[00:00:00] Welcome to the #CriticalFewActions™ to Improve Your Business podcast. I’m John Downes and I’m here to help you cut through the overwhelm and prioritise what matters most to improve your business. Let’s get started and discover the #CriticalFewActions™ that have the biggest impact.

John Downes: Today I’d like to introduce you to Mark Cameron, who I [00:00:30] believe is one of Australia’s foremost experts on artificial intelligence. I’ve known Mark for over 15 years, and he’s helped a number of my successful CEOs with their businesses.Mark has been an enthusiastic, and some would say obsessive adopter of AI since before the days of GPT; implementing AI and robotic process automations, AI agents, et cetera, for his large government, healthcare and university clients, as well as medium businesses to improve [00:01:00] their business processes, create strategic competitive advantage, and improve profitability. Mark, Welcome. Thank You,

Mark Cameron: Good to be here, John.

John Downes: Well, it’s great to have a chat. So, today we’re gonna talk about how medium business founders and CEOs can use AI to improve their business efficiency and profitability. But first, tell me a little bit about your AI journey.

Mark Cameron: Okay. So my journey has been going on for a very long time. I mean, [00:01:30] probably. So from a, a theory, and strategy perspective, I was talking about predictive analytics, and the growth of data and what that was gonna mean, for businesses and the way that they interact with their customers, and how they create value a very long time ago.

So as, as social media was starting to evolve, you know, way back in the days before, the global financial crisis,

and even though I wasn’t calling it generative AI at those time, I was very much, those terms of predictive analytics and [00:02:00] artificial intelligence and so on, were definitely terms that I was using back in, in those days.

and I remember, actually doing some work in the US probably, probably about 14, 15 years ago now, where I actually wrote a paper and that paper was really talking about an age where. Data was gonna be everywhere. That predictive analytics was gonna be increasingly cheaper, that the pace of compute, was gonna sort of keep up to where we are now.

And we were gonna be entering into a world with digital [00:02:30] assistance or digital agents, that were gonna be able to do a lot of things for us. and was probably a little bit ahead of its time at, at that point, because I remember having a few people turn around and look at me sort of with slightly confused looks.

But, it’s nice to know that a, a couple of them have reached back out in, in years, and sort of recently and said, I finally get what you mean. So.

John Downes: Well,

I mean, one of the things, mark, that I’ve enjoyed about our conversations over the last and is that, you know, you really are a thought leader in this.

Mark Cameron: Yeah, I mean, I [00:03:00] really, really love kind of understanding this and understanding how it impacts, so how technology impacts society, how it impacts businesses, how, how it changes the nature of work. It’s been something I’ve been thinking about and writing about for many, many years.

it’s as I think, you know, I mean, I was, digital strategy economist for BRW for about six years and wrote for, you know, wrote for Forbes and lots of other organisations. So that journey of writing and thinking about how to frame up technology in new and interesting ways has been [00:03:30] part of who I’ve been, been for a very long time.

Um, and then take that into, take that thinking and that way of simplifying complex ideas into the way that I work with customers these days.

John Downes: Yeah, and I don’t think there’s been too many weeks gone by where I haven’t seen in the last, 24 months. You, completing yet another certificate, learning, exercise on LinkedIn or, and elsewhere on artificial intelligence.

Yeah. I think I’m at to, when it comes to generative ai, I think I’m about to about [00:04:00] 120 of them at the moment. So, yeah. I think their term obsessively, adopting is probably a good one. Yeah, I got, I’m glad you got that in before I did. And so in the context of medium businesses, what is AI?

Mark Cameron: look, I mean, ai, AI sort of means a lot of different things to a lot of different people. obviously you’ve got artificial intelligence is a technology category, but there’s also lots of other technology categories that, before chat, GPT. We are called things like predictive analytics. they were called, [00:04:30] you know, machine learning, you know, sort of neural networks.

There was all these sort of technology terms, you know, big data and, you know, all these different terms that have really collapsed down into one term, which is called AI Now. and the reason for that is much like when Google brought out search and there was all this, you know, really complex, this hugely complex environment around search, which was really sort of the domain of the technical expert.

And then Google sort of, you know, took it down to one simple interface, which was essentially the search [00:05:00] bar. that’s what chat GBT has done for, for ai. It’s just driven this really, really complex landscape of, of very, very technical, and powerful tools. And the, you know, collapse it down to a very, very simple interface that anybody can access.

so for most, most businesses, what that means is. for most, you know, medium business anyways, what that means is there is the ability to access this information and this technology in a way that you’ve never been able to do before. And without having [00:05:30] to spend a lot of money on technical, like large consulting engagements or LA or huge amounts of technical capability, we’re able to get a lot of value very, very quickly.

John Downes: Hmm. and I guess part of that is, you know, mark, my feeds are being swamped at the moment by post telling me how I can create a hundred days of social media content in 20 minutes. I mean, two minutes. and that I can replace all of my staff with AI agents. So let’s put the bullshit [00:06:00] aside. You’ve been doing this in the real world for at least the last five years.

What do you see are the top three areas that AI can be used by medium business founders and, and leaders to improve their business?

Mark Cameron: Well, in order to explain that, I think kind of probably gotta talk about it in two different ways. So I think before chat, GPT and generative ai, there was a lot of AI that was being used by businesses and it was in the category of automation. So process automation, robotic process [00:06:30] automation, et cetera.

and that’s reasonably simple workflows, but, but access to data, rules based, so, you know, so processes that are rules based and you could, you can automate a lot of that and add a lot of technology to that and get huge efficiency gains from that. And this can do that still. So things like accounts payable, you know, sort of, a lot of those

processes where there’s lots of documents, lots of information that needs to be processed or inputted into systems or moved between systems. Automation [00:07:00] can do that in an incredibly powerful way. And the current form of AI is able to do that very, very well. and in many cases, much better than the previous versions of process automation we’re able to do.

But what I would say is that area is very, very powerful, but it’s also kind of a scarcity mindset. It’s where organisations are, where people are going. We need to do the same thing that we’ve always done, but faster and cheaper.

John Downes: Yeah.

Mark Cameron: A lot of organisations now are moving more [00:07:30] into the growth mindset. so if they’re saying, well, AI is gonna unlock trillions of dollars of new value in the global economy and maybe doing the same stuff that we’ve always been doing, but faster and cheaper is not the way to capture some of that value, maybe we need to look at new ways of doing things.

And, and AI gives us the ability to look at our business, our business model, and our sector in new ways, and think about it from a new angle and capture new value. So two different ways of [00:08:00] thinking about it. One’s obviously the harder one, but the, you know, the short term versions you can capture, you can save money very, very quickly.

you can get productivity improvements. medium businesses on, on average, if they’ve got mostly knowledge workers can expect to get 30 to 40% productivity improvement with a really well structured AI strategy. but then if you start moving into that next phase of, of, you know, the growth mindset phase, there is potentially unlimited, value to be, to be accessed.[00:08:30]

John Downes: Yeah.

Okay. Well, look, that’s interesting and, I guess you’re talking about not only a level of sophistication, but a maturity model. You know, I’m, I’m still speaking to, founders and CEOs. well, pretty one’s, pretty much the ones that I’m meeting now. Most of the ones that I speak to about how much are they using ai, how, how often do they personally use ai?

I’m still getting,somewhat quizzical looks, the clients that I’ve been working with over the last 18 months, if they’re not using it daily,and they’re not looking at [00:09:00] opportunities to, to implement it in their business, then I’ve already failed. and that’s not the case. So, you know, I guess, I guess we’re sort of seeing a crawl, walk, run sort of model.

so let’s talk about that from the perspective of, okay, well if we think about that, that crawl, walk, run model, how do,business leaders get into and get their organisations into AI in a tiptoe way, then how are they getting [00:09:30] early solid wins, as you were describing in the former, former case where they’re doing things better, faster, cheaper, and then how are they, uh, doing with the transformation of their business model or at least just challenging should we be doing this stuff at all?

can you take us through a couple of case stamp case examples that, that demonstrate each of those three, phases of maturity? Would that work?

Mark Cameron: Yeah, yeah, for sure. so in the initial phases and, and these, these are all [00:10:00] interconnected in many ways, so, I think an organisation that, you know, wants to most organisations wanna sort of capture that, that short term value and be able to get those productivity improvements really quickly, often the best way to do that is through, the training and uplift of your people.

so the shortest path to value quite often for organisations Is how do we help our people become, move up that maturity curve themselves? So if, if you’ve got a lot of, particularly if you’ve got knowledge workers, and so the organisation, well, you know, so office [00:10:30] workers, being able to help them become more comfortable with the tools you give them a good solid policy and framework to be able to work around.

you, you help them, adopt the tools and then work up, a training sort of model. You don’t have to, they don’t have to be amazing, but you’ll see very, very quickly, that’s where the 20 30% productivity improvement will, is able to happen. so for example, you know, like organisations that have to respond to a lot of tenders, for example, or, you know, those kinds of processes, things that may have taken days will be be, you know, shut, you know, shrunk down to [00:11:00] hours fundamentally.

So those are and processes that may have taken two or three people could be done by, you know, by one person now. So you’ll see really big productivity gains off the back of that. and that’s the starting point.

John Downes: So, you know, when I, I’m still, and it amazes me, but that’s because I’m deep in it now. You know, I’m still speaking to, business people and, and saying, you know, how much are you using AI and how, how are you using it and what’s innovative for you? And, and I’m still coming across people and it’s [00:11:30] continues to surprise me, although it irritates me,still coming across people saying, well, you know, all your information gets shared with the world, you know, it’s not really secure.

And I sort of say, well, hang on, tick. What you’re really demonstrating here is you dunno, Didley squad. and, and so. If we, if we take a leader of an organisation that’s at base zero, how do we encourage them even to get started and take those first baby steps? Because I, you know, for them to, to do that, [00:12:00] there’s still a ways from trying to figure out how to actually write a 10 to faster, better, cheaper.

Mark Cameron: Yeah, absolutely. the, the easiest starting point is just do it, you know, a, a very short, sharp, current current state assessment. so, you know, process, would we work with the leaders? So to help them identify areas of their, of issues with inside the business processes can be improvement.

areas where they can see that there are is, maybe productivity issues. And then help ’em sort of articulate what a, what they could, where, what a, [00:12:30] a more automated workplace could look like. Then run a survey across the rest of the organisation where we’re looking at, AI, aptitude attitude and mindset.

and then what that does is provide feedback to the, the leaders and say, here’s where. Some of your people are, are excited. Here’s where some people are, interested. Here’s where some people are worried. and if we were to design a training and change program for your organisation to get the most out of, out of these tools, this is what it would look like.

and very, very quickly [00:13:00] you can move that organisation from being, you know, sort of disorganised to a very, a very coherent and unified approach to the way that they’re adopting AI and starting to experiment and find those use cases internally that are gonna start to have the big impact. and, and I think the reason why I sort of tend to focus on that first and foremost is the data out of, Harvard and MIT recently has shown that the organisations that prepare their workforce to be able to [00:13:30] adopt AI at scale rather than.

The way that technology has been bought previously, which is by a technology piece and push it, you know, push it down. But if they help the organisation broadly get prepared to adopt it, they end up getting far greater return on investment because the use cases start getting identified within the business.

So, so you get top down and bottom up and, and of course the people, the bottom up ones is the people who are actually doing the processes can actually see the [00:14:00] areas where things need to be improved. and as they start to understand how AI works, they’re the ones who actually start to innovate and, and drive the change themselves.

John Downes: Absolutely. Okay. So yeah, let’s, let’s go through a couple of case examples there in that context.

Mark Cameron: Yeah, sure. so. I’ve got, I mean, there’s quite a few, I mean, there’s lots and lots of cases, studies out in the world. But the ones that I’m simply working on at the moment, it’s a manufacturing company. really, really, really interesting manufacturing [00:14:30] company. and they, you know, got, got me into look at their sort of AI strategy broadly, what, as we’ve been working on that piece, you know, that strategy, we’ve sort of identified lots of different areas of the business that can be proved.

So customer experience, knowledge management, so the way that people inside the organisation access knowledge, which means that new hires will be, you know, be able to be effective far more quickly and customers are gonna be able to be serviced far more rapidly. [00:15:00] and the means that the chances of losing orders, you know, drops significantly and so on.

What is ended up happening is they’re starting to identify the fact that there are lots of other competitive issues at play. And maybe, you know, and what, what we’re sort of working on is actually how do we sort of use this process of investing in technology to slightly shift the orientation of their business.

So their manufacturing business, they’re, they’re by default, quite product focused. And what they need to do is, [00:15:30] is capture, market share and gain and protect their market share. So using this process to be able to start to focus much more on, customer experience and bringing customer centric thinking into the organisation.

again, it’s, and the reason I use that as an example is, you know, AI is a very, very powerful tool wherever you point where, wherever you point that gun is where you’re gonna start running and you’re gonna run really, really fast. Because, because that’s what it does. and. it’s always really helpful to think about is this process [00:16:00] or is this direction that we are gonna point the gun really where we need to go?

and in some cases, you know, maybe it’s all you’re doing is making it, making the wrong process go faster,

John Downes: Yeah.

Mark Cameron: which mainly would be producing. So, you know,it can really, I mean, this is the thing, but exploring AI broadly and how it could impact your organisation can really help, can really sort of.

Allow the, allow the organisation to lean into those harder questions that they may have been putting, you know, putting off for a while. not always, you know, so quite often, you know, there’s just things [00:16:30] that can be done. as I said, you know, you can pick off parts of the business and get dramatic improvements.

but, you know, often the, those, questions do come up. but with one, another organisation working with, in the, in the travel space, they do a lot of communication, a lot of marketing, a lot of,you know, that the search engine optimisation, all that, that kinda stuff. and for them it’s been like adding 40% more staff to the organisation without actually having to make the hire for that organ.

Well, because [00:17:00] we’ve gone through and looked at all the different tasks that are happening internally, so, and how. And how they’re structured and realizing that if we actually augment a lot of those tasks with AI and we sort of get really clear about which processes are better off being done by in an automated AI sort of technology way, which processes are best being done by by humans, and then which processes are best being a combination between the two.

What ends up happening is we find massive, massive [00:17:30] improvements that that business can make. So. For example, doing, um, search engine remote optimisation analysis, strategy and execution, sort of something that was taking them months before. now that process can be done in days, you know, doing,social media marketing before, you know, there would, it would be quite a process of, you know, working on what they needed to do, working with their internal design team to be able to get a couple of variations.

Those variations would be put on, the, the, you know, put on social media and, and with some advertising [00:18:00] dollars and tested. So ab tested in the best one wins. Now in a process that takes quarter of the time, they can get 40 variations and they’re basically running those at all times. So it kind of becomes like a darwinistic approach to, to, to marketing where, you know, just the, the best one’s winning all the time.

so it just gets better and better and better and better. So they’re. You know, their ability to be able to get very high quality marketing outcomes as well as much better data for their market is, has gone up exponentially.

John Downes: [00:18:30] Mm mm And, and coming back to that manufacturing example, what were some of the early wins at, you know, at that sort of very tactical level that, that you were starting to see?

Mark Cameron: so with a lot of organisations, where they’re just starting on, on the journey, the early ones are usually quite similar. so sales and marketing is usually one area. so, you know, most organisations have a CRM. Most organisations have salespeople. Most of the salespeople hate using the CRM.[00:19:00]

John Downes: That’d be right. Yep.

Mark Cameron: so, and, the reason they hate using the CRM is because in many cases, apart from filling in the form saying, I met with this customer, and click, click, click, and many to get the nuanced information you’re asking the, you’re asking the sales guy to essentially be a creative writer, go, you know, tell us what happened in the meeting and what, you know, try and remember everything.

And it’s just, it’s just not how they’re wired. so, finding a way, so using AI tools so they can essentially just have a conversation with the [00:19:30] tool. So it says, tell us what happened in that meeting. Who are you meeting? And it, it runs through and asks them the question and they can just talk. It then translates, and then automatically puts all of that stuff into the CRM.

The quality of data goes up really, really, really quickly.

and then as a result there becomes a cultural shift where all of a sudden the sales guys almost start competing with each other to have, you know, who’s, who’s entering the most information, you know, as opposed to, you know.

Yeah,

Yeah, yeah, exactly.

so that’s, I [00:20:00] see that, that that one happens often. the other, the other one is around knowledge management. So how does the organisation access information internally? so, you know, old reports, tenders, proposals, you know, any document that’s been created, and particularly if there’s some technical documents, often, you know, business has been running for a while, have, you know, a vast, vast library of all that information.

And when, and it’s quite hard to navigate. And, you know, usually you’ve got sort of a couple of. Key [00:20:30] person risks. So you’ve got, you know, one guy, you know Jeff over there, he knows everything about that thing. So if you’ve got any question, give him a call. He’ll know where all those files are. you know, or, and, but if, you know, if Jeff gets hit by a bus, we’re all buggered.

and then you’ve got, then you’ve got, new people coming into the organisation who, you know, they go through an onboarding process. but then after that they kind of feel a bit embarrassed to ask questions. And so that time to value of, of a new hire becoming really getting [00:21:00] up to speed and earning value can sometimes be, you know, like six or seven months for example.

So, so having really good knowledge management is a good way to mitigate those risks and those issues. and there are, there have been obviously knowledge management tools around for a long time, but they required. Even a bit of knowledge to be able to navigate them. So, you know, you’d have to understand the, the jargon and the nuances, probably some of the structure.

You’d have to maybe understand some of the product names, all of those kinds of things. whereas now [00:21:30] with, a chat interface, you can just ask a question and it’s able to find that immediately for you and bring it up to you and, and help you use it really, really quickly so you know that, that time to value or that ability to be able to get information and use it in, in the way that it’s needed right then is can just be, go down to seconds fundamentally.

John Downes: yeah, yeah. And, and I know that, you know, with one of my chemical clients, one of the things that they’ve been been doing is, using, GPT as a tool for helping them with their, [00:22:00] their process documentation. You know, if you, if you go into any business, it’s under a hundred million and, and talk to them about their processes, quite often they’ll talk to them, talk to you about them with intricate detail.

If you ask to actually see an SOP, you’ll get a slightly embarrassed look or, or you’ll be told about something that was written seven years ago and it’s not what we follow now. but with With having GPT as the accelerator or the catalyst, to be able to actually have a, have basically a live conversation, get it to draw [00:22:30] on, the vast expertise of the web, to come up with the SOPs for handling, flammable liquids in a confined space,in process chemistry reactions, at scale, you know, being able to pull together,materials, handling procedures or safe work, procedures, you know, is, is a hell of a lot faster because it’ll get you there 80% of the way and then you can actually add to it very quickly and customise it to make it real, and context [00:23:00] sensitive.

Mark Cameron: you know, that’s, that’s completely transformed the way they document their organisation. absolutely. And I, agree with that. There’s, you know, really good ways of doing, SOPs for, as you see, and getting it to do business analysis more broadly. and, and then, uh, you can also, there’s a, I think the thing, you and I spoke about this a while ago, but there’s a, an AI model called Tree of Thought where you can get it to actors as several different experts in a particular topic.

And they all debate each other and see whether, like, not only a document [00:23:30] to meet your needs, but it’s almost being peer reviewed. So it’s been looked at by lots of d from lots of different angles.

John Downes: Yeah. It’s actually fascinating to watch it do that,because it’s almost like there’s a, there’s a live conversation and, and, tribal council determine who wins at the end.

Mark Cameron: Yeah.

John Downes: But, but it is a great way to actually get different perspectives on the same question, uh, and,

and instantly.

Mark Cameron: absolutely. And I think, for services businesses, so, so what I’ve found for services businesses, some of the things that they do, it is more [00:24:00] the nuanced changes that they can do that, that actually have the biggest impact. So, for example, with one organisation’s working with them to be able to build out, prompts that act as, different types of customers.

So before they go in and have a meeting, or they’re doing a pitch, it can then, you know, they can upload a proposal or whatever, and it acts as different types of different types of customers and it’ll, it’ll grill them and, you know, challenge them really, really quick. You know, so it can be a conversation, but what it’s doing is they’re preparing them for that [00:24:30] meeting, um, in ways they haven’t been able to do, or, You know, or understand, you know, like they’re going in for it to meet with a client. I, I did this myself a little while ago. I was meeting with a client and I got the AI tool to spend, you know, half an hour talking to me about that, that cut that client. So, so it was multinational client, so it’s quite well documented, but their, but their, their culture and their, and their ways of working, and understanding how that came to be and so on and so on.

By the time I sat down and talked to them and [00:25:00] was working through the program with ’em, I understood those, that language and that way of working in a way that, you know, they were sort of surprised about. ’cause normally it takes people quite a while to sort of get up to speed. so that, that kind of stuff I think is really interesting, as well.

And, you know, using it in those, those ways.

John Downes: Well, that’s actually a really interesting use case because, you know, good sales practice is that if you’re preparing for presentations or you’re putting in a proposal or a, a tender and you’ve then gotta go and do a pitch, then obviously,one of the best [00:25:30] things to do is to actually rehearse and practice.

and often that’s sort of done in the, in the Uber on the way to the pitch or, or in the 90 seconds before you hit, um,connect on Zoom,uh, whichever works best for you. But, but because of the ease of being able to access AI now being able to actually have that conversation, I love the fact that it’s called conversational ai as well as generative ai, means that you can actually have that conversation and, safely interrogate it and [00:26:00] not even your peers realise what a deal you are. and you know, there’s a whole bunch of reasons why everybody knows that rehearsing for a presentation is actually leads to great performance. and it is remarkable that that 99% of people don’t rehearse before they go into a presentation. And, and it’s train on the day, which is always a recipe for,innovation can fly.

Mark Cameron: Definitely it’s

John Downes: Or a surprise.

Mark Cameron: approach. Yeah.

John Downes: Yeah. Yeah. I call that courageous. [00:26:30] Uh, people who know me know that that’s another word I use for stupid. okay. So, so that sort of gets us to, our tier one users, I mean, you talked about, you know, data analysis, and its ability to analyze data of just having, having a lunch with, with a client.

And we were talking about how they have a, a, a manufacturing process that deals with liquids and,when they, repeat that process 50 times over the course of two or three years, and they’ve got all the data to prove it. [00:27:00] Sometimes,the liquid that comes out is, is, beautiful and clear.

looks like white wine and sometimes it’s cloudy and looks like apple juice. both work pretty well. they both achieve the same outcome,but they don’t meet the clarity, test that their client expects. and,on any given day you can actually ask them, so why is it clarity today?

And the answer will be,

Mark Cameron: Not sure.

John Downes: dunno, because, you know, we’re, we’re, we’ve run this thing, a lot of times, and we test the temperature and the pressure and the [00:27:30] filtration and the additives and all that sort of stuff. using our QA processes. but you know, as we were talking about it, they were sort of saying, well, you know, we’ve now got 50 sets of data for, for each of these things and you know, it’s only $20,000 to run one of these processes.

You know, we can probably now feed this data into, AI for some, some data analytics and get it to do an analysis that we would not thought of.

Mark Cameron: Absolutely. That’s a hundred percent true. yeah, I mean the, you can, and particularly if you use [00:28:00] models that are more inclined for, data analysis or you tell them to act as a data scientist, um, it’ll, it’ll find things that, and ask questions you won’t even think of, that’s for sure. and, and the great thing about.

So that’s what I was saying before about thinking about the difference between automation versus a ai. So, you know, automation, the way it was, and the way we think about it is automating a process, versus ai, which is essentially adaptive. and it can have a lot of automation in it and [00:28:30] process efficiency.

But for that example, it’s a really good one where it kind of needs to learn and understand over time and understand what data’s missing, what variables it may not be seeing, and start to learn and try and work out what’s going on. and that’s where the, these kinds of models can be incredibly, incredibly powerful.

as you say, it can, it can do all the analysis on that data, but if you keep it running, it’ll continue to improve and continue to learn and continue to make that process better, rather than just locking, [00:29:00] locking it in and, and automating it.

John Downes: Okay.

so we’ve, this is a, we’ve, we’ve largely spoken about, do it better, faster, cheaper, of the stuff we currently do. Let’s jump forward perhaps into a case study of, of someone who’s now taking it to the next level of maturity.

Mark Cameron: that’s really good. So there’s, some of the best case studies around this and there’s a lot to be thought about, with this around how to, how to even think about the strategy and the [00:29:30] governance of this is, so, one of the best case studies I think is, I’ll use this one ’cause it’s, it’s a bigger, it’s an interesting one.

but there is, this is being replicated at all size businesses everywhere. it’s just, it’s much more to do with the way that it’s being approached. So, the Mayo Clinic and the US has done a really incredible job of their AI adoption. they, they spent rather than going really heavy on governance and writing about risk and then, you know, buying lots of technology, what they actually did is, [00:30:00] is go, quite light on governance, considering some really good governance principles and guardrails, but not overly governed.

So, you know, lots they didn’t do. Lots and lots of subcommittees and lots and lots of risks, risk committees and so on. there was, and then. They spent a lot of time with staff enablement, training and uplift. And then they had, this is one of those organisations that had hundreds and hundreds and hundreds of bottom up use cases.

the last measurement was close to 40% productivity improvement across the board, which is [00:30:30] enormous for an organisation of that, that size. But not only that, they found hundreds of millions of dollars of new revenue because they were using these tools to identify ways. I mean, first of all, some of their products were, had AI baked into them, but they were taking new products out to market off the back of the way they were using these tools.

Now that, that’s a, case study, even though it’s a big organisation, what it’s doing is it’s showing an approach. I think that’s really important and that approach, I’m working with several companies right now [00:31:00] following the exact same model, which is, and even though they’re, you know, smaller companies, you know, I mean Mayo Clinics, you know, I.

Tens of thousands of people. Um, but with this,with organisations, you know, that are much smaller with, it’s exactly the same approach, which is rather than getting caught up in all the idea of the risk and what it could be, and, you know, here are all the problems and we don’t believe it. It’s actually just let’s, let’s get put a wrapper around all the things that could go wrong and just be really careful about how we manage those.

Let’s not overfocus on [00:31:30] those and let’s not start moving into everything that could go right, and how do we set your organisation up to capture that value as quickly as possible. Um, and that’s where the training the people, right, you know, really well thought through change management process, smart investment into technology and enabling people to be able to, as they identify, use cases, work out really quickly how to prioritise them and invest in them, and, and work, you know, how they’re, how they’re gonna scale throughout the business that.

Work of making, [00:32:00] of enabling that. it’s sort of, it’s the, it does, I’d say go slow to go fast, but it’s not slow. It’s a fast process. but what it means is that pace at which return investment happens starts going up really, really, really quickly. and it’s just a, yeah, as I said, I love that case study because that mindset, I think is, is the one that in general gets the biggest return on investment.

yeah.

John Downes: So it is a really, you, you’re bubbling up not just a series of ideas. funny how how old [00:32:30] processes, don’t change. They just get renewed. So instead of having an ideas box, now what you’ve actually got is a series of use cases where people have, have either used the AI to actually do something themselves, or they can conceive and perceive how exactly it could be done by connecting different systems or interacting differently with the employees or the customers or the service providers of an organisation.

Mark Cameron: And, and what it means is the organisation starts to have, early on, they start to have [00:33:00] the challenging discussions. So there’s a term that I like to use a bit, which is they use just purely to raise eyebrows and get people thinking in a different way, which is, you know, how does your organisation look when your carbon base workforce, which is your people, is outnumbered by your silicon base workforce Three to one, which is what you know, what is gonna happen with ai.

agents, how does that change your structure? What does that change from a leadership standpoint? how does that change the way that your business operates? so those questions, even [00:33:30] though you don’t have to answer them all, it’s good to be able to be thinking about them. and realizing that as, as you’re investing in AI and moving up that maturity curve, you’ll then increasingly start to want to have sort of autonomous processes, which are AI agents and those process, and those, those agents will work with people and all of a sudden that’s gonna change the nature of the way that your business operates.

So the, the, and you’ll have an AI enabled workplace and workforce.

Are you looking to pinpoint the [00:34:00] areas that will make the biggest impact on your business? Check out the organisation performance and value diagnostic. This powerful tool lets you evaluate your business on multiple facets and provides email reminders to keep you on track. Head over to www.CriticalFewActions.com.au and get started now.

John Downes: Okay, so, so tell me a bit more about AI agents. So obviously, you know, most, most people are very familiar with the concept of, [00:34:30] of ai, which is really a GPT or Claude or Gemini, or copilot as just being a replacement for Google search,on steroids. And, and you can ask it to do intelligent things and it will come back with, answers either surprisingly quickly or just answers that surprise you.

but now we start to introduce the concept of AI agents. So tell me a bit more about what that is and how it’s applied in, say, a medium business.

Mark Cameron: [00:35:00] so yeah, so yeah, as you say, so chat, GPT is essentially, it’s a tool with a human. Is there sort of asking it to, to do stuff on your behalf. so gimme answer about this and then it can be a conversation. So you’re going backwards and forwards and it’s kind of like dealing with an intern really, or a, you know, a smart junior.

they’ve got lots of knowledge but don’t really know what to do and they need instruction. and that’s kind of where we’re AI agents,increasing, gonna be increasingly autonomous. and they’re going to have a [00:35:30] role to play. Now, some agents may be sort of human in the loop agents where it will run through and do a bunch of, say data analytic, sort of analytics, pulling together insights and then asking a human to say, okay, this is what I think.

Do you agree? And then he rather say, yes, no, or give it some feedback, and then it will continue on with a process. or it’ll be a, an agent that just has no human interaction whatsoever. and so it’ll go be able to go away and do things on. Your [00:36:00] behalf as a consumer or as a, as a person within inside the business or on behalf of the business itself.

and, and only just report back once, once processes are finished fundamentally. Um,now at a consumer level, what that, what that could look like, the one that people like to think about all the time because it’s a, it’s a, you know, it’s a part of the market that’s, that we all don’t like, which is as travel for example.

So, you know, rather than, you know, booking, if you, you know, gotta travel to another state, for example, rather than booking the airline, then booking the hotel, then booking the Ubers and [00:36:30] doing that, you just say to the agent, I need to go to Sydney on this date. And all of that stuff just happens in the background.

and it just organises everything for you. If there’s a change, it’ll read, book everything and make sure that it’s just keeping up with your all’s schedule. now that, same concept. Probably happen over many, many, many different market categories. and we’ll see that, you know, we’ll that’ll be the battleground for, from a commercial standpoint, for the big technology.

So the Googles and the apples and so on. but at a business level, so, [00:37:00] so, you know, meeting business level,agents are gonna be able to do things that, you might’ve engaged and it’s somebody outsourcing to do or a, a part-timer to do so, you know, bookkeeping, to some, some of the bookkeeping stuff, you know, so look sort of, data, you know, as I said, some analysis work, some reporting work, coming in and doing, as you said before, you know, doing, building out standard operating procedures, for example.

agents will be able to do that, that work for you in your organisation. Uh, standardise things, speed it [00:37:30] up, reduce costs, and then help you find ways of becoming much, much more efficient.

John Downes: and I guess one of the, I guess fairly basic, but, but.

Somewhat obvious. examples of that is, is web chatbots now that are, that are basically sitting on, on a lot of people’s websites, which 24 7. Anybody could go into it and say, okay,tell me all about your CEO masterclass service.

what, what’s the,what’s the curriculum? how long does it take? What am I [00:38:00] gonna look like before I start? What am I gonna look like at the end? what sort of diagnostic tools do you have? how do I engage with those? and they can self-serve comprehensively information, in a conversational style that they wouldn’t be able to get by just searching through a website.

Mark Cameron: A hundred percent. I mean, I, I’m a,

a director and invested in a business which has built a number of AI agents, and one of them is in the psychological space, but for all, for businesses, that have issues with psychological health. So they might have been [00:38:30] going through significant change or they’ve got a, you know, a, a sort of a remote workforce or diverse workforce that is, you know, that has, can go through periods of vulnerability.

they’ve, we’ve built an agent that runs, it’s basically just a conversation over a long period of time. Unlike chat bots where every time you log in, often you have to start from scratch. Go, Hey, who are you? And gimme your details and so on. This remembers everything about you. And we’ll say, you know, last week you said this, you know, how are you feeling now?

and it’s got like about [00:39:00] 160 different points of diagnostics in there. and it’s designed to help people understand when they’re entering periods of psychological vulnerability sooner, and direct them to appropriate help before it hits crisis point. That’s really what it’s designed to do.

So it’s, you know, smoothing out that demand curve for psychological, support services. but that there’s essentially, that’s an agent. That’s what it’s, what it’s doing. Now, of course, there’s lots of privacy in there and, you know, the organisation can’t see any of those that, you know, that wouldn’t work obviously, [00:39:30] but, What it’s doing is acting on, it’s learning for about the person it’s learning. You know, it’s starting to build a psychological profile of who they are, understanding where their issues may be, where some, an anxieties may be, and starting to help that person deal with that on a day-to-day basis. And that’s, that te you know, without agent ai, that would’ve never been possible to, to do.

John Downes: Fantastic. So. If I’m now, uh,thinking about it again, you know, as a medium [00:40:00] business owner or leader, what area of AI application do you think is gonna give me the most benefit in my business?

Where do you think the quick runs on the board are gonna be?

Mark Cameron: Yeah, so as I said before, I think I. those areas where you’ve got systems that are being underutilised. so as I was talking about before with sales guys and the CRM for example, or you’ve got information that’s difficult to, to capture or to be able to get, get value out of. So how are [00:40:30] people accessing information?

Those two areas tend to be the, big cornerstones of value quite quickly inside the organisation. there are, there will be, without a doubt, there’ll be some processes that absolutely could be automated, you know, so there’ll be some, probably some reasonably, mundane and basic processes that you could automate quite quickly.

and they, you know, there will be, you know, some of the admin and support services probably within, inside the organisation. but what I would say to that is, So previous technology ways have been quite functional. So, you know, they’ve [00:41:00] been, like, cloud for example was really about how do you, is really about making it more effective.

and there’s been technology ways around marketing, around finance, et cetera. AI is different in so much as that it is actually impacting every single knowledge worker. Every single person who sits in front of a computer will be impacted by this technology. and so the biggest bang for Buck, almost always for most organisations is that piece is [00:41:30] how do we get our people going from.

Aware and not really knowing what to do to, we’ve got a unified approach and we are working together to learn on, an agreed pathway basically. So getting, getting the organisation as a whole started and moving in the same direction is without a doubt, one of the best ways to get bang for buck.

John Downes: Yeah, yeah. Simply actually harnessing the masses. even if they’re only just starting out, getting them started out, giving them the [00:42:00] tools to actually do it. Giving them the encouragement to, to have a crack, make some mistakes, make some learnings, and, and share ideas.

Mark Cameron: Well, and and probably what’s going on with most organisations, I mean, we do the surveys. I find this almost always is that there is. there is a perception of often by leaders that my organisation, you know, our people won’t like it. They’ll be afraid. They’ll be worried about, well, you know, what’s gonna take their jobs and all this sort of stuff.

almost always when we do the [00:42:30] survey is that the organisation is actually really, really excited about ai. They want to learn, they want to do more. They want to have a bigger impact on the business. and what they’re looking for is direction,

John Downes:

Mark Cameron: and permission. And, and they’re, and all, and often what they’re doing is because they haven’t got that, they’re doing it sort of on the, you know, they’re doing it secretly, you know, they’re doing it in their, or using their own tools and so on, which does create some risk.

so I think, that process of just saying, okay, this is here. It’s gonna impact our organisation, it’s gonna impact our sector. Let’s get on the front [00:43:00] foot and turn it into an advantage rather than a threat, and do that quickly. and, and decisively is, is really gonna give you big bang for buck.

John Downes: Absolutely. You know, one of the things that, that I was reminded of by,one of the students at,that I work with at,at the universities at the moment in their, their final, business degree years, is, you know, when, when GPT became,easily and readily available,all the universities clamped down on it and said, you know,

you can’t cheat, you can’t use it. We’re gonna [00:43:30] test to make sure that we can see whether or not you’re getting GPT to write your essays for them. and fortunately, they’ve woken up to themselves that A, they can’t win that battle, and B, they’re missing an incredible opportunity.

And I look at, the fact that what is it, 40 50% of our tertiary education, marketed is students with, that come from a non-English speaking background. And how amazing it is to help somebody who has, English as their second language for them to be able to actually, do and [00:44:00] prepare their work and their analysis, but actually use a, an incredibly competent English tool to actually help them write an essay, or finesse an essay that actually is easy then for it to be marked.

and actually puts them on a level playing field with their peers. So rather than being penalised because they’re from another country now, they’re actually now democratised again.

I think that’s, that’s one of those amazing unintended consequences, [00:44:30] of these tools. And there are so many unintended consequences of, of ai we’re, we are limited only by our imagination.

Mark Cameron: Yeah. And there’s a lot, there’s a lot differently, a lot, a lot. I think, you know, there’ll be some, you know, some curly, unintended consequences, but a lot of very, very positive ones, I think.

John Downes: well that leads us into a really important question because I’m so, so tired of listening to the naysayers. So, mark, what do you see as being the gotchas for our medium business? Starting to [00:45:00] embrace AI and encouraging their staff to embrace it? What can go wrong?

Mark Cameron: so I mean, that’s where, I mean, having a policy is really, really important. so, and you know, it doesn’t have to be over baked. but I think having a policy, remember the days when social media. Just started and every business got really confused about what it all meant. And eventually, and then they started getting really surprised that all their staff were using social media and eventually they had to get a policy and say, Hey, you’re allowed to do this, but you can’t do this.

And, [00:45:30] you know, don’t talk about our business online. Just gave them some great, you know, gave them some, some, a framework to think about what was, what was acceptable and what was unacceptable within the context of work. Well, that is kind of the same thing, when it comes to AI is just a few more risks, obviously, you know, if, you know, you don’t want to be using, you know, even, even free models, if you set them upright, won’t, Exposed data, but you have to set ’em up, right? So you’ve gotta help people understand what that means and how to do that. with, if you, you’re using a paid model internally, [00:46:00] you know, there’s a way to do the setting, so none of the data that gets put in it is gonna leak. or at least, you know, you know, you can sort of mitigate that risk enormously.

and then there’s other ways of making sure, you know, you get increasingly secure, as you’re going up to more and more, sort of sensitive information. so I think that that’s a really good starting point.

John Downes: So one

issue is a security slash privacy issue.

Mark Cameron: Yeah. Yeah. So that’s, that’s definitely one. I think the other one is, and I, you know, you’ll probably see a theme coming through here is, is getting myop [00:46:30] myopically focused on a, on a short term innovation opportunity.

and thenAllowing the business to get disappointed if that one doesn’t work. So, as I said, if you have a, an an automation mindset with ai, it may not deliver the business case you’re expecting. but if you, you sort of take a slightly different route, you’ll get a much, you’ll get much greater return on investment.

so,

John Downes: About that.

Mark Cameron: for, yeah, for example, for example,like, so you were saying, what you want it to do is take every [00:47:00] customer email and write a response in a certain way. and that’s gonna save us lots and lots of time. And then you realise that what it’s doing is not, it’s not actually responding in the way you want it.

For example, it’s quite an automation mindset. So it’s just saying, this is, this is a process or a project that is taking a lot of time and we want to, we wanna save money on that, and we wanna reduce the head count on that. And you sit there and you do your business case, and then you run it and you realise, actually we need a human in the loop.

Oh, yes, it is. Making it a bit more efficient, but nowhere near as efficient [00:47:30] as we thought it could do. And as a result, the, the narrative in the organisation goes, well, AI isn’t working and it’s not delivered the promise. And, and then that you allow that narrative to continue, which slows down the pace of, you know, the belief in the technology and the pace of the potential pace of change.

John Downes: So that’s a, that’s definitely something to avoid and think through the possibilities very, very quickly. Whereas a smarter way to use that stream of customer inquiries would actually be to look at it from a data analytics [00:48:00] perspective to talk about, so what are the key themes that we’re learning from this? And what does that tell us about our customer service and our customer experience, and how can we actually, improve on that, which ultimately may end up reducing the number of, emails we get in the first place.

Mark Cameron: Yeah.

Yeah, absolutely. Absolutely. That’s exactly as I, you know, rather than, you know, thinking about the symptom, going back to the root cause and Exactly. Absolutely.

John Downes: Yeah. And, and use a tool that’s competent to look at a couple thousand emails and actually find out what some of those root causes might be.

Mark Cameron: yeah, [00:48:30] absolutely. I think the other thing that, organisations, although I’ve seen a lot, is, and again, these, these, you know, again, there’s themes here, as organisations, You know, seeing the promise or the, or the, you know, the over promise often of the technology vendors. So, you know, like, so there’s, you know, you know, buy the technology, the licenses, you know, roll it out across the organisation, go, well, that’s gonna deliver us 30% productivity improvement.

Yay. And buy it all and then it doesn’t happen.

John Downes: Yeah,

now was that a [00:49:00] fault of the technology or was that a fault of change management? play the.

Mark Cameron: yeah, yeah, exactly. Exactly. and that’s definitely a gotcha.That can cost a lot of money and create a lot of issues within, inside the organisation because it’ll, as you said, it’ll, you know, there’ll be change management issues, belief issues, it’ll slow things down.

It’ll, you know, all of those kinds of things. where, you know, just spending a little bit more time, we would’ve avoided all of that to, to start with. so those are, those are definitely, definitely gotchas. and then also on the [00:49:30] other side of it, Just being overly confident about what it can do. you know, I mean, ob obviously there’s, you know, we all know that AI can make stuff up from time to time.

So if you’re not helping your people understand that they have to develop deep knowledge about a topic so they can ask sensible and insightful questions of the tools and then have critical thought, so they’re reviewing those with a really critical lens, then there’s, you know, there’s chances of things going wrong.

but I think that’s, you know, and it’s, [00:50:00] again, it’s why I try and avoid the, you know, it will create efficiency gains and, and so on. But I, I try and avoid that language often because if you’re putting the pressure on the people to efficiency, efficiency, they’ll try and find ways to be faster and faster and faster.

And there’s a chance you could increase risk. Because the quality goes down. So you know how to manage those things, as, as the business is learning to adopt the tools.

John Downes: Yeah, I think you also hit on a really important point, which was,you know, be, become, adept [00:50:30] at, asking deep and intelligent questions. being adept at, framing up your questions using, the rich experience and understanding of a topic or a subject matter, to help the ai, really hone in on what it is that you’re trying to get it to do.

then, and then be, diligent about, interrogating the responses that it’s come up with to ensure that it’s, that it’s reasonable. When you think about that, really what we’re sort of saying is our [00:51:00] knowledge workers or our people in the organisation, if they’re worried about their. Their jobs, becoming obsolete.

yes, they’re, their repetitive low value transaction processing, tasks will probably reduce dramatically, but actually their importance to the organisation because they bring this thing called wisdom, which is the experience and the understanding and the ability to actually see what’s [00:51:30] not quite right, is actually where they’re actually now adding a far higher value to a process which might be better, faster, cheaper, which

Mark Cameron: Absolutely.

Well, that’s, I mean, it’s a,

John Downes: guarantees their job

Mark Cameron: Well I think that,

John Downes: and also allows them to get, do all those things that they wanted to be able to do but couldn’t because they never had the time.

Mark Cameron: yeah, yeah.

John Downes: it’s a real value to the organisation.

Mark Cameron: Absolutely. I think that’s the key. I think that’s the key. is, If you think about, if you think about what, what is AI [00:52:00] actually doing? So it’s speeding up. you know, so it’s giving the access to information for all sectors and all knowledge workers. It’s speeding up the pace of change, you know, so all of those kinda environmental factors are at play.

So what, how does an organisation, so without thinking about AI or, or the technology piece, what does an organisation need to do to be able to respond? You need, the organisation needs to be able to set itself up to be able to survive and thrive in that world, which means it’s gonna need to be able to be really efficient and also very agile.

it [00:52:30] needs to be able to respond to market dynamics into a competitive, into a changing competitive landscape very quickly and be able to have the insights and the ability to do that.

John Downes: Hmm.

Mark Cameron: AI gives you ability to be able to do on parts of your business really efficiently and really effectively. and so does automation allow you to do those things really, really effectively, efficiently?

Your people give you the agility, the people give the ability to be able to shift and change. because, you know,

AI I, I mean, at the moment [00:53:00] anyway, AI is not particularly good at dealing, you know, it’s, it’s by definition, it’s looking at data, it’s looking at the, it’s looking at history and trying to, trying to map a pathway forward.

but it’s not able to navigate the future the way a human can,Um, and I think, you know, that’s, that’s the way the, if you really want to create a highly competitive business, I’d say is, you know, there’ll be probably, you know, 60% of your business that can be probably massively impacted by ai.

And then you allow that, Big chunk [00:53:30] of, of business processes and impact to be able to be even more impactful by leveraging the people in new ways.

John Downes: Yeah, absolutely. So.

How do you help in this space for medium businesses?

Mark Cameron: so,

probably the best way to put that is, a lot of what I do is that early stage guiding. so, I help organisations understand their current state.

So I’ve got a process for being able to map, map out. Where they are in terms of an an AI maturity model. and they, it may, may be a little bit further along than they [00:54:00] think once they understand the entirety of, you know, where their, their staff sits and so on. and

then where they need to be. and, you know, like, you know, so say it’s on a five step, you know, like maturity model.

Most organisations don’t need to be at level five, but they might wanna be at level three or four, for example. so being able to help understand where they are and what’s required to be able to get to that next leap along that journey.

so be able to map that out, help enormously with, so the strategy to do that, the change management frameworks to be able to make that work and the support and the training around [00:54:30] that.

and then bring in and help them adopt the lots of the technology. So help ’em go through the technology selection,work out how best to, you know, best to align that with internal capabilities, if they need to change KPIs so that that business strategy piece combined with the technology advisory piece to be able to be able to make that actually work with inside their organisation and then, you know, hold their hand through that process to make sure that they’re getting the return on investment.

John Downes: Yeah. [00:55:00] Fantastic.

so Mark, if, if a medium business is wanting to get up to speed, are there any resources, courses, videos, prompt guides, assessments that you’d recommend that they have a look at to actually get them ready to have a conversation with someone like yourself?

well, we’ve published a lot, so on our website, and I’ll share that with you if you wanna put this out, there’s a lot of guides around prompting case studies. a lot of, articles around what it’s going to do [00:55:30] for, goal, internal goal measurement and what ai, how AI can impact strategy and so on. So there’s, we’ve published a lot.

Mark Cameron: there’s also a lot of great podcasts out there. One called me, myself and ai. I particularly like that, where it’s looking at businesses around the world and how they’ve used ai. Now it’s usually large enterprise businesses, but the stories and the way the stories are told and the process they went through to be able to have that impact, they’re, they’re relevant for most organisations.

’cause it’s, you know, [00:56:00] it’s, even though we were, some of them were making big investments, that what they were doing was, you know, able to be adopted by anybody in business fundamentally. So I think that’s, I think they’re, um, that’s a really, really good, good resource as well. and then, I think it’s, I mean, it’s pretty hard to avoid, content when it comes to AI at the moment.

John Downes: Yes.

Mark Cameron: It’s kind of everywhere. it’s, it’s really about finding the ones that, that resonate with you and your, your the way that you see the

world. Yeah.

John Downes: Yeah. Fantastic.

[00:56:30] If you really want to kick off your strategic planning and work in a supportive environment with other CEOs and leaders, the CEO Masterclass in strategic planning and implementation is for you. This 10 month small group course helps you craft your strategic plan presented to a group of peer investors and develop a habit of monthly implementation.

Our wait list is open now at. [00:57:00] www.CriticalFewActions.com.au

John Downes: So, Mark, finally, if a founder or a CEO is looking to start utilizing AI in their organisation, what #CriticalFewActions™ should they start tomorrow if they did nothing else?

Mark Cameron: So if it was a founder, I would be asking the question

that

most organisations never get the luxury of really being able to ask, which is, if we could design our business tomorrow with AI at its core, how would it look? Founders have the ability to [00:57:30] be very agile and to be able to pivot, often, their, their business, depending on what, how old the business is and where, what sector it is it’s in.

But they, they have that ability that a lot of organisations don’t do, and they’re able to actually think about.

If we could redesign this sector or this business with all the technology ability that we have now, how would it look? And it probably wouldn’t look anything like the business does today. Um, many, many sectors, many, many businesses that you could say that absolutely.

You know, if you had AI at the [00:58:00] core, the way those businesses operate and the way those sectors operate would completely transform. if you’re a CEO of an established business and is thinking about this, the question I would be asking is, how much do we, do you think AI is gonna impact your sector?

And what does your organisation need to look like to be able to survive and thrive in that new world? And, and am I, as I would say, my [00:58:30] suspicion will always be that. If you get, you know, smart people in the room and come up with some good ideas, you’ll find a few use cases that will get you some incremental improvements.

But if you take your organisation through a, through a change that can give them the tools, you will find huge, huge, dramatic improvements and, and return investment shifts. and I think those, those would be the things that yeah, I’d start to think about.

John Downes: Hmm. Fantastic. Mark Cameron, as always, it’s been a delight to [00:59:00] chat with you. Thank you so much for your time.

Mark Cameron: Okay. Thanks John. Absolutely. My pleasure.

John Downes: Thanks for listening to the #CriticalFewActions™ podcast. Don’t forget to subscribe and grab the downloadables from the show notes. And if you found this episode helpful, I’d really appreciate it if you share it with your peers and your team. Stay focused at take action.

The #CriticalFewActions™ [00:59:30] podcast, including show notes and links, provides general information only and is not individualised business advice, nor can it be relied upon. As such, you must take responsibility for your own advice, decisions, and actions.