WEBVTT 00:00:00.520 --> 00:00:03.040 Hi everyone, before we dive in, a quick heads 00:00:03.040 --> 00:00:05.820 up. The Podcast Learning Festival is happening 00:00:05.820 --> 00:00:09.519 on February 26th at Mary Sumner House in Westminster 00:00:09.519 --> 00:00:12.880 in London next year. It's a day for people who 00:00:12.880 --> 00:00:15.500 care about how podcasting and learning come together 00:00:15.500 --> 00:00:19.719 to make work smarter and more human. You'll join 00:00:19.719 --> 00:00:22.199 a mix of creators, learning leaders and curious 00:00:22.199 --> 00:00:25.100 practitioners, all exploring what happens when 00:00:25.100 --> 00:00:27.960 we stop talking about learning and start recording 00:00:27.960 --> 00:00:32.259 it. Tickets and details are at podcastlearningfest 00:00:32.259 --> 00:00:35.700 .live. There is a link in the show notes. It 00:00:35.700 --> 00:00:38.380 would be great to see you there. Now, on with 00:00:38.380 --> 00:00:52.719 the episode. Hello everyone and welcome to this 00:00:52.719 --> 00:00:55.359 episode, the more content one of the Women Talking 00:00:55.359 --> 00:00:58.950 About Learning podcast. I'm Andrew Jacobs. We 00:00:58.950 --> 00:01:01.329 recorded the content one in March of this year, 00:01:01.429 --> 00:01:04.650 and our guests had felt that the time went too 00:01:04.650 --> 00:01:06.969 quickly, and they didn't cover everything they 00:01:06.969 --> 00:01:10.189 wanted to. All of them decided they wanted to 00:01:10.189 --> 00:01:13.250 record another episode. It's been tricky to arrange 00:01:13.250 --> 00:01:15.930 a day when they're all available, but with some 00:01:15.930 --> 00:01:18.370 assistance and support from them, we've managed 00:01:18.370 --> 00:01:21.689 it. Our first guest is Gaynor Aitken. Gaynor 00:01:21.689 --> 00:01:25.450 has been working in L &D for over 25 years, specialising 00:01:25.450 --> 00:01:28.530 in systems training. Currently working as a learning 00:01:28.530 --> 00:01:30.590 solutions lead for an international shipping 00:01:30.590 --> 00:01:33.230 management company as an L &D project manager, 00:01:33.489 --> 00:01:36.090 she gets the opportunity to make a demonstrable 00:01:36.090 --> 00:01:39.849 impact on L &D and wider business. Our second 00:01:39.849 --> 00:01:43.109 guest is Karen Curitan. After working in sales 00:01:43.109 --> 00:01:45.870 and marketing for over 20 years, Karen was an 00:01:45.870 --> 00:01:48.849 early adopter of social media. As soon as it 00:01:48.849 --> 00:01:52.290 emerged, she embraced it with open arms. Roll 00:01:52.290 --> 00:01:54.829 on to today and she has become a results -driven 00:01:54.829 --> 00:01:58.140 social media specialist. Her aim is to unpick 00:01:58.140 --> 00:02:01.239 the often complex world of algorithms and analytics 00:02:01.239 --> 00:02:05.400 and share her expertise. Our third guest is Katie 00:02:05.400 --> 00:02:08.629 Swandell. Katie is currently working at Immediate 00:02:08.629 --> 00:02:10.830 as a Senior Learning and Development Partner 00:02:10.830 --> 00:02:13.629 and has over a decade of experience in designing 00:02:13.629 --> 00:02:16.969 and delivering strategically aligned experiential 00:02:16.969 --> 00:02:20.030 learning. She began her career in teaching before 00:02:20.030 --> 00:02:22.469 moving into apprenticeships and developing leadership 00:02:22.469 --> 00:02:25.370 and management qualifications, including collaborating 00:02:25.370 --> 00:02:28.969 with the CIPD to shape HR and L &D qualifications 00:02:28.969 --> 00:02:32.930 at all levels. KT champions inclusive real -world 00:02:32.930 --> 00:02:35.389 learning that empowers individuals and drives 00:02:35.389 --> 00:02:38.090 organisational success moving beyond traditional 00:02:38.090 --> 00:02:42.909 models. Our final guest is Laura Giles. Laura 00:02:42.909 --> 00:02:45.930 is Head of Learning at DESC, a Sheffield -based 00:02:45.930 --> 00:02:48.969 digital learning agency who design award -winning 00:02:48.969 --> 00:02:52.280 experiences for world -leading brands. She's 00:02:52.280 --> 00:02:54.639 been on a mission to fix e -learning's bad reputation 00:02:54.639 --> 00:02:57.860 since 2009 when she took her first job as an 00:02:57.860 --> 00:03:00.219 instructional designer and came face -to -face 00:03:00.219 --> 00:03:03.500 with the dreaded Page -Turner course. Today, 00:03:03.599 --> 00:03:05.879 she draws on the fields of cognitive and behavioural 00:03:05.879 --> 00:03:08.800 science, user experience and storytelling to 00:03:08.800 --> 00:03:12.280 create learning that truly sticks. Recorded at 00:03:12.280 --> 00:03:15.759 the end of October 2025, this is Women Talking 00:03:15.759 --> 00:03:19.340 About Learning, this is Gaynor, Karen, Katie 00:03:19.340 --> 00:03:24.800 and Laura. talking about more content. All right, 00:03:24.800 --> 00:03:27.280 well, good morning. It's brilliant to see you 00:03:27.280 --> 00:03:31.659 all again. We spoke last in March for the original 00:03:31.659 --> 00:03:36.699 content podcast, and we had so much to talk about 00:03:36.699 --> 00:03:39.919 that we've been invited to speak again, which 00:03:39.919 --> 00:03:43.159 is great for me. I always have so much to talk 00:03:43.159 --> 00:03:45.400 about in the realm of content. How are you all? 00:03:45.969 --> 00:03:48.870 Good. Good, thanks, Zora. And I think you're 00:03:48.870 --> 00:03:51.689 absolutely right. We agreed that we could have 00:03:51.689 --> 00:03:54.110 talked and talked and talked because it's such 00:03:54.110 --> 00:03:58.009 a fascinating topic, isn't it? Yeah, and it's 00:03:58.009 --> 00:04:01.830 always changing. There's always something new 00:04:01.830 --> 00:04:03.669 happening with content. It's never something 00:04:03.669 --> 00:04:06.810 that just stays the same. So it's really, really 00:04:06.810 --> 00:04:08.770 exciting to work with content, particularly, 00:04:08.830 --> 00:04:11.889 I think, at this time. Yeah, and I think that 00:04:11.889 --> 00:04:16.939 with AI, obviously, rapidly developing and changing 00:04:16.939 --> 00:04:21.139 the realm of content there's no matter what even 00:04:21.139 --> 00:04:23.819 if we spoke every week there would be new stuff 00:04:23.819 --> 00:04:27.980 to talk about I imagine Yeah, I was trying to 00:04:27.980 --> 00:04:30.879 remember, Katie, if we spoke about AI at the 00:04:30.879 --> 00:04:33.139 end of the last podcast, whether it was when 00:04:33.139 --> 00:04:35.779 the podcast ended, because I do remember we kept 00:04:35.779 --> 00:04:38.199 chatting for about half an hour after the podcast 00:04:38.199 --> 00:04:40.819 stopped recording. And I think Andrew had to 00:04:40.819 --> 00:04:44.660 kick us off of the podcasting app. So this is 00:04:44.660 --> 00:04:47.420 why we were so keen to get back together. But 00:04:47.420 --> 00:04:50.639 it is like Gayna says, I think it's a topic, 00:04:50.720 --> 00:04:54.319 even like it's such a fast moving topic. I think 00:04:54.319 --> 00:04:58.100 if we spoke again. in six months time, we would 00:04:58.100 --> 00:05:00.339 have even more different elements and layers 00:05:00.339 --> 00:05:02.740 to it. And I think that's the thing is from a 00:05:02.740 --> 00:05:08.519 learning and from the two elements of it, I think 00:05:08.519 --> 00:05:10.980 there is so much happening now. I mean, I see 00:05:10.980 --> 00:05:13.720 it and we touched on this the other day, we were 00:05:13.720 --> 00:05:17.120 chatting and I've got clients who when we spoke 00:05:17.120 --> 00:05:21.740 six months ago, yeah, six -ish months ago, wouldn't 00:05:21.740 --> 00:05:25.910 have touched AI. were terrified now they're almost 00:05:25.910 --> 00:05:28.589 seeing it as this is a friend that can help them 00:05:28.589 --> 00:05:32.269 yeah and i think that if you look at the mckenzie 00:05:32.269 --> 00:05:35.930 report that when we spoke i believe it was around 00:05:35.930 --> 00:05:40.430 the sort of mid 60s as um sort of ai being used 00:05:40.430 --> 00:05:45.089 inside a function or bau and now already um from 00:05:45.089 --> 00:05:47.110 going from like pilots to mainstream it's up 00:05:47.110 --> 00:05:52.290 to 71 percent wow so um my company we we have 00:05:52.290 --> 00:05:58.790 uh AI adoption, our BAU is 86%. But we've done 00:05:58.790 --> 00:06:01.889 a lot of work since it came about. We have our 00:06:01.889 --> 00:06:04.589 Summer of Learning program that happens each 00:06:04.589 --> 00:06:08.269 year. And we've done a lot of multi -touchpoint 00:06:08.269 --> 00:06:12.889 preparation for AI. But the fact that since we 00:06:12.889 --> 00:06:19.089 spoke in March, to have gone up by 5%, 6 % into 00:06:19.089 --> 00:06:23.189 the mainstream is huge. And so... There's a lot 00:06:23.189 --> 00:06:26.870 of content around what content do we produce 00:06:26.870 --> 00:06:31.769 to support people in the best ways of adopting 00:06:31.769 --> 00:06:35.629 this and working alongside and with AI as a critical 00:06:35.629 --> 00:06:38.649 thinking partner. But also what do we need to 00:06:38.649 --> 00:06:41.610 be aware of, of what content is out there, what's 00:06:41.610 --> 00:06:44.649 being produced and where do we actually need 00:06:44.649 --> 00:06:47.149 to put our weight and effort? Because if there's 00:06:47.149 --> 00:06:49.730 already so much mass production of X, Y and Z 00:06:49.730 --> 00:06:53.480 and content is being churned out. We don't need 00:06:53.480 --> 00:06:57.740 to add to the noise necessarily. Really? Yeah, 00:06:57.779 --> 00:07:00.139 I agree, Katie. I remember actually last time 00:07:00.139 --> 00:07:03.180 we spoke quite a lot about does learning design 00:07:03.180 --> 00:07:06.759 have a content problem? Is there too much content? 00:07:07.040 --> 00:07:10.339 And again, we spoke about content dumping and 00:07:10.339 --> 00:07:14.560 too much information. So I think adding AI to 00:07:14.560 --> 00:07:16.800 that space is really interesting, really, when 00:07:16.800 --> 00:07:20.279 you think about, well, is this knowing that so 00:07:20.279 --> 00:07:24.620 much of AI in learning? at an L &D is being used 00:07:24.620 --> 00:07:26.920 for content creation, you think, are we adding 00:07:26.920 --> 00:07:29.060 to the problem? Are we making things better or 00:07:29.060 --> 00:07:34.759 worse? I think it'll find its equilibrium point. 00:07:34.860 --> 00:07:38.920 I think it's almost like it's become a buzzword. 00:07:39.019 --> 00:07:41.579 And if you ask the majority of people, what is 00:07:41.579 --> 00:07:44.459 AI? Probably a lot of people wouldn't be able 00:07:44.459 --> 00:07:46.040 to tell you. It's just something that's appeared 00:07:46.040 --> 00:07:49.769 in... they'll know about ChatGTP, it's appeared 00:07:49.769 --> 00:07:51.910 in Canva, it's appeared in Rise, it's appeared 00:07:51.910 --> 00:07:54.009 in all these products that potentially they're 00:07:54.009 --> 00:07:56.910 already using. So it's almost like it's become 00:07:56.910 --> 00:08:01.910 part of the way content developers work without 00:08:01.910 --> 00:08:04.310 them really having to think about it. Yeah, I 00:08:04.310 --> 00:08:08.449 see that a lot with, so a lot of my clients are 00:08:08.449 --> 00:08:12.029 people who are a little bit... frightened wary 00:08:12.029 --> 00:08:15.569 of social media and when they heard AI come in 00:08:15.569 --> 00:08:18.089 they were frightened and now they've actually 00:08:18.089 --> 00:08:20.730 had a complete pivot I think it's my word of 00:08:20.730 --> 00:08:23.610 the week a complete pivot a complete about turn 00:08:23.610 --> 00:08:27.569 and we've done some exercises where I've shown 00:08:27.569 --> 00:08:31.670 them how they can use AI to save them time and 00:08:31.670 --> 00:08:36.970 energy and headspace And all of a sudden, they're 00:08:36.970 --> 00:08:40.029 totally obsessed with it now to the point of 00:08:40.029 --> 00:08:44.570 saying, you need to create a custom GPT. But 00:08:44.570 --> 00:08:47.990 for me, as somebody teaching people to use social 00:08:47.990 --> 00:08:52.070 media, what I've found is that... They almost 00:08:52.070 --> 00:08:54.269 hook into it. Oh, this is easy. This is going 00:08:54.269 --> 00:08:56.549 to be easy. But it's getting that. And it's like 00:08:56.549 --> 00:09:00.230 you said, it's about they're not losing. I say, 00:09:00.330 --> 00:09:03.190 you know, their unique special source is what 00:09:03.190 --> 00:09:06.690 I call it, is their personality, who they are, 00:09:06.850 --> 00:09:09.570 because there's only one version of them, whereas 00:09:09.570 --> 00:09:13.389 AI is going to create multiple, multiple, multiple. 00:09:13.710 --> 00:09:15.789 And I think that's the challenge. And from a 00:09:15.789 --> 00:09:18.750 learning point of view, you know, when we're 00:09:18.750 --> 00:09:22.279 creating. new workshops. I think that then becomes 00:09:22.279 --> 00:09:25.299 another challenge, doesn't it? So, Karen, do 00:09:25.299 --> 00:09:27.799 you think we're losing, we're in danger of losing 00:09:27.799 --> 00:09:32.120 authenticity with AI? I think there's a risk. 00:09:33.580 --> 00:09:37.419 I sort of am using AI slightly differently with 00:09:37.419 --> 00:09:41.220 my learners now. So, one of the challenges, and 00:09:41.220 --> 00:09:43.179 this comes from my sales training background 00:09:43.179 --> 00:09:46.220 forever ago when I used to do a lot of sales 00:09:46.220 --> 00:09:50.809 training, that Humans are very good at talking 00:09:50.809 --> 00:09:54.190 about facts. So back in the day, I worked for 00:09:54.190 --> 00:09:58.070 a national house builder and I taught our team 00:09:58.070 --> 00:10:01.529 to demonstrate show homes. And they would say, 00:10:01.610 --> 00:10:05.450 and here's the kitchen. And they were very good 00:10:05.450 --> 00:10:08.450 at talking about features rather than the benefits. 00:10:08.470 --> 00:10:12.610 Now, where our AI is coming in and that's where 00:10:12.610 --> 00:10:15.429 I found it really useful for my clients. And 00:10:15.429 --> 00:10:17.129 I think that's the thing with any tool, isn't 00:10:17.129 --> 00:10:20.700 it, girls? It's finding how you can make it work 00:10:20.700 --> 00:10:23.100 for you. So I'll say, well, OK, this is what 00:10:23.100 --> 00:10:26.940 you do. What does that mean to your perfect clients, 00:10:27.100 --> 00:10:29.879 the people you want to work with you, your future 00:10:29.879 --> 00:10:33.139 clients? Then let's put that into AI and say, 00:10:33.259 --> 00:10:36.460 change these facts into a meaningful series of 00:10:36.460 --> 00:10:40.840 benefits to my perfect client who is a busy HR 00:10:40.840 --> 00:10:44.340 director. And then the output then becomes something 00:10:44.340 --> 00:10:47.120 that they can use. And that's when you get the 00:10:47.120 --> 00:10:51.460 light bulb moments and they go, oh, yeah. It 00:10:51.460 --> 00:10:53.539 makes a big difference, I think, at that point. 00:10:53.600 --> 00:10:57.039 For my particular learners, yes, definitely. 00:10:57.299 --> 00:11:00.000 I agree with that. And I think that this is where 00:11:00.000 --> 00:11:03.399 we need to make sure that when we're using AI 00:11:03.399 --> 00:11:06.879 and not only adopting it for ourselves in the 00:11:06.879 --> 00:11:10.100 content that we're creating, but... When we're 00:11:10.100 --> 00:11:11.879 supporting people in their skills development 00:11:11.879 --> 00:11:14.159 to keep our humanities, that critical thinking 00:11:14.159 --> 00:11:17.700 type of prompting. So, you know, I would almost 00:11:17.700 --> 00:11:22.139 take that a step further and saying when when 00:11:22.139 --> 00:11:27.120 you put it in saying what what tweaks could I 00:11:27.120 --> 00:11:30.559 make or analyze this to show the gaps of where 00:11:30.559 --> 00:11:32.600 the benefits could be done or what tweaks could 00:11:32.600 --> 00:11:35.320 I make to create benefits and kind of bouncing 00:11:35.320 --> 00:11:37.860 around the ideas. So then when it gives it back 00:11:37.860 --> 00:11:40.139 to you. you've still got some thinking to do 00:11:40.139 --> 00:11:44.200 or some analysis to do. There's a really big 00:11:44.200 --> 00:11:47.220 push on, well, there's a lot of buzzwords going 00:11:47.220 --> 00:11:50.580 around, but I think that talking about using 00:11:50.580 --> 00:11:54.580 AI in the centaur mentality where you are using 00:11:54.580 --> 00:11:58.419 it as part of your thinking, part of your process, 00:11:58.519 --> 00:12:02.320 it's humanity and AI collaborating together to 00:12:02.320 --> 00:12:06.470 come up with an answer rather than... it giving 00:12:06.470 --> 00:12:08.850 you the answer and it not being fact -checked 00:12:08.850 --> 00:12:11.129 because AI hallucinates so much so unless you 00:12:11.129 --> 00:12:14.929 put in a grounding prompt and a really strong 00:12:14.929 --> 00:12:17.710 prompt it can hallucinate whatever it wants as 00:12:17.710 --> 00:12:20.690 a language model so I think what you're doing 00:12:20.690 --> 00:12:22.610 Karen is brilliant like that is that critical 00:12:22.610 --> 00:12:27.090 thinking talking to AI and bringing it into I'm 00:12:27.090 --> 00:12:28.929 trying to create something how can you help me 00:12:28.929 --> 00:12:32.789 create it I love that. I love that. And I think 00:12:32.789 --> 00:12:35.870 that for me is a big thing is as a collaborator. 00:12:36.009 --> 00:12:38.110 I think that's a great way of putting it actually. 00:12:38.330 --> 00:12:40.909 And there's also, so there's Centaur, but there's 00:12:40.909 --> 00:12:43.350 also going to, and we'll see it more as it comes 00:12:43.350 --> 00:12:44.769 along. And I think this is what people are worried 00:12:44.769 --> 00:12:48.110 about is that cyborg element where if you embrace 00:12:48.110 --> 00:12:50.350 it, there's a really great opportunity inside 00:12:50.350 --> 00:12:54.210 it. But the cyborg element is where AI is literally 00:12:54.210 --> 00:12:57.679 working in tandem. where you where you're not 00:12:57.679 --> 00:12:59.799 asking it to do things it's pre -empting and 00:12:59.799 --> 00:13:03.539 doing it for you so as we have um for example 00:13:03.539 --> 00:13:08.539 our like automatic uh spell correction on on 00:13:08.539 --> 00:13:11.200 our phones we're not asking it to do it each 00:13:11.200 --> 00:13:12.840 time sometimes it's just trying to do it for 00:13:12.840 --> 00:13:14.259 us sometimes it gets it right sometimes it gets 00:13:14.259 --> 00:13:17.679 it wrong but there's an opportunity inside AI 00:13:17.679 --> 00:13:20.860 for as an example if you were like a graphic 00:13:20.860 --> 00:13:22.759 designer that you're going around making little 00:13:22.759 --> 00:13:25.960 changes and that your AI is working on something 00:13:25.960 --> 00:13:27.360 in the background so you don't have to make as 00:13:27.360 --> 00:13:29.679 many changes when you come to it so you're working 00:13:29.679 --> 00:13:31.919 in tandem to the same purpose and the same drive 00:13:31.919 --> 00:13:35.460 and it's learning your behaviors to try and support 00:13:35.460 --> 00:13:40.100 them preemptively and there's if you see it as 00:13:40.100 --> 00:13:42.399 there are risks in inside things that people 00:13:42.399 --> 00:13:43.980 might be worried about there's also opportunity 00:13:43.980 --> 00:13:46.580 inside that so I think that really understanding 00:13:47.399 --> 00:13:49.720 the prompt in the model and the more knowledge 00:13:49.720 --> 00:13:53.480 you have of the capability can also reduce that 00:13:53.480 --> 00:13:56.000 threat risk. Because at the end of the day, our 00:13:56.000 --> 00:13:58.480 brains are always scanning for threat. And the 00:13:58.480 --> 00:14:01.679 way to reduce threat and risk in our perception 00:14:01.679 --> 00:14:04.899 is to feel that we have control. So the more 00:14:04.899 --> 00:14:07.100 knowledge we have, the more control. So I think, 00:14:07.100 --> 00:14:10.080 Karen, the way that you're building out the sort 00:14:10.080 --> 00:14:14.100 of understanding and showing people. how to use 00:14:14.100 --> 00:14:16.059 a critical thinking prompt and how to use it 00:14:16.059 --> 00:14:18.379 in the way of using it as a partner is that great 00:14:18.379 --> 00:14:21.419 first step into building into those areas. What 00:14:21.419 --> 00:14:24.480 do you think, Laura? Yeah, I have to say I'm 00:14:24.480 --> 00:14:28.460 hearing a lot about the time and efficiency gains 00:14:28.460 --> 00:14:31.519 of AI. And as you described there, Katie, about 00:14:31.519 --> 00:14:33.899 the working in tandem with the tool and it's 00:14:33.899 --> 00:14:37.399 helping you to arrive at your product a bit quicker. 00:14:38.120 --> 00:14:40.600 Totally agree about the critical thinking skills. 00:14:41.100 --> 00:14:44.059 I'm not concerned AI is coming for my job just 00:14:44.059 --> 00:14:48.860 yet, but I can see how I need my experience and 00:14:48.860 --> 00:14:52.039 my knowledge to select what is a good output 00:14:52.039 --> 00:14:55.000 from the AI and indeed to put the correct outputs 00:14:55.000 --> 00:14:58.549 into it. into the ai um current to speak to your 00:14:58.549 --> 00:15:03.570 point around using ai to to uh predict your client 00:15:03.570 --> 00:15:06.929 behavior or or the clients of your clients i 00:15:06.929 --> 00:15:09.230 think that's such a phenomenal use of ai and 00:15:09.230 --> 00:15:12.049 it's a space where i've been using ai the most 00:15:12.049 --> 00:15:15.169 really is to kind of stay in the the analysis 00:15:15.169 --> 00:15:18.740 space so rather than quickly trying to produce 00:15:18.740 --> 00:15:22.340 content as quickly as possible it's really promoted 00:15:22.340 --> 00:15:24.620 me staying in the analysis space maybe a bit 00:15:24.620 --> 00:15:27.200 longer than I might have done because you know 00:15:27.200 --> 00:15:29.600 when you have a fixed budget for a project it 00:15:29.600 --> 00:15:32.960 can be really tempting to move as quickly as 00:15:32.960 --> 00:15:35.539 possible to spend the money that you have to 00:15:35.539 --> 00:15:39.259 make this thing the best it can be and and not 00:15:39.259 --> 00:15:42.019 all Clients don't always appreciate the value 00:15:42.019 --> 00:15:44.100 of the analysis space. They might think, why 00:15:44.100 --> 00:15:46.279 are you asking me these questions? I already 00:15:46.279 --> 00:15:48.279 know what the problem is. Or why do you need 00:15:48.279 --> 00:15:50.340 to speak to my learners? I've told you what the 00:15:50.340 --> 00:15:53.879 problem is. And what I found with AI is the more 00:15:53.879 --> 00:15:57.120 relevant input that you can give it, the more 00:15:57.120 --> 00:15:59.039 information about your learners, about their 00:15:59.039 --> 00:16:00.940 behaviors, about the business context, about 00:16:00.940 --> 00:16:03.340 the solution that you're trying, the problem 00:16:03.340 --> 00:16:05.519 that you're trying to solve. the better able 00:16:05.519 --> 00:16:08.139 it is to help you to create that content. So 00:16:08.139 --> 00:16:10.940 what it's encouraged me to do is not skip past 00:16:10.940 --> 00:16:13.320 analysis stage, make sure I have all of those 00:16:13.320 --> 00:16:16.500 inputs, make sure I have the learner interviews, 00:16:16.659 --> 00:16:18.779 make sure I have the stakeholder interviews so 00:16:18.779 --> 00:16:21.519 that I can have something to give AI to understand 00:16:21.519 --> 00:16:24.240 the challenge that I'm working on. So yeah, that 00:16:24.240 --> 00:16:26.720 critical, I think critical thinking, as you said, 00:16:26.779 --> 00:16:31.279 Katie is so important and Karen that using your 00:16:31.279 --> 00:16:34.299 AI to personalize your content for us. specific 00:16:34.299 --> 00:16:37.059 audience i think again is a really a really great 00:16:37.059 --> 00:16:40.759 use case how is everybody else using ai on their 00:16:40.759 --> 00:16:43.620 in their day -to -day work gainer can you speak 00:16:43.620 --> 00:16:47.080 a little bit about that i use it um in a couple 00:16:47.080 --> 00:16:49.659 of ways again i can kind of tend to use it within 00:16:49.659 --> 00:16:52.480 the tools that i have things like canva the the 00:16:52.480 --> 00:16:56.059 ai and that is is good but it's really slow i 00:16:56.059 --> 00:16:58.980 find maybe i'm just maybe i'm just really impatient 00:16:58.980 --> 00:17:01.039 but it's like it's almost like a help when they 00:17:01.039 --> 00:17:03.080 like come on give me give me what i want but 00:17:03.500 --> 00:17:06.140 agree and it's time to even say that you'll say 00:17:06.140 --> 00:17:08.640 this is a developing technology and you go yeah 00:17:08.640 --> 00:17:11.599 but hurry up so so yeah and i think that's really 00:17:11.599 --> 00:17:15.640 it's really you can produce good things quickly 00:17:15.640 --> 00:17:18.259 with that another place that i use it again i'm 00:17:18.259 --> 00:17:20.740 also a yoga teacher so i use things like chat 00:17:20.740 --> 00:17:24.180 gpt for i'll put in i want to have a class on 00:17:24.180 --> 00:17:27.380 back bends for example and again it can give 00:17:27.380 --> 00:17:30.140 you up those in that information what it's not 00:17:30.140 --> 00:17:34.480 so good at yet use the word yet is it's not telling 00:17:34.480 --> 00:17:36.839 me contraindications for that so if you know 00:17:36.839 --> 00:17:39.059 where the dangers are and people doing that it's 00:17:39.059 --> 00:17:41.779 not particularly good sometimes at the linking 00:17:41.779 --> 00:17:44.579 movements from one to the other so it again it's 00:17:44.579 --> 00:17:46.279 that sort of you know that sort of symbiotic 00:17:46.279 --> 00:17:49.240 relationship between it can give me the basis 00:17:49.240 --> 00:17:51.319 of what I want and I think oh that's that's quite 00:17:51.319 --> 00:17:53.839 nice but I still need to do some work to make 00:17:53.839 --> 00:17:57.079 that something that is a is going to be useful 00:17:57.079 --> 00:18:01.279 for for me to prepare a class plan. And I think 00:18:01.279 --> 00:18:03.859 it's improving. It's definitely improving than 00:18:03.859 --> 00:18:06.440 before, but I just keep going in and that's kind 00:18:06.440 --> 00:18:09.920 of my test of it to see where we're getting to. 00:18:10.039 --> 00:18:12.079 So yeah, that's kind of what I'm using for it 00:18:12.079 --> 00:18:15.259 now. I'm actually doing some, just a course on 00:18:15.259 --> 00:18:18.539 future learn on AI. Again, it's one of those 00:18:18.539 --> 00:18:21.400 things, if you're a generalist in learning and 00:18:21.400 --> 00:18:23.460 development, there's all these different things 00:18:23.460 --> 00:18:26.200 that you need to do. We can't all specialize 00:18:26.200 --> 00:18:29.579 in. necessarily e -learning or you know ai and 00:18:29.579 --> 00:18:31.400 stuff so again i think there's a lot of upskilling 00:18:31.400 --> 00:18:34.240 needs to be done for people in general like corporate 00:18:34.240 --> 00:18:37.880 uh for corporate learning and development people 00:18:37.880 --> 00:18:40.200 to understand it could be well be out there but 00:18:40.200 --> 00:18:42.279 if the company's not looking at it you're not 00:18:42.279 --> 00:18:45.160 using it then you're kind of going to get behind 00:18:45.160 --> 00:18:49.819 the curve very very quickly yeah i think that's 00:18:49.819 --> 00:18:52.220 the same for i'm sorry i think that's the same 00:18:52.220 --> 00:18:55.869 for um anyone in any job at the moment if you 00:18:55.869 --> 00:19:00.309 don't start adopting and embracing the foundational 00:19:00.309 --> 00:19:05.769 ai um then you're going to get left behind so 00:19:05.769 --> 00:19:08.069 found and what i mean by foundational ai is really 00:19:08.069 --> 00:19:13.109 understanding good prompting so the the things 00:19:13.109 --> 00:19:14.730 that sometimes when people are talking about 00:19:14.730 --> 00:19:18.130 oh ai turns out you know such trash or like it 00:19:18.130 --> 00:19:22.329 just makes stuff up it's not the ai model Most 00:19:22.329 --> 00:19:25.670 of the time it is the prompt isn't detailed enough 00:19:25.670 --> 00:19:28.369 or grounded or asking it not to hallucinate. 00:19:28.490 --> 00:19:31.849 So like to your point about the yoga bits, it's, 00:19:31.849 --> 00:19:36.309 you know, the prompts that I put in are paragraphs 00:19:36.309 --> 00:19:40.890 long. They're not it's because it's not being 00:19:40.890 --> 00:19:44.930 used as it's not the same as Google. It's it's 00:19:44.930 --> 00:19:51.029 a detailed, deep prompt. And the the. more detail 00:19:51.029 --> 00:19:54.009 your prompt is and like you guys have said the 00:19:54.009 --> 00:19:57.630 more detail that you're giving it the the better 00:19:57.630 --> 00:19:59.970 the model will come back with because the language 00:19:59.970 --> 00:20:02.809 model is just trying to find a best fit it's 00:20:02.809 --> 00:20:04.609 not trying to base it on anything else unless 00:20:04.609 --> 00:20:08.130 you put in the control mechanisms and the wording 00:20:08.130 --> 00:20:11.329 to say i don't want a best fit i want you to 00:20:11.329 --> 00:20:13.349 cite your sources i want you to do it like this 00:20:13.349 --> 00:20:15.690 i don't want any hallucinations i want x y and 00:20:15.690 --> 00:20:19.630 z i've actually built um And I do recommend this 00:20:19.630 --> 00:20:23.309 for anyone listening, build yourself a GPT or 00:20:23.309 --> 00:20:26.190 even a project. So I have the paid for chat GPT 00:20:26.190 --> 00:20:28.910 account and I have projects. And one of my projects 00:20:28.910 --> 00:20:30.650 where I've written the instructions and everything 00:20:30.650 --> 00:20:35.390 is to do, to help me write better prompts. So 00:20:35.390 --> 00:20:38.230 I put my prompt into there and then I say, like, 00:20:38.269 --> 00:20:41.869 analyze this of, this is what I'm, this is my 00:20:41.869 --> 00:20:44.230 goal, but analyze this prompt to see if there 00:20:44.230 --> 00:20:46.470 are anything, anything that I've missed that 00:20:46.470 --> 00:20:49.250 could help me. have a better prompt when I put 00:20:49.250 --> 00:20:52.410 it through and it's such a quick and easy tool 00:20:52.410 --> 00:20:55.730 it's just almost like a checker I love that I 00:20:55.730 --> 00:20:58.150 think that's great I think it's great because 00:20:58.150 --> 00:21:01.869 I learned just last month to create my own I've 00:21:01.869 --> 00:21:05.490 created two different types of custom GPTs and 00:21:05.490 --> 00:21:07.490 I too have got a paid one and I think that's 00:21:07.490 --> 00:21:10.450 a question I get asked a lot is is it worth paying 00:21:10.450 --> 00:21:15.819 it and is AI using AI to help us do this stuff 00:21:15.819 --> 00:21:20.119 worth all the fuss? And I think that it's, I 00:21:20.119 --> 00:21:25.480 think it's about, so as a, for example, we came 00:21:25.480 --> 00:21:27.779 out from holiday and I was really unwell and 00:21:27.779 --> 00:21:30.380 wasn't feeling great at all, but I had some work 00:21:30.380 --> 00:21:35.700 to do. And I actually put some of the stuff into 00:21:35.700 --> 00:21:40.289 my custom GPT and said, Give me some output and 00:21:40.289 --> 00:21:44.289 ideas around this. And I input a recent proposal 00:21:44.289 --> 00:21:47.470 that I'd written when I was well, when I was 00:21:47.470 --> 00:21:51.430 on full Karen mode. And then it then gave me 00:21:51.430 --> 00:21:55.009 and it really helps me out. And I think that's 00:21:55.009 --> 00:21:58.049 the other thing is, it's like you can make it 00:21:58.049 --> 00:22:02.390 your friend to help you when you have, you know, 00:22:02.390 --> 00:22:04.930 when I was. When I wasn't feeling great, I was 00:22:04.930 --> 00:22:08.029 asked to design a new workshop about using Facebook 00:22:08.029 --> 00:22:12.369 personal profiles for your business. And it's 00:22:12.369 --> 00:22:15.250 something I've been physically doing, but I hadn't 00:22:15.250 --> 00:22:17.730 designed and delivered a workshop before. So 00:22:17.730 --> 00:22:22.789 I put it into my custom GPT and said, this is 00:22:22.789 --> 00:22:26.410 what I want to do. Write me an outline so I can 00:22:26.410 --> 00:22:29.710 market this out and create a sales page. And 00:22:29.710 --> 00:22:32.380 it did it for me. And I think that's the thing 00:22:32.380 --> 00:22:34.759 is I think it's finding, you know, I don't think, 00:22:34.779 --> 00:22:36.920 you know, you don't have to use it. But I think 00:22:36.920 --> 00:22:41.180 I know that last month it paid its monthly fee 00:22:41.180 --> 00:22:46.759 more than because of how much it helped me. Yeah. 00:22:46.779 --> 00:22:49.180 I think there's that element is deciding what. 00:22:50.319 --> 00:22:52.900 I suppose it's thinking about what do you need 00:22:52.900 --> 00:22:56.380 most and where can it help you save time, energy, 00:22:56.500 --> 00:22:59.259 headspace? Yeah, I definitely think that that 00:22:59.259 --> 00:23:03.380 is true. And I would just, oh, that when we teach 00:23:03.380 --> 00:23:06.859 it out of there are things you can do where exactly 00:23:06.859 --> 00:23:08.339 as you said, where it'll come back and be like, 00:23:08.380 --> 00:23:11.119 here's the outline, here's the da -da -da. And 00:23:11.119 --> 00:23:13.579 then that's when it saved you time by pulling 00:23:13.579 --> 00:23:15.339 up that first idea. So you're not looking at 00:23:15.339 --> 00:23:19.660 a blank sheet of paper. And we need to... remember 00:23:19.660 --> 00:23:22.019 that there's that next step of, okay, now you 00:23:22.019 --> 00:23:24.819 analyze it and take it and run with it and do 00:23:24.819 --> 00:23:26.940 whatever you need to do. And I think that where 00:23:26.940 --> 00:23:31.680 there's maybe a bit of lag is people just taking 00:23:31.680 --> 00:23:33.779 whatever it produces and going, oh, perfect, 00:23:33.940 --> 00:23:38.240 done, copy, paste. And that's where we're going 00:23:38.240 --> 00:23:42.400 to, if we don't readily teach the steps that 00:23:42.400 --> 00:23:45.019 need to come with analyzing what's being put 00:23:45.019 --> 00:23:48.299 out and not taking it as face value. And you 00:23:48.299 --> 00:23:52.740 can even fire it back into the chat, right? You 00:23:52.740 --> 00:23:54.640 can say, okay, that's great. I can see you've 00:23:54.640 --> 00:23:56.339 done all of this. I've tweaked it down below. 00:23:56.359 --> 00:23:58.740 Now analyze it to tell me how it's going to come 00:23:58.740 --> 00:24:02.119 across. For example, in comms. So I use it for 00:24:02.119 --> 00:24:05.380 comms writing. Like I'm dyslexic. I find it super 00:24:05.380 --> 00:24:07.759 helpful to not look at a blank sheet, to look 00:24:07.759 --> 00:24:11.119 at a kind of starting place. But I can also ask 00:24:11.119 --> 00:24:13.819 it. Like we've said about analyzing it for your 00:24:13.819 --> 00:24:16.480 audience, Karen, you can then, once you've changed 00:24:16.480 --> 00:24:18.140 it, you can ask it to look back at it and go, 00:24:18.200 --> 00:24:20.799 how will this land? Will this engage? What could 00:24:20.799 --> 00:24:23.140 I do to better it? And I just think that there's 00:24:23.140 --> 00:24:28.119 making sure that we're really clear that AI isn't 00:24:28.119 --> 00:24:30.720 going to fact check. It isn't going to generally 00:24:30.720 --> 00:24:33.000 check whatever it produces. Even if you don't 00:24:33.000 --> 00:24:34.880 have facts in there, it's not about that. It's 00:24:34.880 --> 00:24:37.740 just it's producing whatever it thinks is roughly 00:24:37.740 --> 00:24:40.900 right. And it's to help you not start on a blank 00:24:40.900 --> 00:24:46.250 page. 100%. I'm just listening to you there, 00:24:46.289 --> 00:24:49.230 Katie, and I'm remembering my nephew who's almost 00:24:49.230 --> 00:24:53.509 30 now. I got him using computers almost as soon 00:24:53.509 --> 00:24:56.009 as he asked what it was. And he got a new computer 00:24:56.009 --> 00:24:58.349 game and he turned around to me and says, Gaynor, 00:24:58.490 --> 00:25:00.630 what does whatever is look like? And I'm like, 00:25:00.690 --> 00:25:02.349 what are you talking about? He says, well, I 00:25:02.349 --> 00:25:05.029 press the start button. I go to Disney Interactive. 00:25:05.029 --> 00:25:07.390 I go to start programs, Disney Interactive. What 00:25:07.390 --> 00:25:10.009 does whatever is look like? And at that point, 00:25:10.029 --> 00:25:11.450 I realized he was three and he couldn't read. 00:25:12.090 --> 00:25:16.630 But he's just in there. He's in there and he's 00:25:16.630 --> 00:25:19.490 pressing those buttons. And there's a little 00:25:19.490 --> 00:25:21.789 part of me thinks, how many people are doing 00:25:21.789 --> 00:25:23.450 that with chat GPT? It's like they've just got 00:25:23.450 --> 00:25:25.589 this tiny little bit of knowledge. And a little 00:25:25.589 --> 00:25:28.470 bit of knowledge can be a dangerous thing. And 00:25:28.470 --> 00:25:30.049 a bit like my nephew, it's just that he's going 00:25:30.049 --> 00:25:31.650 into it and just, like you say, they're just 00:25:31.650 --> 00:25:33.910 pressing the first couple of buttons and almost 00:25:33.910 --> 00:25:37.980 using it like a search engine. Again, for those 00:25:37.980 --> 00:25:40.059 of us who are old enough to remember pre -Google, 00:25:40.079 --> 00:25:42.880 all the things that you had to put into a search 00:25:42.880 --> 00:25:45.299 engine, you had to put your quotations and all 00:25:45.299 --> 00:25:47.839 your symbols and stuff like that back in the 00:25:47.839 --> 00:25:50.099 early days of search engines. Nowadays, we just 00:25:50.099 --> 00:25:52.299 Google it. I think someone here said earlier 00:25:52.299 --> 00:25:54.920 that it was, I think it came from Buffy the Vampire 00:25:54.920 --> 00:25:59.059 Slayer. Yeah, Buffy. There's an entire generation 00:25:59.059 --> 00:26:01.339 of, I'm saying younger people, like I'm really 00:26:01.339 --> 00:26:04.839 old, but they don't remember. You just put it 00:26:04.839 --> 00:26:06.700 into Google, you just do that. And I wonder whether 00:26:06.700 --> 00:26:10.759 people are using chat GPT like a search engine 00:26:10.759 --> 00:26:14.019 and kind of missing a trick. Yeah, certainly, 00:26:14.200 --> 00:26:17.099 especially if they haven't had access to sort 00:26:17.099 --> 00:26:19.720 of an education about it. I was very fortunate 00:26:19.720 --> 00:26:23.440 at the end of last year to go on Dr. Philippa 00:26:23.440 --> 00:26:26.319 Hardman's AI Learning Design Bootcamp. And the 00:26:26.319 --> 00:26:28.460 amount I learned in four weeks, I was kind of 00:26:28.460 --> 00:26:30.519 reeling. I needed to sort of sit down in a dark 00:26:30.519 --> 00:26:35.200 room afterwards because there is so much about 00:26:35.200 --> 00:26:38.339 the input that you put in. And as you said, Katie, 00:26:38.420 --> 00:26:40.539 about constructing those, you know, I would say 00:26:40.539 --> 00:26:42.920 several paragraph prompts. You know, she talks 00:26:42.920 --> 00:26:45.720 about the CD prompt, which is context, instructions, 00:26:46.240 --> 00:26:49.880 directions and inputs. And so you're really giving. 00:26:50.430 --> 00:26:53.349 the AI a lot of detail before it answers. And 00:26:53.349 --> 00:26:55.490 then there are additional conversational prompts 00:26:55.490 --> 00:26:58.130 that you can use to guide it into the right direction 00:26:58.130 --> 00:27:01.430 and then to validate the results that it's giving 00:27:01.430 --> 00:27:03.789 you, making sure it's not hallucinating, as we've 00:27:03.789 --> 00:27:07.049 mentioned. But I was very fortunate to be able 00:27:07.049 --> 00:27:09.170 to do that. I wonder if that's available for 00:27:09.170 --> 00:27:11.769 everybody else. And with all the hype about AI 00:27:11.769 --> 00:27:14.450 that you see, hey, we can make content quicker. 00:27:14.589 --> 00:27:18.069 Hey, you can make content even if you don't have 00:27:18.069 --> 00:27:21.599 a video. have a background in video design or 00:27:21.599 --> 00:27:24.980 graphic design but I don't see the same kind 00:27:24.980 --> 00:27:27.680 of noise about hey this is how to prop to engineer 00:27:27.680 --> 00:27:30.500 a good prompt this is how to validate your results 00:27:30.500 --> 00:27:32.740 it's just not as shiny it's not as glamorous 00:27:32.740 --> 00:27:35.740 and I think that isn't that the risk is that 00:27:35.740 --> 00:27:40.099 we're getting sometimes that human creativity 00:27:40.099 --> 00:27:44.420 gets lost sight of a little bit because we get 00:27:44.759 --> 00:27:47.579 I'm terrible. I'm proper geeky with new things. 00:27:47.680 --> 00:27:51.160 I love new things. But then sometimes you think, 00:27:51.220 --> 00:27:53.960 well, what's at the fundamental? What's the foundations 00:27:53.960 --> 00:27:55.799 of this? And you're absolutely right, Laura. 00:27:55.839 --> 00:27:59.900 I think there is an education piece for people 00:27:59.900 --> 00:28:02.880 to understand because people will dip in and 00:28:02.880 --> 00:28:06.160 out of it without really understanding it. Yeah. 00:28:06.220 --> 00:28:08.779 I think also you need to have a purpose for it. 00:28:08.920 --> 00:28:11.240 It's like the other day I went and I created 00:28:11.240 --> 00:28:14.329 a... an avatar of myself one of these little 00:28:14.329 --> 00:28:17.009 cartoon things and I was like that's really great 00:28:17.009 --> 00:28:18.329 and then I thought what am I going to do with 00:28:18.329 --> 00:28:19.910 it you know it's like it's like just because 00:28:19.910 --> 00:28:22.069 you can do something doesn't necessarily mean 00:28:22.069 --> 00:28:25.150 you should and I think that personally these 00:28:25.150 --> 00:28:28.230 sort of AI avatars that you can create the ones 00:28:28.230 --> 00:28:31.470 that are supposed to be like real life. I took 00:28:31.470 --> 00:28:36.029 a photograph of myself and AI -ified it with 00:28:36.029 --> 00:28:38.769 a press of a button, which is all well and good, 00:28:38.869 --> 00:28:41.369 but it didn't then look like a real person. It's 00:28:41.369 --> 00:28:46.009 almost good enough, but not quite. I did that 00:28:46.009 --> 00:28:48.950 too. It looked nothing like me. I actually showed 00:28:48.950 --> 00:28:51.670 it to my husband and he started laughing. He 00:28:51.670 --> 00:28:56.609 went, who the fudge is that? From that perspective, 00:28:56.849 --> 00:28:59.980 it's like it's almost there. but not quite. And 00:28:59.980 --> 00:29:02.660 I find I'd rather have like an animation where 00:29:02.660 --> 00:29:06.019 it's quite clearly not the real me rather than 00:29:06.019 --> 00:29:09.680 something that's 98 and a half percent there. 00:29:09.759 --> 00:29:12.240 Because if I was doing layering or something 00:29:12.240 --> 00:29:14.359 and I saw someone and I see something that is 00:29:14.359 --> 00:29:16.140 clear, it's almost like the eyes aren't quite 00:29:16.140 --> 00:29:20.900 right. It really, personally, it puts me off. 00:29:21.099 --> 00:29:23.420 So I'm waiting for that point where that gets 00:29:23.420 --> 00:29:25.920 to that, you know, actually, where I can create 00:29:25.920 --> 00:29:29.039 a video that completely looks like me. Maybe 00:29:29.039 --> 00:29:33.180 that's a dangerous thing. I think that it also 00:29:33.180 --> 00:29:34.799 depends on the tools that you're using. Some 00:29:34.799 --> 00:29:37.960 tools, they're putting their effort into producing 00:29:37.960 --> 00:29:41.740 that and some aren't. But our brains are wired 00:29:41.740 --> 00:29:44.779 to be able to, there's a certain level of detection 00:29:44.779 --> 00:29:46.680 that it will look at something and think that's 00:29:46.680 --> 00:29:49.400 slightly off, which is getting harder and harder 00:29:49.400 --> 00:29:51.359 for our brains to process and do, obviously. 00:29:51.519 --> 00:29:54.710 But yeah, I think it's still there. as as a little 00:29:54.710 --> 00:29:57.630 bit but um it's definitely considering which 00:29:57.630 --> 00:29:59.730 tool you're using as well because they're putting 00:29:59.730 --> 00:30:03.349 more money into the production in certain tools 00:30:03.349 --> 00:30:05.470 than others so others are getting lagged behind 00:30:05.470 --> 00:30:08.809 for example this is this is a fun thing that 00:30:08.809 --> 00:30:12.009 i have done with my my nieces where we put in 00:30:12.009 --> 00:30:17.019 a photo into chat gpt and and i've got i've got 00:30:17.019 --> 00:30:19.640 all my memory and data collection all off so 00:30:19.640 --> 00:30:22.460 we put in our photo into chat gpt and turn it 00:30:22.460 --> 00:30:25.140 into a coloring page and it's really good at 00:30:25.140 --> 00:30:28.200 doing that because it's quite a simple outlining 00:30:28.200 --> 00:30:30.980 but if i wanted it because i'm just making it 00:30:30.980 --> 00:30:32.960 into a black and white kind of cartoony outlining 00:30:32.960 --> 00:30:37.740 but if i wanted it to do a replication yeah that 00:30:37.740 --> 00:30:41.099 isn't that isn't where chat gpt shines at all 00:30:41.420 --> 00:30:45.980 and so it's for us as lnd and when we're creating 00:30:45.980 --> 00:30:48.559 content really understanding which tool we're 00:30:48.559 --> 00:30:51.240 using for different the different things that 00:30:51.240 --> 00:30:54.920 we need it for is is also a really important 00:30:54.920 --> 00:30:57.480 aspect and i don't know if you guys have seen 00:30:57.480 --> 00:31:00.640 it but um there's that site that's uh called 00:31:00.640 --> 00:31:03.160 there's an ai tool for that that's quite good 00:31:03.160 --> 00:31:04.700 i've used that a few times when i'm trying to 00:31:04.700 --> 00:31:09.759 find different tools so you know Again, it's 00:31:09.759 --> 00:31:11.799 that critical thinking of, is this the best tool 00:31:11.799 --> 00:31:13.759 or is this the quickest tool that I have access 00:31:13.759 --> 00:31:18.680 to? And will that actually end up me taking longer 00:31:18.680 --> 00:31:21.279 to do it than if I took a little moment and found 00:31:21.279 --> 00:31:24.599 a tool that maybe was a bit more aligned? But 00:31:24.599 --> 00:31:26.420 I think we've always done that, really, or at 00:31:26.420 --> 00:31:29.019 least I have. You never just went into one tool 00:31:29.019 --> 00:31:31.759 and used it if you were creating a course or 00:31:31.759 --> 00:31:33.720 creating an e -learning course. You're always 00:31:33.720 --> 00:31:36.640 in and out of different programs. So maybe that's 00:31:36.640 --> 00:31:39.980 just... That's just the next evolution of the 00:31:39.980 --> 00:31:43.799 tools that we have. Yeah, definitely. I have 00:31:43.799 --> 00:31:46.339 to say one of the spaces that's really helped 00:31:46.339 --> 00:31:50.619 me in content design is writing robust assessments, 00:31:50.839 --> 00:31:56.119 being able to give the AI all of the course content 00:31:56.119 --> 00:32:00.380 and the learning objectives and the learner personas 00:32:00.380 --> 00:32:02.000 or the learner behaviours that I spoke about 00:32:02.000 --> 00:32:04.440 before and coming up with robust assessments, 00:32:04.759 --> 00:32:07.329 particularly around good... distractors good 00:32:07.329 --> 00:32:09.930 quality distractors there will be something that 00:32:09.930 --> 00:32:11.589 I would always struggle with before because they 00:32:11.589 --> 00:32:13.710 need to be meaningful they need to be there's 00:32:13.710 --> 00:32:16.170 a reason why you might have selected this um 00:32:16.170 --> 00:32:18.970 as a wrong answer so they need to be good good 00:32:18.970 --> 00:32:21.990 wrong answers and then AI tools have been so 00:32:21.990 --> 00:32:24.450 helpful for for coming up with that I wanted 00:32:24.450 --> 00:32:27.210 to just circle back to your comment there Gaynor 00:32:27.210 --> 00:32:29.190 about just because we can doesn't mean we should 00:32:29.190 --> 00:32:31.970 and I do think that's so much about some of the 00:32:31.970 --> 00:32:34.869 AI tools that I've been asked to use there is 00:32:34.869 --> 00:32:37.809 one called Synthesia which creates talking head 00:32:37.809 --> 00:32:41.970 videos, which I've always been dubious about 00:32:41.970 --> 00:32:46.119 anyway. you know if uh having somebody read out 00:32:46.119 --> 00:32:49.319 content to you is probably one of the lowest 00:32:49.319 --> 00:32:52.960 forms of video use cases i can think of is people 00:32:52.960 --> 00:32:55.500 can people might prefer to read people might 00:32:55.500 --> 00:32:58.559 prefer to listen so having you're very tied to 00:32:58.559 --> 00:33:01.380 the speed of the of the delivery of this person 00:33:01.380 --> 00:33:05.299 and then to make that person an uncanny not quite 00:33:05.299 --> 00:33:09.220 human not quite robot person i i i have a i have 00:33:09.220 --> 00:33:13.079 some problems with and um i uh the way that I 00:33:13.079 --> 00:33:16.880 would encourage people to use media in learning. 00:33:17.039 --> 00:33:18.420 Well, one of the ways I would encourage people 00:33:18.420 --> 00:33:21.200 to use media in learning is to create that personal... 00:33:21.799 --> 00:33:25.279 aspect just to create that humanity um you know 00:33:25.279 --> 00:33:27.420 distance learning online learning can be quite 00:33:27.420 --> 00:33:29.339 a lonely task you sort of just start on your 00:33:29.339 --> 00:33:31.859 own so when you see a real person in there and 00:33:31.859 --> 00:33:34.460 they're an expert in their field or they're very 00:33:34.460 --> 00:33:36.819 charismatic in a certain way you're connected 00:33:36.819 --> 00:33:39.519 to the content in some way and then to take that 00:33:39.519 --> 00:33:42.400 thing which has its use this of creating human 00:33:42.400 --> 00:33:45.000 connection and replace it with an AI version 00:33:45.000 --> 00:33:47.279 because we can because it's cheaper because it's 00:33:47.279 --> 00:33:50.559 faster you're kind of losing the point of having 00:33:50.559 --> 00:33:52.680 a video you there in the first place so yeah 00:33:52.680 --> 00:33:55.460 just because you can doesn't mean you should 00:33:55.460 --> 00:33:58.319 do a wider realms of content karen maybe it's 00:33:58.319 --> 00:34:01.619 it's linking to your profession but the fact 00:34:01.619 --> 00:34:05.440 that we're getting kind of ai actors and um you 00:34:05.440 --> 00:34:09.079 know ai reporters and and things like that what 00:34:09.079 --> 00:34:10.820 are your what are your thoughts on that guys 00:34:10.820 --> 00:34:14.250 i've always felt that we should be using ai to 00:34:14.250 --> 00:34:18.289 be doing the the boring tasks the difficult tasks 00:34:18.289 --> 00:34:20.650 the monotonous tasks the tasks where there's 00:34:20.650 --> 00:34:22.909 too much information for us to synthesize in 00:34:22.909 --> 00:34:26.210 one in ourselves using it as a co -pilot in that 00:34:26.210 --> 00:34:28.989 way i really don't like the idea that we're giving 00:34:28.989 --> 00:34:32.269 it the enjoyable creative tasks that we would 00:34:32.269 --> 00:34:35.269 do ourselves that we would just don't worry you 00:34:35.269 --> 00:34:38.630 do the acting i'll carry on the other night i 00:34:38.630 --> 00:34:40.730 was watching the latest episode of the morning 00:34:40.730 --> 00:34:44.380 show And behind the scenes, they've been developing 00:34:44.380 --> 00:34:50.059 these AI hosts of the shows. And they launch 00:34:50.059 --> 00:34:55.079 it at this live event. And the AI that is the 00:34:55.079 --> 00:34:58.460 boss, that is of the boss, has been learning 00:34:58.460 --> 00:35:02.079 from the boss. And the whole thing implodes. 00:35:02.099 --> 00:35:06.780 And then repeats back things that the real person 00:35:06.780 --> 00:35:11.610 has said. And it totally went, oh, my God, no, 00:35:11.730 --> 00:35:14.710 you don't want AI doing this sort of stuff. That 00:35:14.710 --> 00:35:17.610 to me was, and I think, Laurie, you made the 00:35:17.610 --> 00:35:22.610 perfect point, is the tasks that take up our 00:35:22.610 --> 00:35:26.570 headspace that isn't, you know, like back in 00:35:26.570 --> 00:35:28.849 the day, my first business coach said to me. 00:35:29.519 --> 00:35:32.420 Best advice I can give you is get a virtual administrator, 00:35:32.780 --> 00:35:35.079 somebody who can do the jobs that you hate to 00:35:35.079 --> 00:35:38.059 do, that isn't a good use of your time, that 00:35:38.059 --> 00:35:41.619 will cost you a lot less. I see AI a bit like 00:35:41.619 --> 00:35:45.400 that, is that I'll use AI to do the stuff that 00:35:45.400 --> 00:35:48.860 I can sit and probably procrastinate about, but 00:35:48.860 --> 00:35:50.840 then I can look at it and say, do you know what 00:35:50.840 --> 00:35:55.119 it is? Okay, this is my idea, bring it to life. 00:35:55.500 --> 00:35:58.550 And I think that's the thing is, is that... To 00:35:58.550 --> 00:36:02.190 me, that's the biggest help it can give me is 00:36:02.190 --> 00:36:04.889 I can say, well, okay, you're going to save me 00:36:04.889 --> 00:36:07.730 some time now and some headspace. And that to 00:36:07.730 --> 00:36:11.210 me is invaluable. And that's how I get around 00:36:11.210 --> 00:36:15.030 it. I don't want to see an AI version of Karen. 00:36:15.130 --> 00:36:18.010 That would be bloody scary. It would be bloody 00:36:18.010 --> 00:36:20.769 scary. In fact, oddly enough, I just recorded 00:36:20.769 --> 00:36:23.369 a new lead magnet and I sent it to some clients 00:36:23.369 --> 00:36:27.650 for some feedback. I didn't have myself on the 00:36:27.650 --> 00:36:30.869 screen. It was just me narrating it. And the 00:36:30.869 --> 00:36:33.969 feedback I got was we wanted to see you. And 00:36:33.969 --> 00:36:37.710 I think that's the disconnect, isn't it? When 00:36:37.710 --> 00:36:40.750 it moves more advanced. Yeah. And I think we 00:36:40.750 --> 00:36:43.050 can also get stuck into, I mean, we have it in 00:36:43.050 --> 00:36:44.650 social media, right? We go into like a social 00:36:44.650 --> 00:36:48.190 media vacuum where it's kind of because the algorithm 00:36:48.190 --> 00:36:50.130 keeps you inside a vacuum. And I think that AI 00:36:50.130 --> 00:36:52.530 is similar. If we don't put in our humanity and 00:36:52.530 --> 00:36:56.139 our almost disruptors. in because because of 00:36:56.139 --> 00:37:00.559 how we think and we're curious human beings that's 00:37:00.559 --> 00:37:02.079 the whole thing about the critical thinking is 00:37:02.079 --> 00:37:06.320 that we put in almost random acts that that it 00:37:06.320 --> 00:37:08.980 can't necessarily predict and that shakes up 00:37:08.980 --> 00:37:12.239 the vacuum and the ai kind of loop and that's 00:37:12.239 --> 00:37:15.559 where we can disrupt the the loop it creates 00:37:15.559 --> 00:37:18.199 where it might be hallucinating or kind of if 00:37:18.199 --> 00:37:22.780 if you stay inside like a one chat and and you're 00:37:22.780 --> 00:37:25.539 trying to do emails, for example, because it's 00:37:25.539 --> 00:37:28.980 learning each time, your emails become more similar 00:37:28.980 --> 00:37:31.139 and more similar. Or if you're doing it for like 00:37:31.139 --> 00:37:33.800 posting, your posts would become very repetitive 00:37:33.800 --> 00:37:36.980 because it's learning that, oh, that's what you 00:37:36.980 --> 00:37:39.519 like. So I'll keep making things very similar. 00:37:40.119 --> 00:37:43.719 And so that humanity and that disruption element, 00:37:43.860 --> 00:37:46.179 and what you were saying, Karen, about being 00:37:46.179 --> 00:37:50.920 a person behind it. people notice people notice 00:37:50.920 --> 00:37:53.019 when the pattern is becoming too repetitive it 00:37:53.019 --> 00:37:55.300 starts switching people off having you on screen 00:37:55.300 --> 00:37:57.440 and those little sort of whimsical things that 00:37:57.440 --> 00:37:59.800 you might do that's what disrupts your brain 00:37:59.800 --> 00:38:02.480 from kind of turning off and so i think that 00:38:02.480 --> 00:38:06.880 those are really really oddly and this actually 00:38:06.880 --> 00:38:09.400 aligns exactly with what we're talking about 00:38:09.400 --> 00:38:12.440 i was reading a report from linkedin about the 00:38:12.440 --> 00:38:16.239 algorithm changes and one of the ways in which 00:38:16.239 --> 00:38:19.679 linkedin's now telling you love it or loathe 00:38:19.679 --> 00:38:22.619 it best way to disrupt the algorithm is to go 00:38:22.619 --> 00:38:25.860 live be a real person back to what we're talking 00:38:25.860 --> 00:38:29.659 about and I mean I go live every Monday anyway 00:38:29.659 --> 00:38:33.199 and have done for nearly eight years but I'm 00:38:33.199 --> 00:38:35.780 now saying to my clients don't just hear my word 00:38:35.780 --> 00:38:40.300 for it this stuff works yeah I think on the flip 00:38:40.300 --> 00:38:43.460 side of that as well and I don't know I certainly 00:38:43.460 --> 00:38:48.039 am very aware that if I am on social media facebook 00:38:48.039 --> 00:38:51.800 linkedin and i've looked at something the algorithm 00:38:51.800 --> 00:38:55.940 is then giving me more of the same and i think 00:38:55.940 --> 00:38:58.380 there's a real danger that you get into that 00:38:58.380 --> 00:39:02.119 little you're just given what this this little 00:39:02.119 --> 00:39:05.760 world view um and i think it's really important 00:39:05.760 --> 00:39:08.639 that you look at things and again whether it's 00:39:08.639 --> 00:39:11.659 political views or whatever that you go and you 00:39:11.659 --> 00:39:14.360 look at the what other people are seeing and 00:39:14.360 --> 00:39:16.500 other views because otherwise you end up just 00:39:16.500 --> 00:39:20.400 in that little sort of bubble of this is my world 00:39:20.400 --> 00:39:25.860 and I believe this and the bots everything is 00:39:25.860 --> 00:39:28.360 coming to your same world view whereas actually 00:39:28.360 --> 00:39:30.260 you need that and again whether that's to do 00:39:30.260 --> 00:39:33.820 with content or or anything it's like you get 00:39:33.820 --> 00:39:36.219 i may be the only person it's like when netflix 00:39:36.219 --> 00:39:38.659 suggests things based on what i've watched i 00:39:38.659 --> 00:39:40.760 deliberately go and look at something completely 00:39:40.760 --> 00:39:43.659 random so that it it screws up screws up the 00:39:43.659 --> 00:39:45.860 bot so that but it's like i don't want to just 00:39:45.860 --> 00:39:49.260 watch all of the same content all the time and 00:39:49.260 --> 00:39:52.519 i think that there's a real um you know i'm really 00:39:52.519 --> 00:39:54.619 conscious that i don't want to be stuck in this 00:39:54.619 --> 00:39:57.739 this bubble of where it's like algorithms and 00:39:57.739 --> 00:40:01.059 ai is giving me what it thinks my worldview should 00:40:01.059 --> 00:40:04.659 be it's giving you what it what his notice is 00:40:04.659 --> 00:40:07.139 a pattern of what you engage with so it thinks 00:40:07.139 --> 00:40:09.920 it thinks on that basis of you engage with it 00:40:09.920 --> 00:40:11.920 you like it therefore that's the right thing 00:40:11.920 --> 00:40:14.500 correct move i'll do more of it you know the 00:40:14.500 --> 00:40:16.739 same as when we're a child and we learn to to 00:40:16.739 --> 00:40:18.719 please someone we try and do it again and again 00:40:18.719 --> 00:40:23.119 and um you know if you've ever been around children 00:40:23.119 --> 00:40:25.599 have children worked with children you say one 00:40:25.599 --> 00:40:27.579 thing that you like you say you like their picture 00:40:27.579 --> 00:40:30.940 of a son and by the end of the day I've got 20 00:40:30.940 --> 00:40:33.099 pictures of sons on my desk and I'm like oh I 00:40:33.099 --> 00:40:37.340 just liked that one but sure um so you know if 00:40:37.340 --> 00:40:39.699 you think about AI in the same way you say you'd 00:40:39.699 --> 00:40:41.320 like it once and it's like great I will give 00:40:41.320 --> 00:40:44.559 you 20 of the same sort of thing rather than 00:40:45.179 --> 00:40:47.500 I liked that, but do something different. And 00:40:47.500 --> 00:40:50.280 even that prompt in itself, I like that, but 00:40:50.280 --> 00:40:52.659 now show it to me differently so I can see a 00:40:52.659 --> 00:40:55.960 different perspective. It'll do it, but it needs 00:40:55.960 --> 00:40:59.679 the prompt. As we sort of come to the end of 00:40:59.679 --> 00:41:05.059 our time together, I wonder, we've sort of discussed 00:41:05.059 --> 00:41:08.420 some of the challenges there of AI. And I wonder 00:41:08.420 --> 00:41:11.159 if before we leave, we want to sort of go around 00:41:11.159 --> 00:41:13.420 and say how we think that we can reconcile those 00:41:13.420 --> 00:41:15.780 challenges. I think for me, really strongly coming 00:41:15.780 --> 00:41:17.920 out of the discussion is the humanity. I would 00:41:17.920 --> 00:41:20.280 wholeheartedly agree, Laura. I mean, I see it. 00:41:20.340 --> 00:41:22.739 And AI isn't going to do this stuff on its own. 00:41:22.780 --> 00:41:25.920 We really need to be putting the strong effort 00:41:25.920 --> 00:41:28.659 into our prompting, into reviewing the outputs, 00:41:28.780 --> 00:41:32.340 into working with AI as a co -pilot rather than 00:41:32.340 --> 00:41:33.599 letting AI do these things for us. It's going 00:41:33.599 --> 00:41:36.780 to save you a lot of time. And I think it's, 00:41:36.880 --> 00:41:40.320 It's not compulsory. You don't have to use it. 00:41:40.380 --> 00:41:42.960 But when you have a moment where you think, what 00:41:42.960 --> 00:41:46.139 could I change or do differently for my learners 00:41:46.139 --> 00:41:48.699 that's going to make a more meaningful experience? 00:41:49.380 --> 00:41:53.719 Then actually, I often talk about the way we 00:41:53.719 --> 00:41:57.039 behave and we do things. It's like turning a 00:41:57.039 --> 00:41:59.679 cruise ship around quickly. That takes time. 00:42:00.500 --> 00:42:04.800 Whereas what AI can help us do is to... shorten 00:42:04.800 --> 00:42:08.699 that time that we make those changes. Yeah. For 00:42:08.699 --> 00:42:13.099 me, I think AI is a tool. I think it's developed 00:42:13.099 --> 00:42:16.300 very rapidly. I believe the first time anyone 00:42:16.300 --> 00:42:19.980 mentioned AI or machine learning was in the 1940s. 00:42:20.000 --> 00:42:24.739 So that concept has been around for a long time. 00:42:24.780 --> 00:42:28.940 So it's here and it's become a tool that we can 00:42:28.940 --> 00:42:32.139 use and we should use. But again, using it with, 00:42:32.159 --> 00:42:37.000 using our heads, using our intellect with it 00:42:37.000 --> 00:42:39.420 rather than just giving it free reign to just, 00:42:39.500 --> 00:42:43.400 you know, take over the world. I completely agree 00:42:43.400 --> 00:42:46.820 with everything that you guys have said of how 00:42:46.820 --> 00:42:50.219 to reconcile. And so I would just sort of leave 00:42:50.219 --> 00:42:55.179 us with one thing that I readily pick people 00:42:55.179 --> 00:42:59.480 up on in sort of training. And hopefully not 00:42:59.480 --> 00:43:03.719 in an annoying way, but just that sort of nudge 00:43:03.719 --> 00:43:08.000 element to psychologically change how we see 00:43:08.000 --> 00:43:11.460 AI. So instead of saying, I'll get chat GPT to 00:43:11.460 --> 00:43:15.960 do that for me, change that wording to, I'll 00:43:15.960 --> 00:43:21.199 get AI to do that with me. Yeah. Never use for 00:43:21.199 --> 00:43:24.559 me, you know, never use, I'll get something to 00:43:24.559 --> 00:43:27.820 do it for me, always with me. I love that. Brilliant. 00:43:28.329 --> 00:43:31.349 Yeah. Well, it was lovely chatting with you guys. 00:43:31.510 --> 00:43:34.190 Yeah, always interesting, always fascinating 00:43:34.190 --> 00:43:38.530 and thought -provoking. It's always good to chat 00:43:38.530 --> 00:43:40.409 and we never run out of things to see, do we? 00:43:43.650 --> 00:43:46.369 It was really interesting how our guests referenced 00:43:46.369 --> 00:43:49.670 AI more in this episode. We can't ignore its 00:43:49.670 --> 00:43:52.349 changing L &D content and it was great to hear 00:43:52.349 --> 00:43:55.389 from our team about how it's changing work. A 00:43:55.389 --> 00:43:57.789 huge thank you to Gaina, Karen, Katie and Laura 00:43:57.789 --> 00:44:00.070 for their insight into this topic and for their 00:44:00.070 --> 00:44:02.949 time in making this brilliant recording. I know 00:44:02.949 --> 00:44:04.650 you want to connect with them afterwards and 00:44:04.650 --> 00:44:06.690 their details are in the show notes along with 00:44:06.690 --> 00:44:08.449 information about how to connect with us here 00:44:08.449 --> 00:44:11.610 at Women Talking About Learning. Please do remember 00:44:11.610 --> 00:44:13.630 to like and subscribe to the podcast on your 00:44:13.630 --> 00:44:16.329 podcast player. It does make a real difference 00:44:16.329 --> 00:44:20.650 in helping people find and recommend us. We'll 00:44:20.650 --> 00:44:22.809 be back in a couple of weeks time and next time 00:44:22.809 --> 00:44:28.019 it's... the Women's Network one. As always, thanks 00:44:28.019 --> 00:44:29.920 for listening and we'll see you again soon.