John Munsell: How EOS and AI Can Give Your Team 10 Hours Back Every Week

SUMMARY KEYWORDS 

AI impact, upskilling employees, AI strategy, AI governance, autonomous agents, AI mastery, AI training, AI efficiency, AI security, AI implementation, AI capacity, AI tools, AI frameworks, AI challenges, AI benefits. 

SPEAKERS 

John Munsell, Debra Chantry-Taylor 

John Munsell  00:00 

Gams can be so sophisticated now with AI. It'd be very easy for somebody to clone your face, clone your voice, and clone a request to wire money somewhere, or to do something. If you're a business owner, one of the things you want to know is what is the actual impact of AI on your organisation if you upskill your employees. We teach them how to reduce their workload by at least 10 to 12 hours every week as an outgrowth of going through the training that increases capacity in an organisation, especially when you multiply it by 50 or 100 or 5000 employees. 

Debra Chantry-Taylor  00:40 

Welcome to another episode of Better Business, Better Life. I'm your host, Debra Chantrey Taylor, and I'm passionate about helping entrepreneurs lead a better life by creating a better business. I work with established business owners and leadership teams to help them live their ideal lives using EOS, the entrepreneurial operating system, and I bring guests onto the show to share their stories about EOS, but also tips and tools that can help you to create that better business. Now, today's guest is the author of Ingrain AI Strategy Through Execution and Scalable Prompting. He's also the creator of the AI Strategy Canvas, and he is probably one of the few people who truly knows how to integrate AI across every department of your business. Today, he's going to share with you the impact of AI on your organisation, how to truly integrate AI across the entire business, what to do with the efficiencies that AI helps to create, and also the pitfalls that you need to avoid. John Munsell is a CEO of Bizukka and author of Ingrain AI. Please welcome John to the show. Welcome back to the show, John. So, we had some really great feedback from your last podcast with me, which is why we've got you back on board again today, and I'm really looking forward to chatting a bit deeper about AI. So, yeah, welcome back. 

John Munsell  02:05 

Thanks, great to be back. Appreciate you having me on. 

Debra Chantry-Taylor  02:08 

Absolutely, so AI, I mean, I know it's such a buzzword at the moment, but there are so many levels of AI, and I think since our last podcast, I got your book, I started doing some research. I started going down rabbit holes. I got challenged online by somebody about how EOS was going to be dead because of AI and agentic AI. And I've got lots of questions that I want to ask you. So, but before we do that, for people who may not have listened to the last episode, would you mind just giving a little bit of your background and how you got into where you are today? 

John Munsell  02:42 

Sure. So I started off in financial services, spent 16 years there, and then when the web came out, I ejected to start a software development company back in 97 building web applications for people, and then we gradually grew into a digital marketing agency and a web design agency, and all this stuff, so I've been basically building software applications and in them and the digital space for 20 plus years, you know. And then about four years ago I sold off the agency side of the business because I saw the new frontier, which was AI, and so I spent a year studying the markets, trying to figure out what problem the market wanted solved, and then I realised the problem it maybe says solve was to figure out exactly what this thing was, how to use it, and how to scale it in their operations, and so that's what we have built over the last four years, a company that is part consulting company and part training company, but we're also building out a network of certified implementers who implement our frameworks inside of businesses literally across the globe. 

Debra Chantry-Taylor  03:55 

Wow, and so four years ago, and how has that business or grown in that four years? 

John Munsell  04:01 

It's been a whirlwind, because everything changes so fast. It used to be that a new model would come out every six to eight months, and then it became every two or three months, and now it seems like it's every six days there's a new model coming out, and it's. it's one of these things where it's. it's exciting, but at the same time you know that people are kind of, they're racing too far, too fast, and that is where it gets scary, because I don't know that people are fully, fully aware of where this could go, and it could go really, really well, but it could go really, really wrong. So we want to try to help people put controls in place and put structure in place, and be prepared, and hopefully keep as many of their employees employed as possible by giving them skills to where they can. Actually, add even more value to the organisation, and the organisation just needs to know how to handle that, you know. 

Debra Chantry-Taylor  05:08 

I think that has been one of the biggest sort of concerns for a lot of people, as they sort of think that AI is going to take their job, but it's not just changes the way that they do their role and the value that they add to the organisation. But we were talking offline before we came on, and we were just talking about the, you know, people get excited by AI, they may have downloaded a chat personally, had a bit of a play with it, seen the possibilities of it, got hold of some prompts from somebody online, and tried it out, and created some amazing stuff, and then all of a sudden they kind of go, "Right, I now know AI, and they go into it full steam ahead without thinking about the consequences, and you use a really beautiful analogy, which I'd love you to share now, if you wouldn't mind. 

John Munsell  05:47 

Yeah, well, when you talk about agentic AI, especially autonomous agents, so we have this thing called the 10 levels of AI mastery, and you know, level one is just really having a conversation with AI, asking your question, and going.. wow, this is amazing. Level two is you're asking better questions, and you realise, well, the way I phrase it is different. When you get to level nine, you're building autonomous agents, which means the AI is actually making decisions and taking actions on its own, as opposed to the earlier levels, where you're telling it what to do, and you've structured what it. what it does, but on the autonomy side, it's making decisions. If it doesn't have an asset or a resource, it will download it. If it needs to go find it, it will retrieve it and insert it into your systems. It will knock down doors that you want it to knock on first, you know. And so I kind of liken it, like, like having, if you, if you haven't progressed up those 10 levels of mastery before you get to level nine, and in fact, instead you skip from level two, which is what we see all the time, to wanting to play around with level nine, that's like walking into a room full of open gas cans with a cigarette, because you just learned how to smoke cigarettes, you know, and you don't want anybody doing that. You have to have far more controls in place, and you know, people don't realise they've got a lit cigarette in their hand, and they're literally playing around with open gas cans. It's, it's, it's not a good idea, but the gas cans, you know, power a lot of machinery, so if you know what it's for and you know how to use it, that's great. 

Debra Chantry-Taylor  07:25 

I love it, and I love the idea that there are those different levels. I'm curious, though, what is level 10? So 

John Munsell  07:30 

level 10 is actually managing people and autonomous agents, so you know if you think about an autonomous agent, that agent has a specific mission and a specific discipline. I mean, it's almost like an employee, but it's even more narrowly focused than an employee. It has a different set of skill sets and a different mission, and theoretically it's going to go on 24 hours a day, right? But you have to have people that manage those things, and you have to have somebody at the top that manages the people that are managing the agents and manages the agents as well, so there's a that level 10 fully understands the consequences of every action and every interaction. 

Debra Chantry-Taylor  08:17 

I think it's what I've been, I've been talking a little bit about online, and I'm not talking about it from the technology point of view, because I'm completely new at that, but it is really going to be where in your accountability chart, using the US tools, you will actually have seats that are owned by an agentic AI. They will have, you know, specific systems and processes that they are responsible for, like, so that they're running 24/7 They will need to have measurables to make sure they're producing the right outcomes. They will need to raise issues as issues come up, and have somebody actually, you know, working with them to resolve it. So, I see it as being you will actually have not just somebody at the leadership team level who looks after AI in the business, but you're going to have real agentic people that are, you know, being managed just like people are. And it sounds wonderful, but there's a lot of risks that come with that, because I remember reading Scary Smart many years ago, and, and looking at the way he described the AI back then was, it is like a, it's a baby that is going to absorb everything around it, and it has, it has all the knowledge, but it doesn't necessarily have the experience, and so, how it takes that knowledge and uses it can be very different to the way we might imagine that it would. So, yeah, it's good. It's fascinating for me. I've still got a lot of learning to do about it, but I just think there's probably lots of business owners out there who are similar to myself, who are going, they can see what the future looks like, but they have no idea how to get there, and I'm sure they must make some pretty big mistakes along the way, 

John Munsell  09:42 

yeah. What, that's what we don't want people to do, right? Is make a whole lot of mistakes, but you don't learn unless you actually make mistakes. But it's, it's, it's a whole lot easier if those mistakes aren't fatal ones, you know? I mean, like, I'll draw some parallels. We may have talked about this before, but 12 years ago we got an email in, or at least my bookkeeper did, masquerading as me. It wasn't me, but said, "Hey, I need your quick attention on this past due invoice. I totally forgot to send it to you. I need you to wire $19,000 quickly. Can you take care of that, and so anyway, wouldn't you know she did it. She wired $19,000 out, and then she got to thinking about it. She goes, maybe that wasn't John, and so she emailed my assistant. She goes, do you think this was John and my assistant? He goes, 'No, indeed, that was not that was not John. So, my bookkeeper literally threw up. She was just a mess. But we were fortunate enough because these bad guys, they did that en masse, and apparently they had scammed somebody moments before us. And what they do is they clean out the bank account and close it. Well, our wire bounced because they had closed it right before our wire hit. I mean, we were just blessed by timing, and so it took us about two weeks, but we got the money back. The point of all that is, that was that was an old school scam, right? Just an email. Imagine now, like I've been on a bunch of podcasts, it would be very easy, as have you. It'd be very easy for somebody to clone your face, clone your voice, and clone a request to wire money somewhere, or to do something. So, so jams can be so sophisticated now with AI. It's crazy. Now, somebody could build an autonomous agent to go out scour YouTube looking for people like you and me, figuring out more about them, cloning them, cloning their voice, and turning it into requests, just like you would a mass email outreach to cold customers. That is stuff that you have to prepare for. It will happen, and it will happen in mass, and it will be orchestrated possibly by a group of bad guys or possibly by a government, you know, so these are real possibilities, and if you don't have the governance set up to anticipate that, and you don't have any, we put people through a role-playing game where they see what will happen and what squirts out which sides when something rogue like that happens, and it's about an hour and a half game, and when you go through that, you start to realise, oh, there's a, there's a whole lot more we need to think through, so that's what what worries me about people jumping like I say, going into the room with the open gas cans with a cigarette in their hands. This is so much more complicated. So that's the dark side of it. The good side of it is there's so much that you can get done in seconds, and most people look at it as a way to improve productivity, right, to increase what we would call increased capacity by by freeing up people's times, because they, they have five hour tasks or 10 hour tasks they can reduce to 20 minutes. One of the CEOs that went through our training, and keep in mind, this is this is a CEO, this is not, you know, somebody on the front line, so to speak, but company, I don't know, they probably did 50 million in revenue or something like that, so not, not a huge company, but not a small company. He went through our training just to say, let me learn what this is for myself, and then we'll put all our people through it, so that we could increase the capacity. 

John Munsell  13:42 

Now we talk about increasing capacity when we teach somebody to go through our training, or when somebody goes through our training, we teach them how to reduce their workload by at least 10 to 12 hours every week as an out growth of going through the training that increases capacity in an organisation, especially when you multiply it by 50 or 100 or 5000 employees, you know, if you could get 100 employees saving 10 hours a week, that's a significant amount of hours, especially when you multiply it by 52 weeks, right? That's that's a monstrous amount of capacity. So he was anticipating this capacity because you have basically three choices with this excess excess capacity. You can either reduce the time that you have people, so you can either give them time off or get rid of them, or you could give them more time to work on things that add more value to the company, or you can sell into that capacity, right? You know, and that was what he said. Look, if I can create this kind of capacity, how am I going to sell into it? So he built on his own the most amazing tool, and because this is what we call a capstone project, so they have to build one of these things. What he did was they would get these RFPs in there, we. The office furniture space. Okay, they would get these RFPs in 350 pages for these giant office buildings, and these office buildings, they're looking for furniture, they're looking for fixtures, they're looking for plumbing, they're looking for lighting, all that stuff. He's got to find in that 350 pages what's what's desk chairs, the, you know, conference room tables, right? Then he has to look at the specs of them, determine exactly whether or not that fits with what he sells, because he sells high-end stuff, and he doesn't want, you know, stuff that you can, you know, put together with a couple of bolts and screws, so he has to figure all that out, and he said it would take him somewhere between three and six hours just to determine whether or not they wanted to reply to the RFP, that's what he calls a go-no-go decision. If they said it was a go, it would take him somewhere about two to three weeks with two and a half people to assemble the reply to the RFP, think about that kind of bandwidth, and how much a resource hog that was, and so he said, you know, we would only reply to two or three of these per year, because it was such a distraction, and then we only had a 5050 shot at getting the business, but we would pour our heart and soul into it anyway, he, he built the tool to get to the go, no go decision within 20 minutes, as opposed to three to six hours, and then he was able to build something that would read the RFP, purge it, sort it, and then pull in all the resources through various semi-autonomous agents and assemble the reply and response to the RFP in two hours with one person, as opposed to three weeks with two and a half people, and he said, "John, this will actually make us millions of dollars, that each RFP is about a quarter of a million to a million and a half in potential revenue, and he said, "We can now respond to three to five per month, as opposed to three per year. That is amazing, that a CEO would do that. Right, he didn't normally - they would delegate that kind of stuff to somebody else, but he figured out how to do it on his own. And those are the kinds of things that people don't necessarily think about. They think about efficiency first, but then he found a way to sell into some serious capacity, so it's, it's, it's monstrous what you can get accomplished with this. That's 

Debra Chantry-Taylor  17:30 

game, game changing, isn't it? For the business and for everybody involved. Yeah, especially the go, the no go, because you know, as you said, it took such a lot of time to even get to that before you pulled together the rest of the thing. Love it. Okay, so that's a great example of where it really truly can change the way that the business is run. You work with a lot of businesses, various sizes and whatnot. What have you seen the biggest mistakes that they make when they start thinking about AI? 

John Munsell  17:58 

Oh man, and you know what, I would, I would have to say, we've identified a bunch, you know, like 20 or so. One of the big ones that we see is that people will, they get really attracted by what we would call the one sexy AI application, you know, and so they go like, oh, we could build an AI application that does this, this, this, and this, right, and so that takes an enormous amount of time and internal resources, but it also takes an enormous amount of collective expertise to drive that, so that would be what we would call like a level seven or level eight application, and when we test the organisation, especially the executive team, we find that the entire organisation is operating at level three or below, and they, they have an appetite, they, they're looking at this giant elephant they want to buy, and it's a level nine thing, but there is no collective expertise to actually do that, so they bring in an outside vendor to build that application, the vendor has expertise in building the automation, but not in the internal expertise needed in the organisation to know what's missing or to know what's necessary. So the big mistake is that they invest in that application first rather than investing in the capability of the organisation as a whole. In other words, when we look at it at a natural basis, what typically happens is an organisation will buy people access to certain tools, like, you know, Chat GPT, Claude, whatever, but they'll, they'll have them trying to figure it out on their own, it will take, or they'll just give them some cursory introductory training and say we check that box. It will still take an individual because they have a real job and they're busy all day, it's hard for them to try to figure out a new toy when they got responsibilities. It'll take. Person somewhere between 19 and 24 months to get to level six efficiency or proficiency, but if you put them through training, you can get them there instead of, you know, two years, you can get them there somewhere between, say, two to eight weeks. Now that's that's a 22 month head start now. If you can get everybody trained up, let's say in six months, now you got this collective expertise in the organisation hovering around level five or six, which then you can bring in these level seven and eight applications, and not only that, but you think of more, right? All these people can now collaborate as a team, and they can share what they're, they're learning and doing, and the way these applications take place. At that point, the success rate is way, way higher, but also they'll look at the original idea that they were like, "Oh, this would be cool, and they realise, oh, that is not nearly as cool as these other four things we can build now that we have the collective expertise. Does that make sense? So, that's one of the biggest mistakes that we see. I can name a bunch of others if you want. 

Debra Chantry-Taylor  21:15 

We're going to come back to the couple in a minute. I was just thinking, of course, I've got my EOS hat on, I can't help myself, but it's like you were saying, that you know, going out and building something without really kind of truly understanding all the well, the people understanding what AI is possible, but also the people who are building the tool not really understanding the way the business runs. Both of those are quite dangerous, and I guess it's like they're within the EOS framework, having your systems and processes really clearly defined is going to help you with doing a lot of this AI technology stuff, isn't it? Because you know what the secret sauce is, and how you operate, and where your levers are to actually drive that business. 

John Munsell  21:52 

Well, you know, here's another thing that people don't really think about. You know, we talked about AI governance a lot, and most people think AI governance is one controlling who has access to what tool, and two, having a document that says here's what our AI policies are. That's not governance, that's just lip service, you know. Really, what you have to have is you have to understand that this is so different for people, and you have to have people thinking differently, so they can't think in the traditional workflows. They have to reimagine workflows using AI, because AI can start to help you make decisions. So, instead of you branching out, if-then do that, you know you're literally, literally letting AI make these decisions, so you don't have to build it the same way, you have to reimagine it, but you need a what we would call an AI centre of an excellence, or an AI council, or all three, there's a governing board in between, but you need a group of people that oversee and champion AI throughout the organisation, and you, you don't have, you don't saddle that with one person, because there's so much for that one person to get their heads wrapped around. It's, it's easier as a team, but it also spreads better throughout the organisation, and, and, and here is, is, is one of the key points that we test for that nobody else really fully understands and doesn't test for, and we test for team chemistry. So, when we have a team of, let's say, nine people on an AI governing board, or an AI centre of excellence, and we'll say that, you know, I say a centre of excellence's job is to get people to experiment, take chances, and learn inside of a controlled sandbox at the very top would be an AI council, which really sets rules and penalties and things like that, but the centre of excellence job is to get people really, really working it with their own fingertips, building their own tools and learning how to make everything better, but take the expertise that's in their head and build little machines that accelerate their job. All right, the team chemistry of that centre of excellence is critical, and so we test for our four things: are they is their primary trait, are they a doer, right? Are they somebody that just likes to do the work, they're producers, so you give me a set of rules and I'll crank out the work. Are they an administrator, somebody that likes to make the rules to keep things organised? Are they an innovator, someone who likes to think of new ideas, loves to push the envelope, loves to get creative with it, or are they connectors? What you would call an integrator, are they the people that really like to build culture and get everybody working together as a team? So everybody has one or more of those as a dominant trait. Sometimes you know some of those things are somewhat absent. It, when you take that as a team, what you don't want on the centre of excellence is all innovators, because they'll take way too many risks. On the flip side, you don't want all administrators, because it'll be way too restrictive, it'll be boring, and they'll be like, "Yo, this is no fun. So, you have to make the right balance. You want a good balance of administrators and innovators on that, and maybe a couple of those connectors, so that you can get the whole thing spread. And as you move up to an AI council, the team dynamic changes a little bit more, where you can play around with it, but most people don't think of it that way. They go, oh, so and so knows AI. Let's put them on the team, you know, and it's.. it's not something that you want to, for instance, just saddle with it, or saddle with compliance, or saddle with a HR. You have to have a good spread, but it really is critical that you have the right team chemistry, and nobody's really paying attention to that, so you see a lot of flops in that area as well. 

Debra Chantry-Taylor  26:05 

I like getting the right leadership team, isn't it? So, making sure that you do have those different representatives at that team. Okay. What a security. We mentioned security briefly when we were talking offline, and you said something along the lines of, you know, people don't really understand some other security aspects that come, particularly once you move from LLMs into agentic AIs, but even at an LLM level, right, or chat level, 

John Munsell  26:32 

that's such a everything depends on the size of the company, too, right, but either, whether you're a small company or a huge enterprise, you still want to make sure that the information that you put into an LLM is not exposed to train the LLM, so you have to know what buttons to turn off inside of the tools that you get, you have to know which licence turns those off for you, and then you need to have some person, some governing body that is literally always looking over the terms of service of those, those tools that you get, and then as you get more sophisticated, you could bring the LLMs on Crown, or you could have your own private cloud. There's a whole lot of complicated things you could do, but you better have somebody that knows how to lock those down, you know. And you're probably familiar with Open Claw. Did we talk about that last time? 

Debra Chantry-Taylor  27:36 

Not sure. Yeah, not remember either. But yeah, let's go through it again. 

John Munsell  27:39 

Very good. 

John Munsell  27:40 

Well, Open Claw is basically a very easy autonomous agent that you could build and deploy and have it deuce, do things with that. You can read all the controversial stuff that they've done. There are other other tools that are coming out now to compete, but the person who created it, he's now working for Open AI, but they, they let that thing loose in the wild, and I've got lots of friends that have loaded it on machines, and they let it do stuff, and you know, you can read about a lot of these people, but one guy, in, for instance, as a, for instance, he, there's no chat layer to this thing, so what you have to do is you have to connect it to to something like a WhatsApp or some other tool that that you essentially are sending text messages to it and it will text you back. Okay, so this one guy, he was, he was used to just sending texts to his open claw thing, and it would reply back after it did whatever mission he sent it on, and then he dictated it, but didn't realise that he was recording an audio file rather than transcribing from, you know, voice to text, and so he ends up sending his open claw bot an audio file. Well, that bot gets the audio file, sees it as an audio file, says all right, I got to figure out how to translate this to text, so it says it looks on his hard drive. There's nothing here to do that, so it goes out into the web, finds an open source product to take an audio file, convert it to text, downloads that, instals it on his computer, converts the audio file to text, understands the command, goes ahead and proceeds, and does all that stuff, and then turns around and has a question. I can't remember where there's a question or whether it was like, hey, I'm finished, tell me what you think of it, but instead of sending a text back, it creates the text, finds another open source piece of software that will convert the text. To an audio file, and then sends him the audio file, and he plays back the audio file, thinking, what's this, and he's like, whoa, how did you do that, and he said, and the, and the thing replies back, well, you send it to me in an audio file, I figured that's how you wanted to communicate from now on, so I just made that happen, but think about what that early means. This thing was making its own decisions, executing things on his hard drive, finding resources in the web, installing them. That could go really, really badly. The things that run Open Call in the background are things that are called skills, if you're familiar with building skills in Claude, and those are markdown files, and so there are a whole bunch of them that started to get populated in these online libraries and GitHub and whatnot, and so these other bots that were out there would go and find these skills and they would instal them so that it could get more capabilities. Well, one in five of those, I think it might be more like two out of five now, were basically malware, and so it would instal malware onto the computer, and some of them would just lay dormant for weeks or months, just so that they would go undetected, and then all of a sudden it would execute something, you know, looking at, you know, recording your screen, figuring out what your passwords are, you never know, so that's what I mean, you don't want, you don't want one of your employees messing around with something like that on their home computer or on their laptop that happens to be a company laptop or maybe it's a personal laptop, but then they plug it into the corporate network. 

Debra Chantry-Taylor  31:48 

It's just such a minefield, isn't it? It's really hard to get your head around in terms of what you're doing. Hey, we've talked about these 10 levels, and I'm just wondering, for people listening in, is there any - have you got any tool where you can actually kind of establish what level you're at. How do people know where they are on those levels? 

John Munsell  32:06 

Yeah, I'm glad you asked that. We do, so that's one of the things that we do with companies, is we test for those two things, we test the level of mastery and we test whether they're a doer, an administrator, an innovator, or a connector, and so that way we have a profile of the person, but we also know their, their, their operating level. We also identify whether they have what we call a skip pattern, which means they're operating very efficiently at level two or three, but they've attempted level seven or eight work, and there's this, they've missed a lot of the foundational expertise needed to really operate safely and securely at a level seven or a level eight, or whatever that level is. So, it'll identify those, and we call those skip patterns, but we test everybody for that, and then when we're working with a company, we create a dashboard that monitors all that, and one of the cool things about the dashboard is what we would call a heat map, and so if you think about it, on the x axis is levels one through 10, and on the y axis is going to be everybody in the organisation, and there's a little light pattern that will show where they're, you know, they're competent, if they're competent at level one, level two, level three, it'll be a bright light if they're less competent at level four, five, and six, and then if there's a skip pattern, it'll be more of like a dark red thing, so when you look at it in this grid where you got hundreds of people, you're like, whoa, instantly you can see the collective wisdom of the organisation hovering in one zone, and it is always level three and below, and your job as a an AI council is to constantly retest people. We recommend quarterly, and so now you can watch those lights go from level one, two, and three and move into levels where we want them to be is like five, six, and seven, and that's the goal of the council, is to not just put somebody in a course and have them check a box because they finished the quiz at the end of the course, we want to actually see the impact, we want to see they demonstrated mastery, and so our assessment, some of it is is multiple choice, but three questions are fill in the blank, right? If they're they're open text, and then we, we analyse that open text. One of them is you have to put together a prompt, submit a prompt, and then we analyse the structure of the prompt, and we see how well you structure the prompt, and that tells us whether or not you, you basically fabricated your answers on the previous questions, but we understand that completely, but yeah, that, that is definitely something that we, we assess, and some people can get through that test in 20 minutes, and other people it might take an hour, it's not easy, because especially. When you get into the personality type assessment, because that you're looking at the questions and the possible responses, you go like, man, I could check all these boxes, but you have to figure out which one you're most like and which one you're least like, and it's a challenging deal, but we've had everybody who's gone through it go, wow, that that's really accurate, that's spooky accurate of who I am. 

Debra Chantry-Taylor  35:24 

Yeah, that's fantastic. I know that you also have an online community, which came out from the book that you've got, and I've joined up to that, and that's got a lot of people kind of helping each other and talking about all this stuff. Where would people find that? 

John Munsell  35:40 

Well, in the book, there's there are links to it. It's it's ingrained.ai and yeah, I think I think you, there's there's a there's probably a link on that page where you can join the community. We're getting ready to bring in some more community managers on that to get more active as the community has grown, so, but the idea is to to help people think through how they're going to be doing this in their organisations, that because that's who buys the book. The book is called Ingrain AI, and it's about scaling AI in your organisation, and I'm fast at work on a second book, which takes it up a notch. The second book goes into these 10 levels of mastery, the three axis AI maturity model, and some other things. So, it's, it's, you know, again, it's more advanced version of the current book, but the current book is very much like the book Traction. It gives you a good idea of the frameworks needed to operate AI in your organisation and scale it, so it's been a great resource for people. 

Debra Chantry-Taylor  36:54 

So, yeah, ingrained or AI. Obviously, there's so much to cover in here, but I suppose I always love to be able to go away and do something when they listen to this podcast, so if you had to say, hey, look, here's three things that you can go and do that will completely change the way you think about it, or maybe give you some some tools to actually help you with AI in the business, what would that be those three top tips or tools from you, John? 

John Munsell  37:18 

Golly, you know what, if if you're a business owner, I'll give you, I'll give you something that Debra, I'll do this for your, your listeners, because I love EOS, and I love your, your show, but here, here's what I would tell you, if you're a business owner, and you're trying to figure out how to do this, one of the things you want to know is, what is the actual impact of AI on your organisation if you upskill your employees? What is the potential upside? And so we have an impact analysis that we'll do for people. It's on our website, and it's $547 but what I'll do is I'll, I'll, I'll give you and your listeners a coupon code, and let's just say it'll be active. 

Debra Chantry-Taylor  38:13 

We'll pop, we'll pop it in the, yeah, we'll pop it in the link, so they've got it here. 

John Munsell  38:16 

I'm thinking, what it might be is, you know, it could be, how about better life, you know? If better life is the coupon code, we, it'll knock 100% off, and say, well, we'll, we can make that until like November. How's that sound, november 26 

Debra Chantry-Taylor  38:38 

Yeah, 

John Munsell  38:38 

but that would give them 100% off. Now, what that does is it's not one of these things that's just all AI. You get it instantly. We use AI quite a bit, but there's a three-step process, and we manually take those three steps. They come back, and we look over it, and we look at your business, and then we assemble it again, and then we, we send it to you, but we also include in it at one, a 45 minute to one hour meeting with you, if you want it, okay? And that way we can go over that report now. It shows you literally, look, if you have, if you have a company with 50 employees, and out of those 5030 use a computer for more than 30% of a day, it will tell you exactly where you'll have to guess as to where they are if they're operating at level two, and it'll explain to you kind of what the 10 levels are, so you can kind of guess, but it'll show you, look, if we took your employees from a level two to a level six, here's what the financial impact would be in year one. In other words, here's the capacity in dollars and cents that you'll recover by teaching them how to build their own tools. Now you can contrast that, and I think it actually does. It will contrast it with you picking an application and trying to. Spend the money to build that application that will take you nine months to build versus just upskilling your employees. It will also give you ideas as to what tools your employees can build when they get to that level six, what kinds of things they'll change in their operation, and I think it even breaks it down by division in your company, so it's a very complicated report, but it's very useful. That would be, that would be the I think the one action I would take is just to figure out what's this actually look like if we were to train our people, because most people think of training as, like I said, let's just go throw them in a course and let them, you know, sink or swim, but you can't. You can't work with AI and not have an ongoing process for learning, you know, because it changes too much. I mean, you know, low Chat GPT 5.5 came out a couple of weeks ago, prior to that, Opus 4.7 came out, and prior to that, just what, three weeks, 4.6 came out, and so these things are coming, and so I know that when we go into an organisation, they're like, well, which one do I use, you know, if you've got Microsoft Copilot inside a Copilot, you have access to multiple tools, and they don't know what tool to use or when, and you yourself said you've got now access to three different tools, and you're 

Debra Chantry-Taylor  41:26 

actually four. I forgot about Copilot, so I've got four, not 

John Munsell  41:30 

what you think about it, right? Yeah, yeah, yeah. And so I have.. I don't know, I hate to, I hate to even mention how much money we spend on subscriptions, but like, for instance, Perplexity, if you've used Perplexity, so Perplexity Computer came out not too long ago, and it's fascinating, and it spins up its own agents that'll run simultaneously, so when you ask it to do something versus just asking Chat GPT or Perplexity Pro, when you ask computer to do it, you know, instead of it going down one path and having to do things, you know, one after another, this thing will spin up agents to do them concurrently, so the speed to completion is insane, but at a cost. Okay, so I was on the pro plan, I think it was like $35 a month, and next thing I know, I got to notice, you're out of, you're out of credits, I think they call them, as opposed to tokens, right? You're out of credits. Okay, I like, well, I can, I can put my credit card in, and they can hit me for more, so they're like, sure, you know, I said, okay, well, let's just max it out at $100 and you bill me in $15 increments. Okay, cool. So I start working again, I get an email, we just hit you for 15 bucks. Like, okay, I work for another an hour and a half. I got another email, we just hit you for 15 bucks. I'm like, whoa, that's going fast. For I think it was six hours later, it said you're out of credits. I'm like, man, I've already burned through 100 bucks. And then I thought, well, I guess I'll have to go top it off again, because I was working on a pretty big project. You can't top it off again. 100 bucks is the most you can do. Your next move is to upgrade to the max version, which was $225 a month versus the $35 a month, you know. So, anyway, but what you have to do is you have to say, okay, how long would this normally take me if I were using either Chachi BT or something else? What I determined is computer does a really cool job and a really nice job. I can't, at this juncture, say whether it's better than Gen Spark for the price. Gens Bark is, is pretty doggone cool, and it'll spin up some simultaneous agent work too. But they're both really good. They both solve different problems, and you have to look at it and say, would I hire somebody in the Philippines? Would I hire somebody in the US to do this kind of work? What would I charge him? Would I get him off of Upwork? Would I get him off of something else? I'd have to, I'd have to pay them a couple 100 bucks to do what I'm doing anyway, and then I'd have to explain to him what I really want, and I'd have to go error check, and it would be a week before I got anything out of them, or I can just put my credit card in this thing and get it done tomorrow. 

Debra Chantry-Taylor  44:31 

Yeah, there's going to be a danger for people like me who like to shop anyway, regardless. But no, I mean, it's a return on investment, and everything in business should be about that. And I think that's really important. I hear people sort of say, but I don't want to pay for this software, and it's like, seriously, for $200 as you said, what else could you possibly get for that? Probably not much. Yeah, 

John Munsell  44:51 

I mean, in this day and age, man, time to market is everything, and you know, you never know whether somebody else is doing what you're doing or going. Down the path that you're doing, but if I can get a prospect an answer and a really detailed, well thought out answer within an hour or two after I met with him, versus next week, and my competitor waits a week, and I got mine turned around the very next morning. I think I did a better job. Odds are I'm going to get the business, you know. Other thing, and I don't know whether you do this or not, but we have developed a series of skills, and we teach our people how to do this, but we've developed a series of skills that do the most insane job of prospect briefing you've ever seen, and they uncover opportunities that no one else is uncovering, and they uncover potential roadblocks that make it really easy for you to open the door for somebody, because you're talking about stuff that no one else has talked about, and it points back to you, and these things can.. this is what cost me so much money with perplexity, was building this set of skills, right? But now that I've got them built, I can run them on any tool, and they're very token efficient now. But I can create, I mean, I could say I could brag about the number of pages, but the number of pages is not really what we're looking for. I'm not looking for bulk, I am looking for actionable intelligence. intelligence, you know, so I can take a market. I'll give you a for instance, I want.. I'm going to be in Houston next week, and I wanted to call on higher education, you know, universities, etc. in that market, and I was like, I want you to do a buyer's brief or a prospect brief in southeast Houston for higher education and what it does is it looks for buying signals in the market, it analyses everything out there and it comes back and it tells me, okay, here, here are some buying signals, this this university just hired this person, they promoted that person, they put this person in charge of that, and it ties it into my products and services, and then it also goes out and says, okay, these are some new new rules and regulations that they have to comply with, according to Texas state law that just passed in December, and then here's some other insurance regulations that they're probably not aware of that are tied to AI adoption, and it goes into this massive amount of information, like, whoa, I didn't even know about Texas House Bill 3125 you know, whatever, but you know, you're when you go into a prospect and you're that fully loaded and briefed, they can't do any gotchas and go, well, you didn't know about this, and you're supposed to be the expert, right? But those are those are amazing things that I couldn't have done in, in two weeks with a team of four, you know. But I can pull that out in about 20 minutes. It's crazy. It 

Debra Chantry-Taylor  48:17 

is absolutely crazy. It takes, it goes back to that original kind of example that you gave the CEO, who just was able to, you know, knock out these three weeks of four people's worth of work into a very short period of time. The possibilities are just infinite, but it's really about don't just go into it and ad hoc start doing these things. You've got to have a plan, you've got to have your people on board, you've got to make sure that you have the right structure, the right governance, that you're not allowing them, the security breaches and stuff. Gosh, mind blown again. Every time I talk to you, I always get something else. I've got copious amounts of notes here, John. I'm going to make sure we have the link to that in the notes, along with the Better Life Code as well. And, obviously, you've got ingrain.ai all the information stuff that you about the book and about the community is in there as well. Honestly, I've just thoroughly enjoyed our chat again. Thank you so much for coming back on. Really appreciate you sharing all your knowledge. 

John Munsell  49:12 

Well, thanks for having me on again, Debra. It's always a blast talking to you, so hopefully your listeners will get a lot out of it again, and hey, you know, we'll keep prog. And 

Debra Chantry-Taylor  49:24 

yeah, you're there to help. Thank you so much. Really appreciate it. 

John Munsell  49:27 

You bet. Bye.