AI in No Man's Land: Who Owns it In Your Business

I just asked the table how many of you are playing with Adi and none of them were aware of it. You come forward to a year and a bit ago saying sort of set up everybody knew that. There's not someone who's responsible for it like CTLC is it's not my job, CTLC is not my job, EIS is still in this no man's land. It's about having champions and showing improving that it works. If people drop off what the process is then you end up with that debacle. But isn't the goal in us in place to prevent that from happening? You would hope it's in place. Remember the issue with the taxation office? That wasn't AI, that was humans doing humans do silly things which you've got to have the tools to be able to verify things. So it's people working with people on how to work with technology. We can't stop it so we need to adapt to it. Hi everyone and welcome back to the DUM Monkey Show. I'm your host for today and for most episodes, Devina Montgomery and I enjoyed as always by the wonderful co-host and founder of DUM Monkey Academy amongst various other things. A merecotub, a mere how are you? I'm good, thanks how are you? I'm really well, I'm especially good today because we've been joined by Special Guest. Which is always fun. Absolutely. So everyone, we'd love you to meet Mark Edmunds. Mark, tell us a bit about how we managed to drag you in. Tell us a little bit about you. Feel free to throw in how we try to. I've known at me for a long time. Yes, prior to the long authority. And hence that's why me because I'm here to come and have a chat about AI because I've been pushing AI and I love AI. But my background is I've been a retailer for donkey's years. I've run a motel and built a motel. I'm non-executive director on a couple of boards. Do a little bit of business consulting and a little bit of podcasting and a bit of podcasting without plugging through business stories. Yes. With my friend Bill Winter. Well, thank you for joining us here today. Pleasure. It's an absolute joy to have you. Mark, what have our favourite things to do when we get our guests to come into the studio? Is ask them how are you using AI? So I probably a bit of an earlier doctor with AI. I remember reading about chat GP2 two, three years ago, maybe further back and thought this sounds interesting. Actually, the day to day we are recording, I think there are things that are at any worth three of chat GP2ES. That's not that long. That's not right. So I remember reading about it and then registering on the chat GP2 site and then getting an email saying you've got access. You go into the early open, I like this. I just started playing with it. But my mistake was enough to obviously learn is I was treating it a bit like Google. So kinging things to ask questions. And then going, oh, that's interesting, but what's different? And then I started to read a bit about prompts and then Amelia with the monkey academy probably in the last year. Yes. I did a series of prompts and I thought, aha, the light bulb went on. So that made it more efficient. But I've I used it a fair bit. I've had to re-document so I've got it opted into chat GP2 five now and had it do a summary for me, give me key points. I'm also in the AICD bar when Seth with regional committee and we were running an event down in warnable and I thought I need to promote this on LinkedIn. And then we had a bit of a phone call and someone said, oh, they'll go and use some creative software and they'll go and do this. And I thought, no, there's a quicker way. Interch at GPT five. Can you create a tile for me? And I did the brief, you know, you're a marketing guru. I want you to. You're really good at this. And I need a photo of Karl Morris and Alex Patley, who are the two speakers and it did a magnificent tile with all the key facts, everything. The only problem was the Alex Patley and the Karl Morris and the photo were not the, that we, so that just reinforced in my mind how you have to be certain. I then uploaded two photos and it fixed that up and all up took me five minutes. And then I posted it on LinkedIn and it was gone. So and do you have any designs? Because would you have been able to do it without chat GPT? No, I don't know. This is because I would be cutting and pasting on word or something and then getting frustrated. And possibly playing with Canva. No, this was so easy. So easy. And then I've helped my daughter a couple of times who said, I, what do you think of this job that I'm applying for? And so I've just gone into chat and said, right, I write me a cover letter. You know, you are the whole prompt thing and then send it to her and she's come back and said, you've used chat. Haven't you heard it? Yes. And she said, well, that's not how I say it. Yeah, but use it as the main tweaking. Don't, don't just take it for gospel. So like I try and use it recently, regularly, but mainly to assist and help. It's a great tool. Yeah. I used to tap into chat GPT or you are saying, I did, I did try a bit of talking to chat GPT. I did a summary of, I'm just trying to think what I was summarizing. I can't remember, but I did read my notes in chat GPT and then ask it to reform and it worked. But I probably prefer to type and upload, I think, I need to play with it more to get more proficient at it. But it's fun. It's good fun. So Mark, you are a non executive director. You sit on boards, you help people sort out governance. And like I always come to you with any governance questions as well. And where I've seen is that sometimes a lot of governance people are very skeptic of AI and they're very conscious when I'm talking to you. You feel like you are very enthusiastic about it. So I just want to understand as a governance person, like how do you sort of see the difference between the risk with AI and what you can do with AI? The risk is probably the starting point in having a policy. So one of the boards I'm on, we put together an AI policy because we realize very quickly, there were people in the organization that were using all sorts of versions of AI and you get nervous about where that information is going in the first and foremost. So we put a policy to get around it to make sure that we'd set some guide routes. The state government is you know tensed at favorite co-pilot because they've got the Microsoft 360. But if we're only using the free version of co-pilot rather than the enterprise version, you just sort of have running with a lead strap beyond your back. So there's obviously got to be a bit of expenditure to get everyone on enterprise to get the bright respect out of it and using it from. And then at a board level, having a discussion around AI, where does the board see the AI going? And that's the interesting bit I think because there's still a fear of AI and there's still people read about the stories that it hallucinates. So that immediately fixates. If you're not interested in it, then you've got an excuse because you could tell on it's not actually hallucinates. Well, no. Let's check what we're doing and it's how we use it and how you check it to make sure you're getting where you want to be. So where do you think some of those few points are? Mark, what's driving those fears at a board level and I think it's a really interesting observation of yours because I feel like it's probably something that's been experienced very broadly. Yeah, talking to friends of mine on boards, a lot of it boils down to a fewer, a reasonably comfortable with technology. And I use the example if you know how to connect you to the Wi-Fi when you're in a strange location, then you're probably okay. You can probably navigate. But if you can't connect to the Wi-Fi and then you blame the laptop and then you say, "I'm the terrible connection." And then you call your kids an awesome person. You are probably going to be reluctant to try anything that's new. So and it's just a fear thing at the end of the day. On the other hand, if you're reasonably proficient at being able to do what you need to do, then you'd be happy to have a chat with AI. And I also wonder if it's a bit of an age thing and that might sound a bit age-ism. You know, I'm not a young man anymore, but I'd love this path. But if you're probably under 30, you're into it because you don't think anything more about it. You know, my kids are, they just use it. So what's the big deal? But I think there's definitely that factor of people that aren't used to the technology, might have a little bit of pushback against it. So it's trying to get people to understand what it's about and what it can do and take away that fear. And something else that I think tends to happen in, particularly in an organisation, space when we're talking about an organisation that's big enough to have a governance board, then one of the blocks that tends to come in is there'll be a piece of work done will identify some possible new ways of doing things, go down the path of research or testing, development, all of those things. And then it gets there and the people in the organisation don't use it. And I wonder how much there's a block in, I'm really good at what I do, I'm good at my job, but I don't know how to use this. I don't want to use something that might not be something that I'm good at. That's interesting. I've found that people within the organisations are really gung-ho on it, they really want to use it. Nobody's thinking, I'm going to be out of the job, I've not struck that. And if you get the right people to come and chat to you. So Amir's had it to a few people that I know. No, the guy Alex Pappley, you know, we've had conversations. We had him on our podcast. Alex, yes. Alex is a great guy. So you've got to get the experts into come and talk to you and understand what you're doing and have a look at your organization. And that's then helps you get through some of that. But I haven't struck anybody saying, I'll take my job. I'm not seeing that anywhere. But though, oh, no, well, that's not true. I did talk. We did interview on our true business podcast, a real estate agent who said they no longer use copy riders. Yes. Yeah. So you could argue that copy riders are a dying one of the first one of the first ones. Because in the firing line, I think how many times can you say, for bedroom has three bathrooms together? In any way. Yeah. You know, so they use it in a big way. So yes, there are some. But up to me, it's in the in where I play, it's more about making things more efficient and it's freeing up time to get the projects done that you want down, etc. One factor that I've also seen is, of course, there is a fear of AI. But a lot of executives that I have got to, they have used AI in their personal stuff. So they'll say, oh, I'm helping my kids do it. Or I'm, you know, I made this funny picture out of it. But when it comes to using it for work, they sometimes do not even understand how they can actually use it for work as well. So I guess there's fear in the initial part, but I know that there are a lot of people who are using chat GPT, but they still are not very clear on how to actually use it for their work. So I guess there's a gap that needs to be filled as well. It's a day of education. I think it's hard to possibly know where to go to get educated. So you're done monkey and me up. I'm assuming it's going and leaps and bends at me because it's lots of great information in this. And then it becomes up to the individual to apply it and use it and learn. But the information I gained and I was a big, as you know, I bought your book and then handed out quite a few copies to lots of people. And that got people talking about the book and not so much the book, but the information within it. And I kept saying to people, use the sections I talk about prompts, use and have a play. So it's, to me, it's the education. And I'd be interested to know, I don't know, if you know, is, is, are the schools in teaching about AI? Or they just seeing it there is something that they're using to do their homework and not, yeah, where's the education in there to bring people into it? I think education is still struggling. I know some tapes that we are working with to develop micro credentials around it. But like schools are in a very uncertain stage where they don't know what to do. I know Kardini, what they have said is, of course, we want to use block chat GBT, but you can use co-pilot, which is, which is at least a good approach. But it's an interesting space. Everyone is confused in terms of what to do. Like last year it was all about BAN AI. At least I am seeing organizations evolving from there and saying, no, we don't, we can't ban it. Like it needs to be there. I remember probably four years ago. So chat GBT is three years also, maybe four years ago. I was sitting at a AI CD function and we broke for lunch. So there was about 10 people, 11 people sitting around a table and I just asked the table, how many of you are playing with AI? And when around the table and none of them were aware of it, none of them were using it except person that came from one of the tapes who said, we're trying to work out. How do we use it and do we allow the students to use it? Because this is quite scary how this is going. You come forward to a year and a bit ago. Same sort of setup and I asked the same question. Everybody knew about it. Yeah. And the unies and the tapes were, yep, we're encouraging use with, we can't stop it. So we need to adapt to it. And I think that's part of it is evolving and saying, yeah, where do you use it? I do some reviewing of AI CD entries for grants and we get about 40 applications sent to us. So you read them and score them very easy scoring system beautifully laid out. And when we had a webinar catch up just to talk about the process and the very first question was, now do we mark down the ones that are definitely written by AI? So my question was, how do you know? Oh, it's easy to tell. Is it? Oh, yeah, yeah, it leaves gaps afterwards and it, yeah, but so wouldn't we encourage people to use AI to make it a better application? But it doesn't take away from what they have to have in their knowledge and what they've used it because it can be a generic response and you go, well, that doesn't answer the question. Yeah. So it's interesting. And that was yeah, that was last year. So there's still that learning process. And even the gelong business awards. Would you nicely one, I mean, we'll just give you a little plug. Business leader of the year. When we sat talking about the entries, there were a couple there where you could see that there was a consistency of the, so had an AI influence on it. And we talked about, well, okay, what do we think about the AI? And to me, it's just another tool years ago when you were using a word document and spell check was there. You used spell check. It should have been using grammarly. Oh, then you had grammarly and you had grammarly. So AI is just another tool in the in the data doesn't change what you write. Yeah, just can make it a little bit more. And it can be the uses that you're talking about, even in reviewing your own document or asking for some suggestions. It could be a really nice entry point if it is something that does make you a bit nervous. But of course, I mean, you've got to be working with a whole lot of businesses and interacting with a whole lot of businesses that are using AI whether they know it or not. That's right. We talk a lot about this. When AI is being integrated into so many parts of you mentioned using Can, but before we're using Can we use AI now? Yeah, that's. But the point that you raised about using it as a tool, I recently was checking job applications. And I saw one job application that was very poorly written. And I generally don't respond to such a team, but I responded back to that person, saying in this day and age, you do not have any excuse to have a poorly written job application with problems or errors and things like that. Like that becomes the standard now. If you're putting in something, if you're sending a job application, you should at least have that checked with chat GPT or some of us. You know, interested in the job. Maybe yes, especially applying to a company that feels in IT intake. Yeah. Yeah. Like it's just become a minimum standard now. Yes. Yeah. You talked about education, Mark. And you said, it comes down to the education. Who's in an organization, especially I'm not talking from an individual perspective. Who's responsibility is it to ensure that that education is provided to the to the board members, to the executives and to the team members? It's probably different. Yeah, there's different levels. I would say the board needs to educate itself. So you need to bring somebody in to just present to you. This is what AI is. It would be ideal if they'd had a bit of a wander through the organization and talk to people. So they could say to the board, look, he has three quick wins. Because boards love quick wins. Yeah. And then probably they're meeting with the exec to just, what are your problems? What are you trying to solve? Yeah. And then helping them understand what they could solve. And then possibly the CEO probably would have to drive that to say, right, how do we educate people? How do we bring people in? The individuals that gets interesting. Because some people are happy to pick up and read and use your dumb monkey academy or whatever. And others just won't. They'll be busy keeping busy. So I think it becomes then that sort of the managerial role. You know, if you've got an exec who's embracing it, it will populate down through the organization. If you've got an exec who tells everybody it's a hypercrap and dangerous. So it's getting over that. And I'm sure I'd still laugh. I still think, you know, when Microsoft Word first came out, I'm sure there were people said, oh, there's this is witchery. So mobile phones, mobile phones. A lot of people didn't want to touch them for a long time yet now. So it's that learning is really important. And I'd be really interested to see whether, you know, if you look at schools and, you know, what year do you bring in AAR? Do you start teaching students around year 10 about AAR? Then you've got to think about the teacher. The person that teaches something at school, sometimes is multi-skilled. So they might be the sport teacher and the English teacher. Then who's the AI teacher? So who does, how does the teacher learn for them to be able to introduce it? Because to me, it's a skill needed in the real world. And if it's not being taught in school, it's when do people pick it up? And that's when they pick up the bad stuff. That's when they know. Yeah. Yeah, I don't know. I don't know. How you had much to do with schools? I mean, I have. And I've been in a, like, talk to few advisory boards. as well. I think the challenge for them is twofold. First is that AI, all that's an amazing tool. It's a very open-ended thing, right? So if you expose students to it, especially people who are young, it can be there with a lot of potential to create harm as well. Like we have seen cases where AI encouraged a student to do so. I'd have seen cases where AI can encourage people to self-arm as well. So you almost need to implement AI and teach them how to use AI, but in controlled environment as well, which current large language models they don't do it very well. It's a great startup idea. If someone wants to take it, if they can actually create an environment which is controlled where students can play with it and be safe. So at one stage, I think that students at a particular age should be taught AI, but they're controlled in a managed environment. But it's very important for us to teach them about AI as well. It's interesting is that because like, whatever there's anything good, there's always someone that comes out with the default sort of. So therefore you get these AI generated graphic pictures. It's so easy to do it now. Earlier, you needed to know Photoshop, you need to know Canva to do that. Now you just have to imagine it or it would come back to you like we use Grock model and then a lot of fun with it. It can start to take you in some dark territories really fast. But we teach kids in school about social media and it was very much the same the same conversations that come in. Obviously, we've got the social media ban from the 16s coming even effect as this week. A couple of weeks away. Yeah, that's right. So we've had to adjust. The education system has had to adjust and individually at school levels, particularly I'm working particularly talking about students under 18. We're not talking about university and type students as much here. But the reality is it's there. It's there as everywhere. It's going to be used. Kids are going to use it if they can get access to it, which most of them can. Right now, I'm not even thinking about school students because I'm first of all worried about the graduates who are coming out. People who are spending five years studying law and what their feature is going to be. But then also the organizations that are prompting. They need to link up questions. Yes. But once agents start to take over what would be their role. But also with the organization now embracing AI, how would you start to upskill people who are currently in the organization? Because they need to upskill them so really fast as well. So I'm seeing a huge skill gap that's going to happen. So for the kids who are about to graduate who are actually hoping that there isn't a big skill take up because that's the pathway to get a job. Be out there. The people that don't get it anymore. We've had this issue before. We've had the technology take up. Then becomes a bit ages. There is a gap. It's really interesting. Because it's it's so easy to use. And if people grab it and use it. But then how do they? It's all very well to sit there like I was playing with Nana banana the other day. Yeah. Which is hilarious. It's really good. I just said give me a put Bill Winter's picture in Hello Bill. Yeah. And so give me a picture of an old man sitting on a hill in a thoughtful way. And bang. There's this Bill sitting on the side of the Himalayas. Yeah. It was so accurate, so good. But I could have done evil with that. So it's how do you put those parameters in? How do you stop people not being nasty? And how do you use it as an effective tool? Yes. That comes with those government boundaries. But also helps you to achieve the potential that's sitting there in your organization or your business. Because that's the whole point of being a tool. That's the whole point of being a tool. And especially around the other the chats with the media, the chats with Alex, you know, how you can use it in helping analyze financials in an organ. I'm saying, sorting out your filing instead of someone sitting there and trying to use SharePoint and put it all in the right boxes, put an agent over the top and set it off. It's a really interesting point, isn't it? Because I think we've talked a lot about the kind of things that, realizing where your problem areas are and then deploying potentially AI into that space and teaching people how to do it and putting some governance around it. But you could almost flip it and say, well, what are the things that we know we should be doing, but we're not doing because we don't have time more, it's too boring or we can't get people to do it. And can we get AI to do those things? Can we start to get Phil as well? Look, I think next year will be a really interesting year because I think it's building leaks that are starting to understand it better and want to understand it better. And then you've got people that are the danger is I suppose on the other side is there are people out there saying their AI experts and could train on it and they're not necessarily because there's no, you could just put a shingle up and say AI expert. How do you know? There's no qualification, no certification, no certification. And I remember talking to somebody after you're very careful about phrases about an issue they had and I said, I said, look, I'm not an expert on this, but it strikes me that a very simple AI agent could help you sort that out. And they said, yeah, I'll talk to my AI expert. I said, who do you use? And with all respect, they were a marketing agency that had rebranded themselves as an AI expert, right? No, more. That's a long stretch that you're used to marketing and I know suddenly you're talking to people in that AI. It's interesting. It's entirely possible. But I think you raise a really interesting point because we're working on what our suspect is going to see a lot of in 2026 and moving forward is that you're going to have more customized AI. And this will feed into the idea of how do we put governance boundaries, how do we deploy it in a way, how do we help people use this effectively? We will have more customized AI's. So whether it's a large language model, a small language model, but it will be specified that we can obviously solve a lot of ways to cut it. It has an agents attached and coding and things. But if we're starting to get very specific AI usage is so it will be me give me a large organisation. It might be West farmers, I'm sure, has their own AI, no doubt. And there'll be plenty of others that do as well. Who's building them? Because that's great, doesn't it? That's a real solution. We don't really want to look. Copilot's fantastic, but does it really fit our use case? Love some of the stuff over at GPT, but we're worried about using chat GPT. But we should get our own because then we'll start to learn our systems and we can start to develop these agents and we can really make this work. But if you don't have a good system from the beginning, but hasn't been built particularly well, that's going to be a problem. And who's teaching you and who's? We don't want to point fingers at the bomb website, but it's not exactly like we don't have a great history of people charging very large amounts of money for things that don't necessarily fit well. And it's interesting too. I was just I read through the State Government Silver Review on the structure of the public service. And there's a big section in there on AI and how that which is great and how it can help the organizer and help the public service. And it was interesting. I think that one of those recommendations was merging waterboards, which has been knocked back. But what was interesting, what was the driver, I think behind some of that was each waterboard was delivered, was working on its own, I think CRM system or some back of Hases. But each of them were different. They're all water boards. They're all in Victoria. But I can just see it. The person who's driving it in each of those places will be thinking they've got the best solution. Oh no, we won't use this because there's his crap. Yeah, well that won't work for our use case. Whereas the push is how do we standardize some of this and using AI to get the best results. So I think it'll be interesting because I did sit in on a public service briefing on AI, but a month, a couple of months ago, I was fascinating. So there were people there from the department health who's openly said we're all over AI. We're using it constantly. And it was, and it yes, there's that, you know, they're using co-pilot, but they're also developing their own models. And that's not also suggests that the Department of Health is one of those areas that has very poor data standardization and not on not just where the Victoria Department of Health, I'm talking about departments of health, of renowned for having really poor standardised data. Well, that was interesting because that's where they were looking at, but they were also really conscious of the whole privacy and the government around it. But it was really interesting. It was a really positive, this is where we're heading, and we need to be working together to get the best results. So I think that'll evolve. Yeah, which is, I mean, that's so positive. That's really good. It's a lovely, I find that really comforting to be honest. I must have attended the call thinking I'm going to be yawning and I'll go off after 10 minutes and I didn't. I stayed all the way through. That was about an hour and a half. Yeah. It was really interesting. Because there's some fascinating ways that that our public service can be using AI. And most of that is very people focused. Most of it's not so much. I mean, there's, of course, there's back end systems and processes that could be put in place, but the ability of AI to help people interact and have a better experience with the public service because we're all interacting with public service. so often is amazing. Absolutely. And it's, you know, it's more so in your palm of your hands, like I've used an Android phone. So I've got the Gemini, the Google version. I'm still playing with it. When you're not going to distract a banana banana, I feel up now. But it does hallucinate. And I thought it's very keen to give you the answer. It's good to give you the answer. And it made me realize how well it knows where you are because I just as a silly, I was looking for champagne for Christmas. I thought, oh, what's the top three champagne in 2025 that's one award side just as Gemini. And bang, it came back at me. But it told me that Ves Felix, grown on the Bellarine was the, and I thought Ves Felix is an on the Bellarine. I live on the Bellarine. Now I'm thinking, wow, that is really interesting. Because it really wants to give you an answer. And it told me the third best was Oak Dean. Well, I've Dean is the one who just ran the corner from it. So I thought, no, you've given me an answer. You might think I want. So then I rephrased it and said, listen, you know, there's a one kind of certain, well, and then I got three good points. So you still have to, you've got to be have your wits around you. Because if you didn't, you'd probably be running around bottle shops looking for Ves Felix from the Bellarine and they'd be looking at you as if you're an idiot. So, yeah, and Bill's family are going to be looking for him up. Yes. So it only comes, but it's again, it's handy to use it. And I find it quite easy to just say, hey, but like any tool, understanding the, the pros and cons, understanding its limitations and understanding its capabilities, that's kind of the point. And we get, you know, we get experts, you know, you can get, you know, expert that comes in and pick up a, pick up a tool and do something pretty incredible with it and you can get those of us who, you know, you hit our thumbs for the hammer and sit there and cry about it. But it's going to be, it's going to be a very broad school, I think. Yeah, it's interesting. Okay, about experts, we looked at, there was a big consultancy that submitted the report for government and there were a lot of issues over there. The reason that case, that is interesting for me is because they would be expert in governance as well. And they would have applied some sort of governance right. But this is a case of their governance going completely wrong there. What do you think would have happened over there? It's an interesting question. I would say someone just got slack and lazy. That instead of reading it and just saying, right, let's just do a, you know, pulse check of what we're put together. It was, oh, that looks okay. And whether it went through their risk assessment, folk to say, is this right? Is this accurate? It was, yeah, everyone just went, yeah, yeah, yeah, it's good. Which is no different to when things go wrong without using it. Like if people drop off what the process is and don't follow the governance process, the risk process, then you end up with that debacle that that happened. But as into goreness in place to prevent that from happening, you, you would hope it's in place. And you would think the consultancy company that got caught out here would have all those risks. And but then remember the issue with the taxation office with people being briefed and leaking it to their clients. Well, that was an AI that was humans. And humans do silly things. So it's about having the right guard rails there to say, right, oh, this report that we're preparing, who checks it, who double checks it, stick it through another AI agent and tell us if it's, hey, yeah, good, this and you might be frightened. So it means just having that E.A. policy or absolutely work is, is, is okay, but it's not enough. I think it's, it's good enough to have the framework, which you've got to have the tools to be able to verify things. There's a danger that if you just read it and say, well, that sounds right. As opposed to, well, let me just spot. How do I know that it is right? How do I know it is right? That's the tricky bit. Right. And I think in putting reports together or or feeding reports in on analysing a business or recommendations to a business, there's got to be that, human check. And part of that process probably should be done issue it, don't submit it until somebody is checked into spot checked. You don't have to go through the whole thing, but at least just using it, using an AI agent, but an AI agent. Why not run it through another AI agent who might go, what are you saying? So, yeah. But like taking it from a skeptics point of view, they are saying, even if you use gopilot 365, it can still hallucinate. Now you are giving this tool to everyone in your organisation could be a large organisation. People can use this tool at their own level. They can create stuff, not check the stuff and start to push things out. So it can cause a lot of reputational damage to the organisation. They can make wrong decisions, talk to the home customers, share wrong information. Like even with the policies and governance framework in place, what can we do to stop that? I'd come back to education. I'd come back to that the individual using it, the need to fully understand it is not perfect. So yes, it looks perfect. But when you get a response, just double check. And it looks perfect and it's quick. Yes. When you're under pressure or you're trying to cut three things quickly, the temptation is there. Yeah. And chat GPT has, you know, quick and different. Yeah. And there is a difference. You'll get a very quick response and it may not be as accurate as you would like it. So to me, it's about the education and having the understanding that don't don't accept it 100% every time. There's some things you'll look at. Yeah. That's easy. But also what can you do at an individual level to be a checking your own responses, so getting those responses back and doing a check. But then we'll think to look for it and you're right. It is an education. It's education, but there's nothing stopping you using 365 and then feeding it back in and say, I want you to review this and tell me whether it's accurate watching. So and give me the sources for the source. That's a really important. Yes. The thing is he's looking when it's asked, and I've noticed chat GP does that gives you the little link on where it's got the source from. And some of it's a bit vague. If you go in and look at it, you go, okay, I understand why you've used that, but that's not quite. But the. And it's also heavy. You would have done this. I've done this when I've put that in and gone back and tried to test the test the source links. Has it given me the right, the right links? Yes. So even things like understanding, you know, as you play with it more, as you start to understand more about effective prompting by all means by going through the dumb monkey. I don't know if we would hope. But things like putting your prompts in that the boundaries around where those sources can come from. So it will be this source, this source, this source, this source, it could be sources from here, this site, all this particular link of this file. So that you're starting to like anything else. Like when any kind of job that we're doing, it's not bound real. It's. It's no different. We always have we box everything. So you have to do anything. I'm trying to come up with a solution for this and I've thought a lot about it. I feel and the way I try to explain this, how you before giving someone a car to drive and getting them out on the road, you actually get them through a test and give them a driving license. I think we need to have that AI license. So people need to go through a mandatory training at test before they are allowed to use AI because it's a powerful, but it's a very dangerous tool. And then you have road rules in place to make sure that they are following those rules that there are penalties in place and checks and so on. There's a job for you. If you came up with a module of AI and educating people had to use AI, and then testing them on it, most businesses these days use online training. So it would just be a matter of slotting a module in and having a compulsory AI module as part of the governance process. So you've got a policy on AI. Start with copilot because we've all done a training and then all those additional training for SharePoint. Teach people just very simple uses of copilot, but also testing them on it. Okay. Like you do with doing fiber security training, don't click on that link. Here's an example of which of these looks a little harder with AI, but at least you could do a very simple 20 minute. This is how you use AI. These are the things to be wary of and then test people's awareness. There you go. Thank you. That's a good idea. But that's yeah, you're right on me. That's what's missing. That education again is that that bits there. I'm self-taught and I'm nowhere near what I've good as I probably could be. Yeah. And I use you, you're done monkey information. But I go in and play with it and then I walk away from it. But if you're in an organisation and you're wanting people to use it, you've got to have some education around it and strong guide rails. So yeah, I think it's interesting. And people hear the hallucinations and they end up, that becomes the reason they don't want to use. That's right. Okay, but how do you check for that? What are some of the tools you can go? And the links are really important because I mean, all chat GPT and co-part, it's scaring the internet for anything they can do. It's finding sources and it's always something up and lots of people have read it. So and it's ended up load of rubbish. And a whole lot of the internet, content of the internet now just on the web is, AI generators. Yes. So is it self-feeding itself? Mm-hmm. Well, he is. Yeah. So we talk about hallucination, but one of the risks that we've talked a lot about this podcast is people putting in the information that you're not supposed to put. And so people say, like, organizations say, "Don't put any confidential information into this." But it's very hard to determine for a layman to first classify an information as confidential or not, whether it is something that they can put or not. So do we need to develop a system where we assume that people are going to put a lot of work related information that could be confidential and putting the guardrails around it? I'll go back to what is confidential. So there's a lot of, you know, I don't use chatchapet because, you know, if you put confidential information, it'll go around the internet and everyone will know what it is. But if you're putting out information about a business plane you're doing, is that confidential? Is that confidential? Yeah, is there any risk? If you're putting out all the names and addresses and phone numbers of all your team and you want to know how many people live in a certain area, well, that probably breaching confidentiality because that's readily identifiable. And a lot of information in the health sector is absolutely confidential and with high risk associated. So it's always making an understanding of what is deemed confidential. And what could harm the organization? Now, if you put up all the financials of an organization if you're in a company and someone's looking to find out, you know, how well you're tracking, then if that information's there and they can find it because all you've got to do is ask chat and say, "Go and find me." That would be interesting. You've got to have that protection on those sorts of things. Again, I'd come back to education. You're back in, you know, this is, don't put this up. And there's no reason why you can't be vague about something to get the response you want. So you don't have to name the organization. You can say, "I am in an organization that and this is the information and I want you to exit with it." You don't have to name the organization. I must say that I've done a little bit of that because I'm just conscious of where it goes. But you can add to an off where it goes. Absolutely. So again, that's a trick to be known as a part of the organization. It's a part of the education. So you're back into education again. Everyone's sort of saying, "I'll co-pilot or enter co-pilot enterprises the way to go." But what's stopping Microsoft skimming in the back? And the reality is that the space is moving so fast that there's going to be new products, new developments, new companies. We are going to have more agents than we're going to have more systems that are built specifically for use cases. And then you just don't have AI and co-pilot. Now you've got AI and Salesforce. AI and this system, that system. So every system is coming up with their own AI as well. So that does being access to a deal. I heard a story the other day. It's like all these things urban myths get created around these things. But they made a comment that some would be in an organization, asked co-pilot or chat GPT. What was the best way to go about asking for a pay rise? And the response came back. No point asking because you were damned to be made redundant. Because all the data and all the information had been fed into chat GPT to formulate some package on where the staff is looking. No, I don't know if that's true or not. It could happen with Google by let because especially if your share point is not so dirty well it can. Where's the source start? That's interesting. It is interesting. So you know that to me is well if someone in HR is uploading all that information in the or they're using co-pilot to sort the filing out and find information. Have they put parameters on it to say you know don't let people below this job grade or whatever. But that's a scary thing. Yeah. And again, I'm not sure whether it's just an urban myth. A lot of them are not. They are. They are. They are. They are. They are. They are. My feeling is that governance in AI is still an unsorted area and the traditional governance is failing around it. It's probably an evolving area. Yes. As we understand what AI I can do or we don't want it to do. Then that governance needs to evolve. And who wants that responsibility for that as well in the organization? Well, I can give you a live example. So I'm on the Great Ocean Road Coast and Pymes Authority board and we've got an AI policy. And we've asked the er auditors to come in and audit our policy and how we're implementing it using it. And we're treating it as any other risk. So we do a cyber risk. We do a payroll order. So doing AI check. That's I think really good. That's the good question. That's a good question. That's because the auditors will tell you they are the ones we use. Yes. We'll only know when we start to get the report back, which is next year, to see where they delve, don't where they've got. Because I'm not sort of boasting. I have looked at a lot of AI policies and I soon as I look at it, I can start to poke holes into it. Yeah. But that's true for the policies that I have had created as well. Like someone can come in and start to poke a lot of holes into it. Because with any of the tools that we were using previously, there was certain number of things that you could do. So the risk were very well capsulated or defined with AI. So much that you can do. So the risk are so broad as well. It's a suit. So that even AI policy, I feel that cannot be a static document. It needs to be that living document as well. Correct. And I'm pretty sure we've got a policy coming back after six months to have a sit down and review. Hey, the auditors coming in to look at it. So do you think it's board's responsibility or? I think the board has to set the policy management or create the policies that board have to make sure they sign off and interrogate it. And how informed does the board needs to be able to use off AI as a result of the education. Yes. So how do you bring all the board up to a comfort level of understanding what's being put in front of them? Because if you just seek a AI policy in front of them, they'll read it as any other policy and think, "Well, looks okay." But is it doing really? Yeah. And I think that's where you need subcommittees within board. So on the Great Ocean Broadboard, I chair the Auditor and Risk Committee. So those policies will come to us. Now, because I've got an interest in it and the other members of the committee are also attuned. We're asking the question, "Is this right? Have you done this? What about that?" And the people writing who've written the policy and brought it to us, they're very much into the AI space. So they're thinking from their end, "What are the things we need to be aware?" So you need that expertise. You wouldn't just give anybody. Don't give the accountant, so write me up. I know I'm policy. Or the marketing person. Yes. It's got to be someone who understands governance and understands the risks that are attached to it. And I wasn't having a very interesting conversation with one of my director friends. And I was arguing that for you to be able to AI, Gov. and AI properly, you need to have a hands-on understanding of what's possible with AI. But this person was saying that as a director, I manage and gov. And a lot of things in my organization. I don't really need to know how exactly everything is work. I need to know how to go on. Especially with the AI, I need to know what other risk and what other challenges. And if I know that I should be able to go on it properly. As a director, you're responsible for a lot of things across the organization. So, and lots of boards work with subcommittee surf to work on so that they can then deep dive into areas of knowledge. But that doesn't mean the board abdicates when that subcommittee report comes to the board. The board needs to interrogate it as well. And that's thing to me comes back to education that the board needs to have it. And they don't have to know the needy gritty. They don't have to know about all the prompts, but they need to understand these are the tools that it can be useful. And these are the risks that are attached to it. So it's back to that education piece. It's funny how it keeps coming, but that's that's the place. Education, you don't need to know the intricacies of how chat GP2 works, but you need to have an understanding of why you need to have some guide rails there. Why you want them there. But I also wonder, is it even possible to educate someone on AI unless they experience what's possible with it as well? Yeah, that's interesting. But maybe that's part of the education. You know, very simple, very simple exercise with somebody. Yes, he goes use the AI to create a document on X. And just see it. Because once you ask the question or prompt it correctly and get a response back, it would be interesting. Yeah. But I don't think you can't have every director absolutely over everything. No, it's like that's a fine line because that's operational. You're getting into it and you want the operators to know. But like the example that you pointed out where you said, you know, that urban myth that that chat GP2 had access to the files. So chat. Generally, do not have access to your files, but something like co-pilot has access to your share point email teams meeting and so on so like so a very simple Case over there is that if you have like a general room where Microsoft teams is connected and people are doing meetings over there and if it's a General email ID automatically everyone has access to that meeting and everyone would have access to the transcripts if they would ask a Question about what happened in that meeting it would provide them with a response now it's a very nitty gritty thing like so it's a very operational level thing but I feel as a director if I don't Understand that risk that it can happen so that would that to me yes see how you would break it down in the risk register on under the use of the I/O Yeah understanding that well the organizations on Microsoft 365 teams is connected everybody that uses teams because you do you get the Feed and oh by the way have any of you tested out asking it what was discussed in that meeting can you give me a summary of what was Discussed in that meeting that would be interesting to test it out so you're talking a little bit of me about having what's the Roll of the of a potential AI champion in the governance so you've been forming how the governance is then said yes but I was talking about that was a Use case that we finally picked up when one of the employees we were having a conversation about the the I/O policy and then someone raised or we have got a Teams in our generic meeting room everyone has access to those meetings and it can come up but we had like some really good decent AI experts in there some good governance experts in there we are already had prepared a policy or a risk register but we didn't Have that in yes have that in there and then it would go to the board board and if again you have got directors were not informed about it they might not even stop to ask that Question as well because you need to know that that's our teams operates yes it would have to be recorded for teams to have stored the information from that So what happens is generally it's set to auto record especially in meeting rooms so you risk management and there's always a board meeting as a member who is that joins that so what happens is in Co-pilot if you as a member has joined a meeting then you can have access to that transcript of that truck settings allowed which means your Co-pilot can have access to that transcript so even if you are not there and you're that particular email wasn't there then you can have access to that transcript yeah so and you can assume that that well even while they might be a policy of having that setting turned off you can assume that Canon will happen so how do you yeah you said some risk to me if you were having a confidential discussion about well we're going to make 100 people redundant next week yeah ten that you would turn off maybe don't do whatever teams you don't well you don't record it yeah but but that's really about confidential discussion but there are a lot of like discussions that everyone does they need to be pivoted off as well yeah it's yeah because there's lots of meetings you go to that you use it to me it pops up in the corner yes me as agent or whatever it is yeah me yes yeah and I was at one today and and on the team's meeting and up popped this a 10 deep which was the agent for one of the people who was taking the minutes yes and I did ask are we comfortable but this is and there was a yes because they're doing it saying we're not talking about any confidential and they're putting it together and the person who was doing it said I use it on everything because it's so good because as soon as I hang up I've got one yes yes and I'll circulate them to you within an hour yeah so it's that balance up yes that is the case you get asked to turn it off for that you long authority yeah and sometimes I actually go in the meeting say can you please let this is it out because it automatically joins all the meetings there should be some modern there so that doesn't join everything but the person who's letting it in needs to have that education whether yes because they're letting it yes what's the policy of letting someone's agent in and they are not saying you are not going on these got two connections without realizing yes it really says I'm a meet geek so you know what yes back to education yes yeah yeah and it is going to be evolving because they're good tools they are good tools and we don't we don't want to discourage the use of them because there are risks associated with them so do you think there is risk in that as well I think we're paranoid when people say that you know amaze agents up there do you think people get paranoid about it absolutely I think so like some some do like because now some people of course but yes absolutely some meetings I joined there like three people like say James in those like more more more notakers a lot of like all of our clients use notakers like says it has become a standard practice yes but then you just need to be aware of what you want to take an engine notes so that's interesting because at a board level when you record things like you've got to be understand what's discoverable so if something goes pear shade yes anything that's been kept is discoverable so you know unless you have a rule that soon as the minutes are done then that recording is deleted and I sometimes wonder when it's being recorded do you are you as full and frank in your conversation because you know someone's recording it and you might say things that in hindsight you wouldn't have worded it that way if you knew it was going to be recording you so just it's just interesting it is interesting yeah I must be going to become a bit oblivious to it I know it's recording and I just yeah okay yeah for my conversation with my team members I always encourage them to record me what I am saying because I don't like repeating myself again but then we are also building like a brain like of me of me saying something and then it gets captured because you can start to capture those raw transcripts and start to convert into structured knowledge policies proceed there so what if someone's to go you're recording Zemeer and for evilness created evil a mere yeah who then how the team's meeting with the book and says hey look I'd like you to transfer this money into this bank account and I'd like you to go and do this and I what's your parameters around that so I've got some safety parameters over there like you know I've got one with my family as well so as a like if you get a call in distress from me we have got a passcode which I'm not going to reveal yes because and so funny ANZ ANZ in the bank they're still in the mobile lab they say this is my you have to to make a payment you have to use your voice and say you know I this is my voice and I confirm my identity it's so easy to clone anyone's voice as a stage yes yeah yeah and they're still using voice to verify the the the the the spam email yes getting better oh yeah better and better because using ii you know stick out phone bill in there and send me a fake version of this there's even me I mean it's been some great stuff that came out even when I was reading news feed just this morning that we're showing that there's a tendency for people even when they know that video content has been generated by i even when it says it's been generated by ii they believe what's being said fake news report look I am really good at identifying the videos yeah these are saying but it's not saying that it's generated by ii and your people are still like oh yeah but it's still like you know it's still yeah sometimes I get confused with the quality of videos that are coming up because it's so difficult to know what's the data yes so this idea of this idea of source this idea of source is really coming from is really important um and you know i've got a best interest in this because that's kind of in majain for a long time is they've had a lot of conversations with with young journalists young writers or even people that are you know they're interacting with media to say just because something is published doesn't mean that it's real or that it's right go and find the source go and ask the questions be curious be skeptical to what's the way yeah and it's in these skepticism it's really important because if you watch the news and stuff these things happen or news feeds or whatever it's getting harder to work out what's fact it's very easy it's very easy to challenge something and just say that's fake yeah indeed yeah when it's not yeah but then it just puts the data yes it's an interesting place where it isn't just indeed my phone went the other day and it was me ringing me wrong really hi mark and you transfer me 150 bucks my my ties just like and call my mom how's my number and that's it's cloning of the number yes of course whoa yeah it's it's just it's just yeah and that the voice is really interesting because there's a bit of that happening where people get this phone call and say oh yeah can you send me money or help me and it's fake yes so that gives me that brings up a government's question for me how as a as an advisor and as a non-executive director how can you take a governance perspective on being mindful that mistakes are going to happen or the problems are going to happen and how do you create an environment that makes people willing to put their hand up and say oh yeah no sorry my bad to me the culture's really important that if someone stuffs it up that they immediately say oh dear I've done the wrong thing we had an example of that for somebody same doubt an email to the wrong email address very easily done What I was really impressed with is they put their hand up as soon as they did it They fixed it because they'd run the person and sort of delete that email wasn't intended for you and They hope the person or the other end are deleted it They could have said nothing. Yes And hoped nobody would find it right. That is a culture win Absolutely. To me it was the culture they put their hand up and said this is what I've done Yes, and then there was a very clear policy around that yes, and and it was managed really well Yeah, and my bit was did we congratulate them for putting their hand up? Yeah, we didn't beat them over the head for sending it because they weren't telling you next time. Yes I just thought that was really good and that's really important in this discussion Yeah, if someone's using AI and does something oops, I've uploaded It's going to happen. You know it's we know it's gonna happen Yeah, then they put their hand up and say look this is what I don't know Can you ask chat GB to delete everything? I've just brought up. I don't know Probably can ask you know So that's a good point because we talked about governance we talked about education but culture I believe is also a big missing piece huge part and yeah on a couple of ways Culture in the organization wanting to embrace AI. Mm-hmm. If the culture's one of trying and testing and learning Great. Yeah, if if the cultures of risk aversion and Fear you're not gonna test it. You're not gonna get it. So it's interesting. Yeah, cultures are really important back So the challenge for me is like it's for me as a leader for the organization that I'm running on managing It's easy to for me to impress the culture But if I'm going to an organization as an AI consultant Easy not easy. Yes. What can I do to remove some of those barriers? That's a really good question Probably the starting point was trying to find out who within the organization is Keen on AI and embraces it and then have that because you're the outsider coming in So you need to find someone on the inside that's people respect and we'll listen to that then you can have your I think you used to word champion have a champion It was it was actually a mere word was it? That making a channel in within the organization. Mm-hmm. So that they then Embrace what's being said and then can help people because people were probably more likely to go and ask them what they think Then be open with you and say oh, well, I'll someone who last someone who then go and ask them who get there Yeah, on the other hand because you're coming in Knowing far more than probably anybody in the room. Yeah There'll be a degree of people wanting to listen and learn And maybe it's addressing some of the urban myths first up So that they know these things happen. Yeah, let's part that And let's get on with the What I identified is that I go in there like we do the training they are keen to learn they do some things But then they get busy in their work And forget about it or they don't have time to like there are we definitely will find some winners from there who are good ones but That challenges how do you then take it and impurement it across the whole organization? Or how do you find how do you as an organization allow a shift where people who are very busy and time pressured don't have a lot of spare time to Conplay with things have some time to play with things, but that's possibly they're showing you Showing those folk how they can free up time by using But it's chicken eggs and you gotta get them and you get some people who are believers and they are like Champion they just get on board and they start to see some quick wins and some people they really want to do it But they're just probably do busy Yeah, and it's interesting because who's driving it within the organization? And is it the IT person? That's where it tends to land And they're probably not the right person to drive it is probably someone in the governance or the process side of things that Needs to be the champion of it Um, and maybe they you know part of it is putting a little team together to help cut through as a project to get like a project How you going to implement it? Where you're going to and working with Folk like yourself for me to work out where Where's the pressure point? Where's the place that's causing grief? That can be implemented reasonably quickly to show that it can work. I think that's that's the starting point It's an interesting idea to us because when we talk about change management or that's often the way isn't Yeah, so it's finding where's that one place where we can make a difference and show that it works rather than just you Because there's a ten a season it because AI is available now I've seen they've done multiple times where we're implementing AI and now it's everywhere and Everyone's still sitting there and been then gone into a meeting and say, oh, he's using it and everyone's Yeah, I'm doing a bit of a Google but To be able to show that use case inside in your organization that works Can it be an ask about a start? Yeah, I think you're right. I think trying to find where's the pressure point that You could solve or help And then you can show the people. This is what it's done Then you might get people that come up and say, oh, if you did that for there, hey, we've got this And at the end, it's all around data. Yeah, it's the data that's the issue And but the fear is real the fear is real It all sorts of wise many To me, it's no different to any other change management program It's about having champions and showing and proving that it works and and you will always have people in the background sniping at it. Yeah, oh no, it's load of rubbish. That's not going to work. Yeah, of course, but they're afraid of their job Hmm, so if you show that you're not losing people, yeah, you're trying to Or afraid of being shown up or afraid of yeah, it's it's it's an it's just a brave new And there are a lot of organizations who are like yes, we want to embrace AI but Too busy this time end of year financial year this is happening quarterly period and so on So they just keep pushing it I'm seeing some excellent Organization that we are working with that they are taking a leap ahead But then I see a lot of organizations still failing miserably or just yeah sitting on it just waiting for something Is that a fear factor in their part or a cultural issue within your eyes? I feel that it is a cultural issue. I think Especially during like with a lot of organization that we have worked with I can see that the fear is not of a big factor over there people have passed the fear it's finding the prioritizing it and sort of like focusing on the important rather than the urgent and I see a lot of Especially SMEs and even large organizations are focused on the urgent stuff They are not there's no someone who's responsible for it like CTO says it's not my job You know C.O. says not my job There's no clear budget item in there as well because it's a new thing as well So like marketing team has someone who has got a very clear KPI They've got a budget allocated to it. They know they have to do that work But AI is still in this no man's land for a lot of organizations And it's interesting and it's interesting whether you have a person who's the AI champion in the organization Or within the different departments you have so in the marketing Do you have the person in the marketing who's the AI champion? And you can get some quick wins. Yeah show that it works And then you have someone in the accounting and you have someone It's currently been driven not by a system but by a person's passion and energy So you'll have someone in operations is really excited about it So you don't have a very clear culture or system in an organization to do that And over time you can change things like onboarding and you know different training in the beginning But doing that in an existing workforce It's no different to other things that I've done If I think back over my career Oh H&S became a big thing So a whole and this is in my Coles martyr days And there was a whole push on Occupational health and safety and identifying risks And report slips and the whole raft of stuff And it was a big imposter on the organ on the stage Yeah, being part of the tide wouldn't it? But everybody A lot of people push back Oh this is the waste of time while we're doing it But it was a cultural change that gradually You know just being educated We kept having refreshures education And Then it was embraced Now it's so embraced in my head That if I'm walking through my Coles or woolly supermarket And I was a plastic bag on the floor, I pick it up I don't walk past it Because I think someone might slip on the So it's getting that cultural shift And this is no different It's a new thing It's suddenly everywhere And people want to learn about it But it's how do you bring it into the organization And get everyone to embrace it And it'll be something else in another five years There'll be something else that comes it Look at cycle social hazards that is now being legislated We would have known two years ago Three years ago that that was a big thing But it's a cultural shift The organization has to embrace it, understand it And again it's education You know There's modules being formulated So everybody in the organization goes through it And understands This is what to look for This is how you report it This is how you'd You know So it's people working with people on how to work with technology Yeah, I do I think you know, that education bit, I think is that key. And I don't know whether you need a single champion in the organization, but maybe within the different departments. It probably depends on your organization as well. It's a structure as well. It's awesome. Yeah. And some CEOs will grab it and run with it and want it to happen. Yeah. And some CEOs will go hide behind the door and say, can someone else please do that? I'm really busy. Yeah. But and it's interesting, the excuse of I'm really busy, then is, it's just a means of pushing it away. If they really wanted to learn and make it happen, yeah. They'd embrace it and and and over time you run it over time because of that cultural spread and how it just will become part of the become part of the CEO role is that you will understand how I know you'll understand how it gets implemented and where the risks are and all of those. And is it scary but is it fun? Yeah. Not a bit. Ah, come on. Yeah. I just think that's hilarious. That's good. It's just it's interesting coming back to you should should people have an understanding and use the way I probably at a low level understanding and at least having a play with it because most of the people I know have a playing with it. My next door neighbor is all over it like a rash loves it. He's found it hard to get it into his organization. But now he's been recognising the person that keeps talking about it. So now he's running things. He's becoming the organization. So he's becoming the champion. So that's going to be a big trend that you're going to see. These people who are fanatics about AI, the is AI champions. Other ones who will find themselves to be in very important positions within the organization. And he loves it. He's reading everything I gave him your book. Like he's he's he's pushing hard. Yeah. And initially found resistance. But now people have stayed to say there's something in this. Yeah. And it's working in his favour. So do you think Mark that I'm not think I know the answer to this but I'm going to ask it anyway. What do you think about the potential for AI to make people enjoy or allow people to enjoy their job as more? We're back to the working from home scenario. Could the AI be you at work in the office when you are not in the office like I don't know. So if people ring your phone and you are not in the office and it goes through to your office, you answer the phone and it takes a message and this is I'll get back to you. And like who knows where it could go. But being not big silly. But yeah, I think the AI should be able to and a mere knows this better than anyone. Remove those tasks that are laborious and consuming and repetitive so that you can do the stuff that's more frustrating and I'm mind numbing and that they're the things they're keeping talked about. And it's always in the back of my head when you know, people throw spreadsheets up in front of me. Like who's sitting there king all this in? Why not take the data and ask AI to produce the analysis? Yeah. That's that gaps there. Yeah. And that's where the fear comes about off out. I upload that information. Where do it go? So it's yeah. Yeah. My view is that those death. They walk is definitely going to be gone. So like you say I can definitely replace that desk kind of work. We would still want people in the office, but generally in the cafes and in the the meeting room, the smoking areas, building those connections. Yeah. Because those connections would be because start to become more and more relevant. So you would have more free time for people to actually be able to enjoy their work and then of course not do this laborious work. But make better human to human connection as well and focus on their health like networking. And I see a future where you would have a lot of new businesses that would come up that would be just running more health programs networking and things like that. Because anything that provides that human to human touch would start to become more and more relevant in the age of AI as well. What about innovation? Yeah. If you free people up from the laborious stuff, yeah. You free people up to be able to look at their role and think, well, what if I did this? Yeah. It is. It is a great body of problem. The AI and say, what about is that human element? It's not going to just happen. Yeah. And even then it's going to largely come back with an answer from the knowns. The great possibility is that you free up people to actually sit back, take away the things that makes them sort of go running for their 3pm coffee and waiting for the end of the day to go, wow, this is in front of me. I haven't had time to actually even think about that, but now I do. If you've got time to think about it and talk about it with a few other people, then what is possible? And it's all that you're right, Amir. It's all the desk type roles, the analyzing or keying in data and all that sort of stuff. Because that people-to-people-person-to-person contact, those things can't go away. And there's a whole raft of roles that need that person. You could argue that one day there'll be a robot that jumps out of an ambulance and says, where does it hurt? But that's probably a while away because you need that human contact. Yeah. It might be some really cool tools though that the Ambo has that can scan people and figure out what's going on. It's going on. Yeah. So it'll be interesting. Yeah. I mean, I love watching science fiction movies. Oh, science fiction. It's science-fuck that hasn't happened yet, isn't it? So yes. It's interesting to see how it plays out. I'm trying to think of the series on Netflix that plays with AI and this one in Black Mirror. Oh, yeah. Yeah. Yeah. There are things in that that you go, we're not that far away. Yeah. But it's interesting, isn't it? Because they're all human stories. Yes. They're all human stories that haven't yet, that have a tick element to them, but every single one of them is about a human story and people learning to be a little bit more human in a lot of cases. Yeah. It's interesting. Have enjoyed that, Jess. What a fantastic- I want you to come back with Bill. Oh, we can bring Bill. Yes. Playing with AI? Yes. Okay. It would be. It's over. We can. We should bring Bill in, maybe give him a challenge. Yes. We'll do that. We can get his own back on on Mark and his Tibetan moment. That was good fun. That was great fun. Thank you so much for joining us. Thank you to all of you for joining us today on this episode of The Dumb Monkey Show. And we hope that you will join us again next time. Thank you for listening.