit of a discussion about chat is this actually useful for somebody who has an issue or not right so some I think some people are saying well this is actually the worst thing for neurodivergent people and I don't like the word neurodivergent people right neurodiversity is a concept about about a lot of people are different have different needs nobody one is you know everybody has different dimensions of neurodiversity so I think in a way we need to start thinking about what is that what what the individual can do with it so I can certainly agree that and I apologize you know my presentation was quite quick and speedy and perhaps confusing it to some people but I was trying to show examples of where people can develop this notion of some scaffolding and help and that's something that takes time so even though I show these things like quick flashes of here's this this this if I'm working with a student what I'm actually saying here are here's the first step this is here's something that you should be creating for yourself when you're reading a text for example a table like that or summary or bullet points or outlines or if you're writing an essay you should again be creating these outlines but these students find incredibly difficult so in a way I'm saying here is your first draft here's something that you can help you with that guide and it's a slow process so this is not necessarily something that even though I showed it in a flash this is what's going to help somebody in a flash right then so some people also have to think to figure out how can I make chat GPT not give me too much information because often it is guilty of that right it just tells you all this other stuff as well as what you asked for and there's ways of there's ways of using it for that and I'm just using chat GPT but that will be true of other things and it's also important to help them understand some of the issues around the prompt engineering what is what does it make sense to ask because somebody said in the chat asked to do count the number of syllables and that's the worst thing you can ask your GPT to do because it's it's it's absolutely awful at that and there's good reasons for it but it's not saying it should be good at it we just you know so that's one of those things that is important to to think about as well and so that's that journey and if you're trying to prepare a secondary school student for university well just prepare them for the fact that that they need to start organizing their thought and organizing their reading and organizing their writing and note taking in even in a more systematic way and if they if that they find it difficult now they're going to be finding it more difficult later and so if they can actually start developing these first steps of using some like chat GPT to help them with this and that can be beneficial but not as a as I may have made it seem you know apologize for this like just snap your fingers and you're done right the hard work is just beginning after you have that done but in a way maybe you could not have maybe before that you could not have even gotten started on the hard work because it would just be too hard to even deal with the text that you're trying to extract some information from so so that's kind of the that that's my perspective on this right this is always going reading and writing and and dealing with complex ideas is hard and it's always going to be hard and it's just hard because some people because some of these difficulties so if we can find some ways of streamlining this like I try to show some examples then that's all to the good but it's still going to be hard because they still have to do some of the hard work of thinking and and integrating that so I just want to make sure that I didn't give a mistaken impression of this is so easy now that we just can snap our fingers and chat GPT gives something to us yeah uh we've down to the wire now Emma's got her hand up Emma um can you hear me now because I think you couldn't hear me when I was talking earlier yeah we can hear you yeah good um there's a couple of questions that haven't been answered in particular one from Rene about assessment um now we've obviously only got a minute so I don't know what Sarah or Dominic could say in half a minute about assessment um but she has had it um in the question channel for quite some time that question I kind of wanted to get it in for her mm-hmm yeah I can maybe say one thing and the one phrase that sort of stuck with me is is the death of home AI is the death of homework so many of the things that uh that we thought AI could not meaningful it could not be meaningfully done by a student you just have to ask somebody else to do it if you want to cheat and now meaningfully contribute to your work and I think that's that's something to keep in mind is that in some ways we don't quite know yet of all the things that are going to come but in some ways we just start thinking about how uh how we can think about assessment in the age of AI and start assessing things in the sort of famous Vygotsky zone of proximal development sense assessing with AI rather than it rather than rather than without it and and finally enough the example I showed you if the link sharing a link of your AI conversation conversation with chagy bt is the one thing you cannot use chagy bt to help you cheat on because you have to do it yourself so in some ways it's one thing that if I ask a student to submit that I know they have they you know they have to do some of the work there unless they ask somebody else to do it for them so that's just kind of my you know one minute of response to that okay Sarah last point I think um I think as far as well because I'm a neurodiversity manager so my area of of that and I agree with my colleague neurodiversity does mean there is no such thing as neurotypicality everyone is neurodiverse but there is some of us that have different types of challenges if we're dyspractic dyslexic etc and so it's those cohort of students that I'm talking about and for me I think assessment needs to be discussed more about there needs to be you know about the individual and sort of I move away maybe from written and typed assessment for some students survivors different ways creative ways and we can use AI as a tool for this creative ways of assessing students there are a whole cohort of students that have just been missed out where they're at very talented very knowledgeable but because they don't work right in a certain way and they've actually sort of they've been missed and it's such a shame yeah and that's I think that's a ideal point to end on Emma don't you think or is am I missing an extra question no run out of time anyway yeah we've run out of time there were a couple