A Job Done Well - Making Work Better
Welcome to "A Job Done Well", the podcast that makes work better.
Each week, Jimmy and James will bring you an entertaining and informative show that will transform how you work. Their backgrounds – everything from running a multi-million-pound business to packing frozen peas – have given them a rich assortment of flops (and the occasional success) to learn from.
Whether you are the leader of your own business, manage an operations team, or just want to do your job better and enjoy it more, this podcast is essential listening. It provides insights, advice, analysis and humour to improve your performance and enjoyment at work.
The podcast is guaranteed to make your commute to work fly and may also help if you suffer from insomnia.
Contact us and let us know what you think.
A Job Done Well - Making Work Better
How to Make A.I. Work for You
Season 3 - we're back with a muddle of James's!
To kick the season off, hosts James Lawther and Jimmy Barber explore how we can all use artificial intelligence at work with special guest James Crawford. Unless you're living under a rock, you'll be aware of the promise of AI revolutionising the workplace. This tends to focus on how organisations, governments, or public services will utilise AI and automate various jobs.
Today, we explore how we as individuals can use AI to help us get ahead and revolutionise how we work. AI expert James Crawford shares his 5P's model and also explains which AI systems work best for specific jobs. He then discusses some of the uses and best practices he has experienced during his travels.
Plus, you'll hear 'what we did last summer', but somewhat less dramatic than the films or songs!
So if you're worried about getting left behind in the AI revolution, check this episode out!
Hello, I'm James. Hi, I'm Jimmy and welcome to a Job Done Well, the podcast that helps you improve your performance enjoyment at work.
James Crawford:So we're back with season
Jimmy Barber:three. We are James, where we have got lots of interesting guests,
James Lawther:excellent subjects, all focused on helping you improve your performance and enjoyment at work. Ah,
James Crawford:Ah,
Jimmy Barber:very good. Which I suppose, Beth,
James Crawford:to the
Jimmy Barber:question, what have you been up to since season two? So I've been enjoying the, the summer, the sun has been amazing and yeah, been a bit in the garden, a bit out in the sun, you know, just generally enjoying it. I, I haven't been on holiday,'cause now my kids are a bit older. It's a September holidays for us when the prices are a bit cheaper. So Santorini, here I come.
James Lawther:I have been inter railing. It's the first time I have been inter railing since I was 18. And I can tell you a inter railing is much more fun when you've got some money behind you.'cause you don't need to eat nearly so much spam. But B three weeks with two teenage daughters is a little bit more than any one man could cope with.
Jimmy Barber:I find a weak is the sweet spot for me with with my kids now. I think it's it. I think the feeling is mutual as well.
James Lawther:so what are we talking about today?
Jimmy Barber:So today, James, we are talking about AI and work and how you get the best out of it.
James Lawther:All right, but we've done AI before a wee bit, haven't we?
Jimmy Barber:We have, but this is a little bit more tangible'cause we have an expert who will get to introduce himself in a second on the actual practical uses of AI as opposed to
James Lawther:All the strategic gumpf.
Jimmy Barber:strategic governance. What did, what do we call it, management fluff. Is it
James Lawther:Yeah.
Jimmy Barber:And then are the robots going to take over the world?,
James Lawther:So without further ado then I shall introduce our guests still. We've got a chat called James Crawford with us. James, I think you'd probably'd be better off introducing yourself'cause you will do a far better job of it than I will.
James Crawford:So I'm a career business transformation leader, and basically what I've done is set up services, taken down services, worked with organizations in a whole range of different programs in different sectors, both public sector, private sector, and charities. But I'm particularly interested in AI at the moment. Uh, because it's suddenly become a very useful tool. In all of the things that I've done in the past, I kind of held off with for quite some time. I did look at AI in 2018. I was doing strategy work and continuous improvements and all sorts of organization development work at the PO at that point in time. But AI seemed very. Mysterious, very technical, difficult to use. If you had an office with no windows, then you're probably the type of person that would've really got into it. But for me, in my work out in the front office, in customer service, in finance, in operations engineering, it didn't really seem to be particularly useful. So I cracked on all the way through to 2023 and at that point I had a chat GPT moment, and then suddenly I could see how it would be useful in all sorts of different areas. A practical tool that people in my kind of role. Could really find useful working with the stakeholders, you know, inside and outside a company. So that's when I really got interested and started pivoting all of my work towards ai.'cause it suddenly became something really useful that I felt that all organizations need to take note of.
James Lawther:So I think what will be helpful then, we've obviously both use ai, but maybe if we started off telling you how we use it and then you give us a mark, are we in the right general parish or not? Or are we missing something? So, you know, a bit of a self audit. Let's
James Crawford:Okay, far away.
James Lawther:So I, we use it, we use it a lot actually in preparing episodes. So if we've got a specific subject that we want to talk about, we use it to flesh out, the, flesh out, I should say, not flesh out and be very careful. Where I go with that is to flesh out the, the topic. And then the other thing I'm using it for actually is I'm ask, you know, I've written one book. I'm in the process of writing a second book, looking at failures and what exactly happened. So I'm using a lot for researching things. So I suppose that's where I'm with it. Jimmy, we what's, where are you mate?
Jimmy Barber:I guess aside from how we use it for the podcast, there's a few things for me similar to you, James, content creation and I, but I do find it's quite useful as a sounding board as well, so I think you can. You know, you can ask it. This is what I'm thinking. Have you got any views on that? And I find it gives you sometimes some challenging perspectives on how you're thinking about things, and it can give you sort of stats and information to back up or contradict your views. So idea development, content creation. A couple. I've got one great story just to share with you, James, where I, I used it recently. I changed one of the batteries in my remote controls for the car fobs. I took it down to Timpson's. They charged me 25 quid to do it. So when the second one went, I thought I'm saving myself 25 quid. I've got this two quid battery, I can do it myself. So I put it in nothing, absolutely nothing. Fiddled about, did all sorts of things, nothing. So in the end I had to ring, ring up the dealer and say, look, I've got this problem. What do you recommend? They said, bring it in. We'll do a diagnostic on it for 264 pounds before they, not to fix it, just to have a look at it. But then. I went on chat, GPT gave me a few things to try and one of them worked. And you know what it was? It was peel the sticker off the battery and then put it in. See it. It worked. Simplest thing, saved me hundreds of pounds so you can judge from that story whether I'm an idiot for trying to save 25 pounds and then taking hours of my time, or if I've saved a fortune. I know, but that was how AI helped me in that example.
James Lawther:So go on. Then. How good are we then? Are we par for the call? Bearing in mind that we're a couple of old blokes, I think, you know, we're embracing the technology, but are we par for the course? Are we missing tricks? Where would you, what would you
James Crawford:I think that's a, that's a great place to start, to be honest. So Jimmy, you were talking about having a bit of an interaction with. Chat bot. So that's the key to it really. So you don't always frame your full question upfront'cause that's just not the way we're primed to talk to people, whether you do it by voice or whether you type it in. So getting something going, starting to think through. Where do you want this conversation to go? What information do you want? What would be helpful? Ask clarifying questions. It. Adopting that kind of natural style with a chatbot is really helpful, whichever mode you actually interact with it. So that's one of the key things which I do try and talk to people about, to treat it not as technology, but almost like, you know, a person, someone who has sat next to you, who knows what they're talking about. How would you frame the question? How would you go about getting the information out? How would you lead and manage the conversation to get what you want? And whether it's researching new episodes, whether it's trying to think up really tricky questions for your next guest, whether it's diving deep into, you know, a, a request for a proposal that you might have if you're going through a procurement exercise at the moment, or lessons learned. That you might want to glean when you're setting up a new project, you know, whatever it is. There are good practical techniques just to be in natural language working through that particular topic. So I'd say that was a great place to start and you can
Jimmy Barber:James, that that's, that's quite different, isn't it, from how we think about search engines where it's just put in a thing and you get back stuff and you, you, you know, it doesn't iterate. Whereas what I've found is one of the big benefits is the more it learns about me and how I do things and how I think about things, or how I talk or how I write, and the, the more it learns about the projects I'm doing, the better it can do.
James Crawford:If an expert came and sat next to you, you wouldn't start with a six minute spiel, tell'em everything, and then fold your arms and say, right now you can respond. You'd start with sort of, you know, general positioning. They would then say, okay, well, you know, here are some things. You might need to look for. Then you'd start talking about your organizational context, and maybe they provide a bit more depth in some of those areas. Then you dive into some of the particular areas that you need to probe in an interview or the kind of skills that people in different organizations have. And it's that natural development of a conversation, which are people have found really quite striking with chatbots and, and very practical.
James Lawther:Well before, I mean before we go into actually how to use the ai, I mean, your point about an expert is really quite interesting. So I am a big fan of Gemini g. Jimmy uses Chat, GPTI. There are a whole host of different AI that you could use. Are they all the same or are there better ones for better jobs? I mean, how would you choose which AI to use?
James Crawford:I think a lot of them are broadly similar in terms of the knowledge, the breadth of the knowledge on the internet that they were trained on. So they're all containing a lot of similar material, but they do present it in different ways and they are fine tuned for different groups of people. So it's partly, which is the best chat bot for a particular job. And they've all got strengths and weaknesses, and it's partly how you interact with it as well. So if you use something for, you know, called pie ai, personal intelligence, ai, that's really good for kind of, you know, short, sharp conversations about, you know, household insurance or fixing the windscreen wipers on your car, or when, when should I mow my lawn or. Troubleshooting things in the house. It's much more kind of short and snappy, but Gemini is kind of more general and it's got good links out to the internet. If you want to kind of do some research in terms of whether it is relationship revis or you know, the the new car that you want to buy, then perplexity AI is very good'cause that's very much fine tuned for searching out information and bringing it together in a really insightful way together with all the citations that you could use. Each of them has got kind of, you know, different strengths and weaknesses. And of course, I've got to mention the Mistral ai, which is the Europe's premier large language model that's based in France. It's smaller than some of the others in terms of its business, but I found it very quick. A lot of people prefer it to chat GPT or others in terms of its style and its language.
Jimmy Barber:It would be worth us putting some of these links in the show notes as well so people can find them easily. But one thing I did notice, and maybe this isn't a fair comparison, but I noticed that Meta have got an AI in built into Facebook now and it it prompts you with some questions you could ask about any particular post and I found that to be particularly inaccurate. It's not the best example, but I'm a, I'm a Nottingham Forest fan, and when there's a post on forest, it will say, do you wanna know more about these transfers or whatever, and it will tell you absolute nonsense when you click on it for more information about players that don't play for forest or stuff in the article, that doesn't exist. So I found that to be quite different, whereas using chat GPT or Gemini, I haven't gone far wrong as far as I've seen it in terms of the information I get back.
James Crawford:well, it's disappointing if the meta one is bad. They, they've particularly focused on open source software, so they've got this big model called lama. They've released it. So anybody can take that, take a copy of it, use it for free. Fine tune it for their own particular use. So that's particularly helpful when it comes to organizations taking it on board to provide private tailored advice for the employees or for their customers, and they're able to kind of take it on board and, and test it fully. It's disappointing that it's got so many hallucinations, particularly about nodding in forest.
Jimmy Barber:They're basic
James Lawther:It's, it's not. I live in Nottingham. There are plenty of people with hallucinations
Jimmy Barber:it knows, it knows about as much as Forrester as James does,
James Lawther:Yeah, not very
Jimmy Barber:be honest.
James Lawther:Yeah. But then there's an interesting point there though, isn't there, because it's a bit like having a conversation with a person and getting in information from a person. The question becomes, how much do you trust this person's judgments?
James Crawford:When it comes to Nottingham Forest, for instance, you'll find that a couple which have got live links out to the internet. Are gonna be more accurate because these large language models go through a huge training process, which takes many, many months. And then they've got a testing and a release process, which takes many more months. So by the time they actually hit the streets, they could be a year out of date in terms of their
Jimmy Barber:That would make sense. Yeah.
James Crawford:And so that, you know, might be a feature of the LAMA model, which Meta uses. Whereas what Google does and Perplexity does is it actually provides an an interface out to the live internet as well.
Speaker 10:As well as recording this podcast, we help individuals, teams, and entire organizations achieve outstanding results.
Speaker 11:If you'd like to find out more about how we can help you, please get in touch email, either jimmy@ajodowell.com or James at ajo dunwell com.
James Lawther:So I've got a model, James, that you suggest people use to get the best outputs out. Yeah. Which you call the five P. So could you run through that and explain that to us a wee bit
James Crawford:Yeah, sure. So the question is how do you get the most out of this, particularly in a business context? And what I did is I put together what I call five Ps, which basically is work thinking a bit like a business analyst. And one of the things you need to do is to direct the chat bot's attention to the area of the internet, which is going to be most useful for. So for instance, if it's ingested everything from Reddit all the way through to Wikipedia, through to Google, through to every published website, every academic website, you've got all sorts of different niches on the internet. And what you need to do is to try and prompt it. So the kind of output it comes up with is relevant to what you're actually asking. So. I was saying that you, first of all, pick a persona. So if you were speaking to an expert, who would it be? You know, is it a university pro professor who's, you know, lead a leader in materials management? Is it someone who's understood the navigation challenges in London and knows all of the streets? Is it someone who is specialist in HR and a leading thinker on better recruitment and how to make better hires? Who is it that you would really want to direct this question to? So if you just describe that in a sentence or two, that will then immediately direct the attention of the chat bot to that kind of area of the internet, and you'll have a more productive sea of information coming back to
Jimmy Barber:If, for example, James, I was wanting to think about how I progress my career, I could turn around to it and say, act as a expert career coach and, and that would get me a better answer than just if I just said, how do I progress my career?
James Crawford:Yes.'cause you're kind of giving it a cue in terms of the kind of conversation you want, the kind of sources of information you'd like it to mind. And when it goes through its probabilistic building up of sentences. It'll give more weight to that kind of output that's gonna be more useful to
Jimmy Barber:Yeah.
James Crawford:So that's, that's my first one. Kind of pick a persona. You don't need to labor it, you know, you could just, you know, do it in a few words or a couple of sentences. Give it a bit of context and you'll find that that should help. The second one is to give it even more context information. So, for instance, if you're looking at your career, you could say, right, I'm currently a software engineer or a taxi driver, or an innovation manager in a manufacturing firm. I've, this is my background. But what I'd like to do is develop these particular skills because I'd like to. Move to France or have 50% onto my salary or break outta the rut I'm in'cause I'm finding it a bit stale. That's all again, good contextual information. You might also upload your cv, press the plus button and upload your cv so it's got that background there. Your conversation may have been triggered by an article that you've read recently about a, a new career or a new type of skills that are coming through that particularly got you thinking You could try uploading that as well and say, you know, this kind of got me thinking, but it's. You know, just one example of something which I could move into. So providing that context. The more you make it your own conversation, the more you make it a completely unique conversation, the more relevant, accurate, and insightful the information that will come back to you. So those are the first two ones that kind of pick a persona, prime it with the context contextual information that you're after. And then you can prompt it, which is what's your ask of it? You know, are you asking it to think of 10 careers that you could move into? Are you asking it for new? Ideas that are particularly up and coming in your particular part of the world. Do you want it to create a learning plan? Do you want it to tell you where you can go for careers advice? You know, what are you actually asking it? So be clear about that.
James Lawther:Sorry, question about prompts. So one of the things I didn't say to start off with is we actually use midjourney, which is a, a graphics package. So it's AI graphics, it will draw stuff for you. But what is really quite interesting is you can ask it to do the same thing and you'll get different responses each time. So similar-ish pictures, but they're not the same, they're far from it. Presumably the same is true with a, a language. You ask it the same thing and it will give you different answers each time. So it's, it's not providing you with the truth if the truth exists. It's providing you with what it thinks is the best answer for you at that moment in time.
James Crawford:That's correct, and obviously the large language model providers take a lot of pain to try and be as accurate as possible. So you know, they will go to relevant and authentic sources of information rather than try and read off Reddit all the time. So they will. Behind the scenes try and direct its attention to the kind of accuracy of information that you want. But yes, because it's got a bit of a randomizer in there, that's partly what makes it a very natural interface and quite conversational. But it also steers it off in different directions each time. And sometimes they're a bit more profitable than others.
James Lawther:Yeah, but it's only as good as the information as it is read
James Crawford:Well, yes.
James Lawther:Yeah. Well, it can't be any better than
James Crawford:It's a yes, it's a combination because it's read so much information from so many different sources, some of which contradicts itself. Um, you know, if you ask it two plus two. It may give you the answer for, but if it's already had a bit of a chat with you and you are talking about people making mistakes, you are talking about maybe assumptions going wrong. If that's the theme of the chat, you may say, well, it's two plus two and it may say five.'cause it's also read that thousands of times. So it'll try and detect the nuance of the chat that you're having and then bring the relevant information to you. So the answer will be different, partly because of the context of the chat and partly because of that randomizing element. Which makes it interesting to speak to and which does make it different for you each time you do it, as well as different for everyone else.
Jimmy Barber:Back to your five Ps model.
James Crawford:So we've done the first four, which is pick your persona, the kind of expert you want. Prime it with the background information and the context you can. Give it a prompt, which is exactly what you want it to do. The fourth one is telling you how, how it should be presented. Are you after paragraphs of information, are you after a table, bullet points, a PowerPoint slide, whatever it might be. Be prescriptive in terms of how you'd like it to. You could say, you know, please give me the high level view at first, and I'll then tell you which bits to dive into in more detail. You could say, give me a six page report. So that's how it presented. And then the fifth and final one is perfect it, which is all about that iteration at the end. So once it's come up with its output for you, then you can give it feedback, tell it what you want more of, what you want less of, tell it to do it again. Start from scratch, whatever it might be. But you perfect the information that you want. And so you know, after a few minutes, you've got what you came from.
Jimmy Barber:It is a really nice, simple model. Five Ps is quite easy to understand and you can see why logically it makes sense and you get you better answers. So yeah, I really like that.
James Lawther:Yeah, because it's actually, I suppose, James's a sort of continuous improvement way of getting an answer, isn't it? So instead of, I need to write the best prompt and it's one shot and done, it really is all, look what I've got. What do I think of that? How am I going to make it better and gradually refining it?
James Crawford:And if you can pause and think at the beginning of a conversation, you generally have a better conversation, whatever the setting is.
James Lawther:that, that makes a lot of sense. that's how to use it. But if you got any examples you can give us of where this has been done. So if I was, oh, I dunno, I make something, if I was a transformation manager in a, in an organization, how would, what sort of problems would I be looking at? How would I use it?
James Crawford:Well, I'll tell you about the first example I used it for and when I first came across it, I was working in the housing association and what we're doing is setting up a, a program to refresh systems, clean a lot of data, completely reset and refocus our processes, and then put them onto a new. Platform. So one of the things that struck me is that chat, GPT, which is the one I was using at the time, has read every lessons learned report of organizations that are trying to. Platform their data trying to update and modernize their organization, both good and bad. It's looked at all the conferences where case studies were presented. It's looked at all of the, you know, government analysis where things went wrong and they were made public. And so I start off by saying which of the sections of a project document. Can it actually start to complete? And of course the answer was all of them. What it did need though, of course, was a lot of information from me about, okay, what are our objectives? What are we trying to achieve? You know, what brainstorms have we done within the meetings? What emails are shaping how we want this project to kick off? So the more information I could feed up to it, the more of a starting point it had. Then, of course. You could, you know, upload your template for a project document, whatever it might be in your organization, and it'll then take all the information it finds relevant on the internet, brings together the information that you've provided it, and come up with a first draft of a project document. But that's the basic model that you can use, and it'll then do a lot of the work for you in just a few minutes. That would've typically taken you several hours over a period of several days. And suddenly you've got a first draft which you can start to work on, which you can share with people, and which contains the essence of what you're trying to achieve, as well as the wisdom of the wider industry.
James Lawther:So it really is using it as a starter for 10 almost, and or it doesn't really matter what your business problem is, whether I know you're an HR person, you need to develop a disciplinary procedure, or if you're a, you doing an outsourcing contract or it really doesn't matter. It's just about being clear what you want it to do for.
James Crawford:And that goes back to one of your early points about, you know, will I ai nick our jobs and the subject of one of your other podcasts. But where it's at at the moment is it's a very, very powerful tool for people to accelerate the research, the thinking, the brainstorming, the auctioneering, the drafting, the reviewing. These kind of documents, for instance, or business processes or the design of services for customers, or the way that different teams will work together where there's maybe been a historical problem, whatever it might be, and because it hoovers up the wisdom from many places and it builds on the information that you provided as well, then you've got a great starting point for discussion. And if it does make some mistakes or point in the wrong direction to some information that wasn't quite correct or isn't quite what you want. That's fine'cause you're just in the drafting stage and that'll all get filtered out with the conversations you have. But you'll be able to start with those conversations, you know, in half an hour's time instead of fix a meeting for two weeks and hope that you've done the preparation in time so it can just accelerate the way that things are done.
Jimmy Barber:Just to pick up on one point you made there. So as an example, if I'm running a team or a unit of people delivering customer service. There are quite a lot of uses in what you said that I could do, depending on what my problems are. You could put in your problems, you know, whether that's team members aren't behaving right, or you're having trouble with your bosses, or you need to update your process maps, or you need to find ways of improving how you adhere to certain regulations. Whatever it is. It could give you some ideas. It might not give you the perfect answer, but as a starting point, it's probably a, a useful starting point, isn't it? For many, many of the day-to-day problems that you have as a, as a kind of manager.
James Crawford:It is, and again, that iteration comes into it as well. So if you're asking for, you know, process improvements. You can say, please give me, you know, five ways I can improve this particular process. You may upload it, you may have a diagram on your screen. You can take a picture of it and then simply pass that to the chat box, provide it some context, and then say, you know, where would I look to improve this process?
Jimmy Barber:Yeah.
James Crawford:got some information about where it's falling short at the moment, it'll then probably come up with some decent ideas, but probably ones you've thought of already. So you say, okay, please give me another five, which are quite different to those. Which maybe other people wouldn't have thought about or tried before, and it goes round again and it gives you five more. And there might be something a bit more interest there. You can then say, all right, give me some, another five really wacky ones this time really off the wall ideas, but they're gonna be really powerful and have a great experience for the customer of this particular process. And you actually find, as you go through that process, it'll actually come up with something which you haven't thought of before that does actually either get you thinking or you think, oh yeah, I, I could build on that. That could be really useful. So it's a great brainstorming tool. Um, and you'll find that again, low risk is, you know, if it makes a mistake and comes up with an idea that's just not practical at all. It doesn't matter'cause you're going through that particular stage. So that's the, that's where I say people should start. Just get familiar with interacting with the chatbots. You don't need to go away. Write out a decent project, project brief, put a business case together for all the time you're gonna spend on it. That's not where to start with using these tools. It's very much about familiarity, incorporating it into the way that you work, using it for inspiration, and then the business case will come later. But if you've got that relationship with it, you've understood what the kind of thing it can do and the kind of thing it's not so good at, you'll find that it just becomes more and more invaluable to you. And eventually that'll deliver. The real process improvement or the real service redesign that really is gonna be a benefit to people.
Speaker 8:As well as recording this podcast, we help individuals, teams, and entire organizations achieve outstanding results.
Speaker 9:If you'd like to find out more about how we can help you, please get in touch email, either jimmy@ajodowell.com or James at ajo dunwell com.
James Lawther:So those are all the things it can do to help us, but what should we be careful about, James, and what would our watch outs be, do you think?
James Crawford:So business confidentiality, chatbots, as I say, go through a long learning process where they crunch all of the data that's made available to it. Now, of course, part of the data that's available is the stuff that you've uploaded. So if you've got documents at work or personal documents about healthcare or insurance or anything else, if you upload those to a chat bot, it'll treat that as fair game. It'll have it in its memory banks, and then when it comes to a future learning program ready for the next model, it'll use that as. An inspiration for how does it, how do things work? What kind of questions do people come up with? What kind of material can I draw on? It won't save it away in a database and then send your insurance document to someone who asks about it in eight months time. But what it will do is draw inspiration from the way that document's put together, and you may find that some of the issues concerned do feature in a future model if someone else is asking a question about something broadly similar.
James Lawther:Well, and that's interesting. Sorry to cut across, but I, I was asking, as I said earlier on, I was writing a second book and I was asking about the post office and I had written a blog post about the post office a while back, and it actually, one of the sources, it referred back to me. Was what I had written. So yeah, the minute it's out there, all the AI has got it. It's becomes public knowledge.
James Crawford:are ways of making sure that the information is private, which is particularly important in a business context for personally identifiable information, company, finances, you know, policies, pricing mechanisms, the kind of stuff which is really useful to put to a chat bot. You don't want that going into the wider learning process, so you can either press the temporary chat button. Which means normally that that document or that chat isn't used in future. But the flip side of that, it's not. It's also not held in your history, so you can't go and restart the conversation at a future point, or for certain levels if you use the paid versions. Some of those incorporate privacy as well, but you need to look at the terms and conditions. So it's worth checking, either anonymize it, go for a temporary chat, make sure you've got privacy, or somehow make sure that your data is protected where it needs to be. A lot of people put their lives out on social media, you know where they're going, you know, where they, when they're traveling with their family, well, you know, what they're thinking, what they're doing, all sorts of information out there, which kind of makes you think, you know, even on platforms like LinkedIn, which is mainly for a business audience. So I think the f the key thing is to be aware how is your data potentially gonna be used? And then you have to ask yourself the question, and do I mind. And because we've got AI bots, which are becoming quite intelligent and not only used by business people, but also by scammers, I think it's definitely worth looking into how are scammers starting to use the information that's out there? How can I protect myself against some of these scams, but also do I want to continue feeding some of this information out into the internet? How do I become more private? Well, how do I manage the risks? And it's gonna become more and more of a topic that we all need to look at over the coming months and years.
James Lawther:Yeah. Okay. It's a little bit like the old adage, you know, if you're going to go on holiday, don't use your local taxi firm because they therefore know that you've all gone away. Now, is that paranoia or is that very sensible? Thing to do. I don't you? I suppose you're right. It depends on how you feel about these things.
James Crawford:Yes, and
Jimmy Barber:I look, I look forward James, to the next time you're trying to get a Nottingham taxi to come pick you up.
James Lawther:Yeah, absolutely. Yeah. What about the future then, James?
James Crawford:So at the moment, the name of the game for many organizations is what we call augmentation, which is how can these tools help? Internal people do a better job in terms of what they're doing, and it may be that someone is in the contact center, they are talking to customers, but there could be a bot, which is giving them feedback on the way the call's going at the moment. Special offers relevant information from the customer's contact. Did they report this same problem six months ago? Is there a trend? Things which would be very difficult for an agent to actually look at. And when they're trying to concentrate on the speaking to the customer, and also look at the customer relationship management database, the AI could be working behind the scenes to augment that particular process. So that's a, a low risk, but very, very valuable way of progressing.'cause it helps the agent to, to do that. Once the technology is mature enough and once the organization's mature enough, then you come to the automation stage where you can actually let the org, the AI lead on the conversation. It will do the searching behind the scenes. It will understand what the customer wants. Provide feedback, reassurance, answer questions, give the information that's needed. But that level of automation for many organizations isn't quite there yet in terms of the quality, the reliability, and the confidence they can have before going out to, you know, an external audience. But that's the direction to travel. And the best systems at the moment can do that, and they can do it very well, but it's up to each organization to move along that journey. Think through what does this mean? Not just in terms of how do I automate the service, but the really powerful stuff beyond that is, and how do I completely change the service? So how do you flip the service on its head to provide a very different experience for the people that are paying the organization's bills? So again, there are several steps of that journey which will unfold over time. There are people involved in thinking that through guiding the process, setting up the systems, providing the expertise, reassuring the customer, providing that human to human contact. So there's lots and lots of work for us all to do for, I think quite some time before the alleged machines take all our jobs, commandeers of spaceship and leave planet Earth.
James Lawther:But the conversation really flows at two levels then, doesn't it? So is the A, how can I as an individual use it? And you've given us a lot of pointers on that, particularly the five P model. But then there really is a much more. Strategic conversation, which is how are you going to use this in your organization? But really you would get what you ask the AI to do, which is your point about Amazon versus the Telco. Yeah. Which really comes down to the company values.
Jimmy Barber:On the form a bit on the, how do I use it? I think also it's been really helpful to sort of think about the breadth of it, how I as an individual can use it rather than how we use it strategically and how both solving problems, interactions, and how you build those interactions. I think hopefully that's given the audience some real food for thought for how they personally can get the most out of AI as well.
James Lawther:And so James, if people want to contact you or want Yeah. Specifically the service that you guys offer, where should they look? Where can they find out more about you?
James Crawford:Well, I think it's really important to bring the people together and make sure that we address people's. Hopes, fears, expectations, concerns, questions, everybody comes. Having looked to listen to the BBC News, watched at the website, had all sorts of conversations with people about what this means for the future of their jobs and their organization, but once we've been through the process of change, processes, changing services. Ultimately it, it changes organization's business strategy. So that kind of challenge is actually gonna be apparent for all of our organizations. But the place to start when I first sit down with people, it's taking'em through a, a, a process I call the three is, which is inform, imagine, initiate, which leads onto benefit in money, which leads on to benefit in quality of service. So I find that's a, a really good positive start for people's AI journey.
James Lawther:Very good. And if people actually, I mean people want to talk to you, James. What's their best way of getting in contact?
James Crawford:So we're either James Crawford on LinkedIn or through the website touchpoint change.co uk. So my contact details are all there. I'm always happy to talk with people about what their next steps are. I'm always curious to find out where people are up to and you know, if they haven't really embarked on the journey so far, you know, what's holding them up? What are their fears and concerns? What do they see other people doing? So starting off with just, you know, a really interesting, informative conversation that then hopefully will lead to a relationship when we can actually work together on their services.
Jimmy Barber:We can put those links in the, in the show notes as well.
James Lawther:absolutely we will. Very informative. Thank you very much for your time, James. That was really very interesting. Oh, I've got, sorry, one other question. So will AI make Jimmy thicker? That's what I really need to know.
Jimmy Barber:Is that possible? James?
James Crawford:Wow.
Jimmy Barber:it make, will it make James clever?
James Lawther:Will that bit get edited out the podcast? That's the real question.
James Crawford:And to be honest, it has the chance to do both. It depends how you use it. So if bets bets have been placed,
James Lawther:bets are off.
James Crawford:how it's gonna
Jimmy Barber:whole nother episode.
James Lawther:Alright, lovely. That was really helpful.
Jimmy Barber:Thanks everyone. Cheers now.
Speaker 4:As well as recording this podcast, we help individuals, teams, and entire organizations achieve outstanding results.
Speaker 5:If you'd like to find out more about how we can help you, please get in touch email, either jimmy@ajodowell.com or James at ajo dunwell com.
