AI Hype Meets Reality

Lucas 0:01

A really powerful question to ask yourself and your staff is what do you love doing? If you just focus on what you love doing daily and everything else on your list of to-do was automated, how would that make you feel.

Magnus 0:26

There is a $112 billion number floating around Australian boardrooms. It's the projected economic value of AI adoption to our economy by 2030. It's large, it's well-sided, and it's doing exactly what numbers like that always do, making everybody want to act before their thought. Meanwhile, on Australian construction sites, productivity has grown around 1% in two decades. 70% of digital transformation projects fail to deliver. And every business owner from Brisbane to Bunbury knows the feeling of signing off on a software project, waiting six months, and getting back something that was technically built but doesn't actually solve the problem we're trying to solve. The gap between the headline and the shop floor, between knowing what is wrong and knowing what to build, is where most of the real work of AI in automation is going to be done over the next decade. And the people doing it are not the ones writing the McKenzie reports. They're the ones sitting across from people getting it wrong. Today on the Mastering Podcast, we're joined by two of them. Lucas Meadowcroft and Dean Kavanaugh are co-founders of Crofty, a Brisbane-based AI innovation consultancy that spent 12 years working specifically with construction and manufacturing businesses on part of the technology projects that almost no one talks about. The part before anything gets built. Lucas is the founder, the strategist, and the rare Australian tech CEO whose origin story runs through TAFE rather than university. Dean is a technical lead, the one who sat across from more developers, more information teams, and more frustrated business owners than most people will in a lifetime. I'm Magnus Olson and I'm joined by my favorite co-host, a powerhouse in education and Australian masters athlete, Don Sanka Small. Lucas and Dean, welcome to the podcast.

Lucas 2:26

Thanks for having us, boys. Thank you. I want to be the favorite podcast host.

Don 2:30

Well, yeah, it's it's weird when you're having another co-host in front of you and calling me favorite, that's me. I didn't write this at all.

Lucas On TAFE And Strengths

Magnus 2:39

Lucas, you and I have got something in common, and academia is not it. You finish year 12 with an OP of 24. That's one point above the worst possible mark in the old Queensland system. I didn't even get to year 12. So university was off the table. You're 24 years old, and you're the CEO of an AI consultancy and co-founder of four companies. What was that kid in Ipswich actually good at that the school system wasn't very good at measuring?

Lucas 3:12

Making friends, influencing people, going to the skate park, having a lot of fun. I had no idea what I wanted to do. Both my parents were blue-collar workers, and a university or or just the education system in general was uh was broken for me because being, you know, having dyslexia, I can't even say that word, like it's just it's just not a thing, right? And so how do you go from following a path of what the schooling system tells you to then finding your passion? And and and school was just about having fun. It wasn't until I left school and really dove into uh technology and build back then was building computers. Um, and once we started building computers, it's that's where I realized what I loved was the outcome, not the actual building part. That was kind of okay, but the outcome was seeing the lights light up and the people using the computers and how they were using it and what what it was making them feel.

Don 4:04

You've become a bit of a voice for Dave Queensland. Um, what changed for you the moment you walked into Bremo Tave that you couldn't achieve from a decade of high school?

Lucas 4:16

Practical implementation of using the tools. So back then it was learning how to develop, learning how to code, and and using the the like physically the teacher showing you step-by-step process on how to use that use that technology uh to physically have an outcome. Back then it was building web pages and you know building really many apps. Um, yeah, chalk and cheese is what it is today. Uh, but it's like the practical application. Yeah, uh, you don't get practical application in school.

Magnus 4:42

Yeah, let's just expand on that because both of you and I we've gone through that TAFE system, me as an apprentice carpenter, and Don's gone through the university sector. Would you say that that's true? Because I think that the TAFE system is very practical, whereas university is a bit more theory-based. Is that true?

Lucas 4:58

It's definitely practical. So the funny thing is that it's I was on a traineeship and I was actually on a traineeship with where I'm at Dean, my first computer job, and Dean was my boss. And so um so this is how we we initially connected. I went through this traineeship program with that with that IT company, and it was through that that process, and that was through TAFE, uh, to go, well, this is what I'm doing in the real world. How does that relate to the actual certification, uh certification, and then bringing those two together?

Don 5:25

I think it's it is becoming better. Um and also there's a false truth to that too. If you look at certain professions, like if you want to become a doctor, you do have to have that practical sort of education. If you want to become a teacher, you you do have to go and do certain hours nurses, is the same thing. There are certain areas of education that still lacks that practicality. Um, especially if you're in entrepreneurship or any of those parts that you could have taken. I think there is a lack of practicality. I think that's what we're talking about.

Lucas 5:54

Well, I think even 20 years ago, entrepreneurship wasn't even spoken about at school.

Don 5:58

That's right.

Lucas 5:58

There was no there's no conversation around, hey, do you want to start a business? No, you you're the you're at school back then to get a job, to get a profession, to go and work in a government or corporate, that's what you get told to do. Uh, I think today it's a whole different cattle of fish, right? Yeah. And most 16-year-olds don't even want to be at school because they know they won't be doing whatever they're told to do. They want to go and start a start a company or be an influencer or do some sort of you know online.

Magnus 6:23

I think that's sorry, just to put some context around it, Dean. So you guys met at TAFE. So you're a bit of the yin and the yang, I understand. Give us a little bit of your backstory, Dean.

Dean 6:32

Yeah, so

Dean On Building The Work

Dean 6:33

I did actually I completed a bachelor at uni, um, but that's where we met was through a computer store effectively, um, servicing both retail uh and to business as well. Um, so I did have that shop front, and that's where Lucas comes in doing the computer build, computer repair, all that sort of stuff through TAFE. Um, that's really where the training ship kind of kicked off from um from the relationship point of view.

Don 6:56

Yeah, so you know, every founder has a co-pilot, and Ian seems like you're the co-pilot, but it's it's the unglamorous job, right? He goes and gets on stages and tells the story, brings the customers in, and then all the problems becomes your problem. Yeah, pretty much. Right, and you have to do the build. Um, what did do you guys see in each other that the rest of the world in that made you connect and started building this whole journey together?

Dean 7:22

I guess for me, um, and I'd I'd tell Lucas this, I I hope I tell him a lot. Um, his superpower is really about um, and I've forgotten it now, which is great. What do I keep telling him? You inspire people. See, this is why you work together, I don't know. He can get everybody motivated in a room and really inspire a lot of people to start thinking differently. Um, once they're on that sort of thought train, that's when I can really kick in uh and start to leverage where people's heads are at and bring some of those ideas to life.

Lucas 7:55

And because of my ADHD, there's no chance in hell I could sit down and actually code something for months on end and build a product. It's just not who I am. I get bored until six until this last six weeks with vibe coding. Uh it's just it's it's too slow. And so I love inspiring and telling the story, but then am I going to physically solve that problem for that for that customer? Hell no. That's I literally handball it over to Dean, and that's Dean's Dean's uh that's where he steps in.

Don 8:22

So t tell us a bit about vibe coding. It's become a buzzword, right? It's is it's um coding for idiots or I didn't call you an idiot, but I'll say yes.

Lucas 8:32

I'll let Dean answer that one.

Dean 8:34

No, yeah, so the vibe coding side and Lucas's experience with this might be a little bit different, but essentially being able to use plain language to develop applications is really what that boils down to. Um and I guess where it's become a buzzword is because there's a lot of expectation that I can just use whatever language, plain language, and I can build amazing massive applications. And I would say we're we're getting closer to that, but we're still a little bit short of it. You can still do some amazing builds through this uh concept of vibe code.

Don 9:04

Before

Vibe Coding And Early AI Triage

Don 9:05

we go more deeper into the technical side of it, right? Um you started Crofty in 2014. Chat GPT was eight years away. Talk to us about the bet you were actually making at that point. Um AI consultancy in 2014 would have been a very different story to sell, the potential that is there now.

Lucas 9:24

And you don't talk to anyone back then about it because no one believed you anyway, or no one to talk to you. So we actually had an IT support business, and and in our journey within a couple of years, the biggest frustration we had was felt like we're running a childcare center. You think, okay, we have engineers, we have technical people, they're solving technical problems for customers, but most technical people don't like talking to people. They don't like managing their own time, they don't like you know that whole accountability. They just want to go, there's a problem, how do I solve it? So the biggest frustration we had and where the AI came from uh originally was to solve our own problem, which was triage. When a support request would come in, we need to analyze that support request and then understand who's the best person to solve that at what point in time and who's what that priority list looks like. Because in the in the technical world, the the what you class as devs now, but even um IT support engineers, most people want to do what's quick and easy. But as soon as you go, well, this is a three-hour, four-hour complex thing you've got to go solve, that gets pushed to the bottom of the list. And because we all have endor endorphin hits, we don't do the quick ones 15 minutes, 30 minutes quickly do those ones, but then the most important ones get left. So it was actually back then, I'll I'll pass this over to Dean. Dean built a really amazing glorified spreadsheet to try and solve this. Was our original first AI AI model.

Dean 10:43

Yeah, so I guess going through that sort of understanding where the priorities are because people did want to manage their own time, they just they would take from the top, they would cherry pick basically. Um yeah, and where we brought AI into the mix, I guess I got frustrated with it and go, There's got to be a better way. I can either throw more people at this, but we're in the tech space. I like building things. I know AI can is meant to be able to tackle this. We can see it play with the numbers for banks and insurance. How's it going to play into this sort of space? And that's where my sort of journey started in that stuff.

Magnus 11:14

The focus literally just around time management initially, yeah.

Don 11:18

I think a lot of people don't understand artificial intelligence has been around for a lot longer than Chat GPT, right? That's it's automation, it's automated coding, which you guys have been doing. Like even spreadsheet is a simple version of the macros.

Lucas 11:33

Yeah. Macros and spreadsheets. A simple form of machine learning back then.

Don 11:37

Fair CD. Most of the things that uh an Apple has been running since 2007 has been around artificial intelligence, right?

Magnus 11:44

Ray Dalio's been using it. I remember reading in his book that it was one of the things he was using it back in the 60s and 70s in a in a in a in a format.

Dean 11:54

Yeah, it was purely uh a lot of the early stuff is very much the numerical forecasting and trending sort of stuff. And that's largely why it was taken up by banks and insurance, you know, to be able to predict certain market trends or insurance risk and things like that. Absolutely. It was very numbers focused.

Magnus 12:09

So

Why Construction And Manufacturing

Magnus 12:10

you guys could have chosen any industry, financial to be able to go into, yet somehow you chose construction.

Lucas 12:17

We helped construction manufacturing organizations to take their traditional processes and bring them into the 21st century. So what we found was a big thing. Everything else that you have to do in construction manufacturing call is going to be one of these big IT of what they love, unless it's a very important thing. It didn't make sense what wasn't being done properly. So once we start picking up that and you go deeper, go deeper and asking all these questions completely. And that's because you're starting with the problem. Starting with the business problem process that we're trying to solve. So we need to start the right direction, we fix the real results.

Magnus 13:18

We focus on practical innovation using automation in AI, where it makes a real two years profession.

Dean 13:25

We really focus on people and the process to make sure we're balancing it against the big thing.

Lucas 13:38

Construction hasn't changed. It's been lining up for years, huge opportunity. And not just opportunity for us, but opportunity for the actual business themselves. Because they've they've done it, they've used the same spreadsheet for the last 20 or 30 or 40 years, and they haven't changed using that spreadsheet. Then you go, well, why do you use that spreadsheet? Well, that's what we've always done. There must be a better way. And so, yeah, honestly, for the last couple of years, we've just we've gone all in in that sector, and we've done probably 100, 150 projects, uh AI projects in that sector. And this is not just builders, this is like the construction services industry, right? So think about all the services that people provide to the construction industry. Trades, trades could be scaffolders, you know, all the way down to painters, it could be uh little sparkies and plumbers and like all the certain tradies that uh that are in the sector. But they all have a really common theme. They quit a job, they win a job, they deliver a job, deliver a project. Same theme. So if you if you find one problem to solve, you essentially can solve that same problem in every other business that's in that industry because it's very, very similar. Uh so I would say majority of the work that we've done so far is very much in that front end around the the sales, the the quoting, the estimating, the tender writing, uh all that upfront sales component, and then all the way back to delivery of the actual project and the means and project and forecasting and so did you did you have to create a problem or did you have to find new ways of solving a problem that already existed?

Don 15:03

Those problems have always existed.

Magnus 15:05

I I think what I found in construction, every project we weren't doing the same project, so very different projects in very different sectors. So some of that was a bit difficult to get that that same approach. Like when I first started out as a carpenter, we built units and you're doing a block of four units, so it was all repetitive work. What you did in one, and then you went and did that, so it was easier to build up systems and processes. Then I moved into highly complex architectural homes where there wasn't two walls that were the same.

Lucas 15:33

That's the good old 80-20 rule. You know, if you can solve it, solve 80% of it the same, which is like the very similar in every single project, then all you're tweaking is that last 20%. Yeah.

Dean 15:43

Yeah. Yeah. I would say that I would say for for the parts that I saw at least across it, there were ways that um often I guess in that sort of industry they they don't see it as a problem. So I don't think we're introducing or trying to find the problem. It's it's there. Um quite often it was a case that they didn't realise it was a problem or that that problem's probably a harsh word for it, but there was a different way and potentially a better way to do it. It just wasn't really in their face for that. And sometimes it is a case of a salesperson to bring that forward, absolutely. Um, but then as we started to see this AI revolution come around, people are starting to ask more questions around, oh, wait a minute, maybe there is a better way I could be doing this. And that's really started to spur things along more recently. Before ChatGPT, very hard to convince anybody to even look at it uh in with a different lens whatsoever. ChatGPT did the marketing for you, right?

Don 16:34

That's an amazing.

Dean 16:35

It absolutely did.

Don 16:36

Mate, we're gonna get technical and ask some specific questions, but before we get there, when the laptops are closed and the the WhatsApp chats are done and when your all your threads go quiet, what does the rest of Lucas Meadowcroft and Dean Kavanaugh story look like? Lucas, I know you're an avid um runner. What makes you get out of bed at 5 a.m. to put those shoes on?

Lucas 16:58

I feel like I have no choice these days because your body just has to do it. But I think if you roll back, you're a better person for having exercise first, in my opinion. You know, I feel better, I can get through the day a lot easier. Whatever stress was going to be on your plate, that stress doesn't exist because you're you're mentally in the right frame of mind because you've started your day off in with the right, right uh mental state. If I don't exercise and I don't run, or if I have a week off because you're tapering for a marathon, oh, that week sucks. It's like oh you just want to do something, but you know you shouldn't because you need to listen to your body, you need to listen to follow the plan. Um yeah, exercise first before anything else.

Don 17:34

Well, how's the work-life balance? Does that exist?

Lucas 17:37

Does that exist? Work-life balance. The last six weeks has definitely not exist because I've become one of those uh gaming sensations where you know people get lost in games, I've been lost in vibe coding. I've had the most fun I've had in forever in the last six weeks. It's been it's been so much fun. It's it's one of my biggest struggles in life is that whole work life balance. It's like if I spend two hours in the morning exercising, then I'm straight into a laptop. Then I'm on that laptop until 7 p.m., 8 p.m., sometimes midnight, and then you switch off, you have four or five hours sleep, and then you get back in the road two hours running again the next morning.

Don 18:11

You might want to come and have a chat to Magnus, talk about his uh body, mind and spirit um uh coaching. So Dean, how about you?

Dean 18:19

Um family is probably my bigger drive. I've got uh 12-year-old son and wife. Um and that's that's my main motivation, I guess. It's that that whole drive. And for me right now, as my son is getting older, it's uh I guess bringing him some new experiences. And when it comes to the work side of things, it's to show him I guess uh good hard work will actually get results and and to try and build that sort of uh resilience into into my boy.

Lucas 18:51

And then then Dean's the complete opposite, and Dean's not an early early morning person. It's uh Dean will be up until 2 a.m. vibe trading. Well actually developing properly, not but not vibe coding.

Mindset Blocks And Productivity

Magnus 19:02

So just back on that construction sector, I'm looking at what is it about nine percent of the GDP in Australia. Is that right, Don?

Don 19:10

Yeah, I believe it is. It's gonna be more moving towards the Olympics. It's it will get better, but I think it is at about 9% now, but it's 100% gonna get a lot uh a lot worse.

Magnus 19:20

I was fortunate enough to interview Jack Hutchinson recently, and it was interesting we had the same conversation around productivity. So just getting an understanding of where you sit with clients in both sectors. Is it a do you think it's a measurement problem? Is it an investment problem or is it a culture problem?

Lucas 19:35

It's a mindset problem. So a lot of these larger custom businesses, they're not pivoting, they're not moving. They're like, nope, we've got the sector, we've got five years worth of projects, we don't need to change. Smaller companies know that they need to pivot. If they don't pivot, they won't be in business in three years. And this is not just a conversation, it's a fact. If you don't pivot in the next uh little while, you will be out of business in three years.

Don 19:58

Every other industry's had humongous growth. Like in 20 years, there's been about 1% growth.

Magnus 20:04

Correct. It reminds me of a story of when I was an apprentice, I was on this construction site, and the builder we're working for, he's about 40 years of age. And one of the other carpenters, he would have been 65 about to retire. And we're sitting around about 20 of us, plasterers, bricklayers, carpenters. And the old carpenter just um the boss said, I think we should do it this way. And old Kevin, 65 years old, listen, Ray, um, mate, I don't mean to tell you what what what to do, but mate, I've been doing this for 45 years. And Ray turned to him and said, Kevin, just because you've been doing this your whole life doesn't mean it's right, and the whole place just erupted. I'll never ever forget that. And it just reminded me when you're talking now in relation to the construction industry. And that's I work with a lot of people like that's like this is how we do things, the way we've always done things, and you'd be up against that daily.

Lucas 20:55

Yeah, so we just don't even try to get into that conversation because if they're not if their mindset's not ready to change, this is not a technology conversation, this is a mindset change conversation. I want to do things differently just because I've used that spreadsheet for 30 years, doesn't mean I still need to. And so if they're not willing to have that conversation, we just move on straight away. We come back to them in six months or 12 months when they're really willing to have that conversation.

Don 21:16

Do you see a genuine opportunity for our that sector, one of the largest sectors that Australia's got to grow exponentially by especially with the new generation, yeah.

Lucas 21:26

The the 25-year-olds, the 30-year-olds, they don't want to do it the way it's always been done. Their mindset is AI first. There is no option. Yeah, there's no, hey, because Kevin did it the same way for 45 years, they don't know that they've got no preconceived ideas. Their preconceived idea is we could only do it as an AI first mindset way.

Magnus 21:44

So, as you know, the development audit up in Melania, we used all modular. I'm a huge fan of modular. Modular is massive over in Europe, it's massive over in Canada. And part of that reason is with the temperatures, they can literally only build outside for six months of the year. Do you find that crazy? crossover between construction and manufacturing, that the whole modular concept, is there some room there as well?

Lucas 22:07

You're talking about the actual implementation of the module or from a tech point of view, modular between the company?

Magnus 22:11

Well, from a tech point, so you're basically looking at manufacturing, that is construction.

Lucas 22:16

Yeah, and a lot of the manufacturing plants we work with actually manufacture for the construction industry. Right. So it's it's very, very much construction.

Don 22:24

So it's pretty well joined, right?

Lucas 22:25

But I'll let I'll let Dean dive in around like what we see in one company is very, very similar to the next.

Dean 22:31

Well at the base level, particularly on the the I guess how they run the the company, as you said before Magnus, around each project can be quite different. You're still selling a certain dream between lots of units versus an architectural house. The way you execute that can be quite different. And everybody I guess a lot of the industry uh out there kind of try to position their brand around certain types of builds and what they excel at. That's really where some of this um nuance comes into finding out well how do you do it? How do you want this to kind of fit with your your brand? This is when we're getting into some of the discovery side of things about them. But I don't I don't fully understand the question that you're asking at the start with when you put that to Lucas.

Magnus 23:16

Because when I look at manufacturing that's obviously in the factory. When you're doing modular that's built in a factory but it's obviously a construction project that's just not built in situ. So that's obviously the two industries that you guys are working in. And seeing some being able to improve that modular side of things with AI, robotics or whatever, that's that's helping the actual construction industry.

Dean 23:38

Yeah I guess robotics I will say I I would love to get more into the robotics space. It's not been an area of my specialty that's that's more my tinkering area rather than professional space. But I do see that sort of path there there is something about the Australian I I don't know whether or not the the mood is changing or not but um modular homes I I actually quite a big fan of it we were looking at getting modular homes for ourselves as our next move but there is still something about getting a Queenslander or or whatever the the very traditional large house sort of scenario I think before there's any sort of big move in that space it might be a bit of adoption around tiny home you know that tiny home movement that's happening a little bit of change there before that really has a big flow-on effect. But that's that's a very non-AI related sort of talk there and going on my own personal experience as I've been going through housing stuff.

Lucas 24:32

Society drives it right so what we're seeing from a robotics point of view is this year to 2026 is a year of experimentation with robotics but 2027 will be the implementation of robotics. So a lot of warehouses today or factories they haven't had the resources or capacity to be able to do the implementation just because they they're not an Amazon. They don't have the millions of dollars of funding lying around but we're seeing a like a million dollar robot for a factory now drop to 80k and new 80k is now affordable.

Don 25:04

If you read Elon Musk's autobiography you can see being a trail b trailblazer the costs of that like building those robotic factories the things that he had to do and the calls he had to make and the costs of actually getting it right um yeah that's not something that a uh mum and pop business can can do in every day and I think you're right.

Lucas 25:22

So this this time next year they will be able to I can guarantee it because so give you an example back to your point before uh Magnus like say we're doing a lot of the the PO sales orders processing a lot of data which is very common in these manufacturing plants as an you know customer order comes in they physically have to key that order onto their system. So we automate that process. Then the conversation turns into okay so once the once the orders are automated how do we then make the the process better from a man the actual manufacturing point of view or the logistics as in it's now being manufactured it's going to be put onto the back of a truck and then it's got to be shipped. That process still requires a lot of physical human labor that's what we're going to see massive disruption in the next say 18 months because the robotics will come in we've automated the front end now we can automate the back end using the robotics that we can now get out of the US or get out of China and then configure it to physically do that pallet loading and so forth.

Don 26:14

AI

Robotics And The $112B Gap

Don 26:15

is already here right the AIIA and Google have projected that AI could add 112 billion to the Australian economy by 2030. AI Lab Australia has been more conservative models putting it at about 45 to 50 billion annually Dean you are actually delivering this stuff right when you hear those numbers um what do you think the gap is between the projected numbers and the work that has to be done for them to be true.

Dean 26:47

A lot of it Lucas has spoken a lot about mindset that is going to be a big driver behind this. There is an element of that where there will be people who adopt and that will be the basically that leading edge for a lot of people I I think in Australia at least we're going to it's going to be a bit of a wait and see type of scenario. There is a little bit of hesitancy you'll always have a few front runners um I think in terms of the the stats behind that it we'll definitely get there. It's not going to be an even race across all industries um some will really blaze ahead and others will fall behind. And they're the ones that'll probably drop off and there'll be a little bit of ripe opportunity potentially that's kind of how I see that playing out.

Lucas 27:32

The biggest investment so far in AI has been into the legal industry and there's really good legal AI platforms coming out in the US to help legal firms be better be faster be more cost effective for the client I talk to loads of legal firms here in Australia. They're not embracing it they're pushing back they're holding back which means you'll have this this all this investment goes into AI all the companies are going and waiting and holding and seeing what's going to happen the next three years what what's going to happen? There'll be this big gap Kodak the Kodak moment but then what I well my personal opinion is all these new law firms will develop as AI first law firms they'll then take the what the customer wants they'll take and take that customer order. They'll use AI to generate that legal document whatever it might be um and then what happens to all these other firms like it's it's an interesting three years.

Don 28:25

Do you think it's a leadership issue I saw I think it was PW uh PWC report that came out saying that you know 75% of the leaders are not ready or embracing artificial intelligence.

Lucas 28:36

Are you finding that uh less hand of the entire population. Yeah.

Don 28:40

Does there need to be more education around the leadership people who are driving these businesses on what artificial intelligence is and what it's a fear conversation in the media.

Lucas 28:51

The media drives fear.

Magnus 28:52

Yeah and so what do you think just expand on that for us so you're saying only one percent of the population are embracing it at the moment and in that's Australia.

Lucas 28:59

In uh in the West in the yeah that's called the Western population I can't really comment for China because hardly many stats coming out of there. But in the Western population 1% of businesses are embracing AI consumers are getting the ChatGPT apps and then whatever and doing the pretty email research or whatever but that's not really adding value to your life but businesses embracing true AI I guess the space that we see every day literally still only 1% of businesses are embracing it.

Don 29:28

So outside of the funny pictures and the fake videos that are coming out that's that's grabbing attention you're not seeing the future of work actually adopting I'd say social media side of things is doing probably a lot in the AI space.

Dean 29:42

So if you're in the business around social media it's a different story. There's a lot happening there.

Magnus 29:47

Is there an industry that's embracing it more than others? So you just mentioned the legal industry and you're sort of mentioning the social media so you're referring to the marketing industry.

Lucas 29:56

I don't know about industry from you from your perspective where I guess maybe this is just what's coming in up in front of me at least maybe maybe more so the departments sales types of roles absolutely um molly agrees uh sales departments there is stuff across delivery particularly where it's very digital sort of delivery documents or or reports being generated that sort of stuff back on people for a minute is it is it true as well that with chat GPT one of the biggest things it's it's from a mental health from a psychology perspective people are using and also from a medical medical diagnosis perspective is that are those rumors true I can't comment on actual stats but uh you do see it a being used a lot for personal um coaching whether it's whether it's mental health coaching or or personal coaching I personally I've put every blood work every statistic every Strava running everything I've ever done from a fitness health and mental perspective is inside my let's call it mini AI model. So he knows he knows me better than I know myself right what's what's the purpose of that for you? Yeah it's a great question Magnus why did I start that journey I think I just wanted to know that I was on the right track and I was I was progressing in the way that I think we as humans want to keep progressing. But I asked my question this the other day because we're moving from Chat GPT over to Claude and I was like do I even need to keep the two years of history and I've been thinking to myself why do I even want that history so I'm not sure about the answer to it but I think the original reason was um just to make sure that what I'm doing from a from a uh mental health and an eating perspective is that am I achieving an outcome?

Magnus 31:36

I think it's great from a record keeping perspective I like the idea of it from and I'm not a a huge data person but just understanding what I've done like you said one I can track the improvement but secondly I think from a private health perspective I know some private health funds but if you've got that data the the data that can't lie for example like your score you're getting your HRV scores your resting heart rate all of the exercise that you do and then you look at nutrition your blood work ADEXA scan like if all of that and I'll be more than happy to upload that to my private health and see if I can get a significant discount when you look at some of the people because I've got my wife's sorry my brother's wife's a nurse one of their daughters is a nurse and the stories that they tell me in relation to to people that are filling up our our hospitals.

Don 32:26

Talking about data data sovereignty is going to become a huge topic of interest right do you think the federal government is investing enough in the AI adoption side of it and then ownership or do you think they're investing in the wrong places? I think the number was something like 22.7 billion of capital but it was in all these places that I've never even heard of future made in Australia AI adopt centres the responsible AI network um is this a substantial investment or or are we just pushing capital at a problem that they're not solving?

Lucas 33:04

I feel like this is a rhetorical question. We already know the answer. But I would love to why do you hear your thoughts on it and deed that's a that's a big word for someone with your educational why do you need a AI adoption center?

Don 33:19

That's what I'm asking my question is are we actually putting the money in the right places?

Dean 33:24

Yeah well the fact that you hadn't even heard of those things before I guess this is less to how is the government uh or are they putting enough towards it? Maybe they are but then what happens afterwards people don't know about it. They don't know the effective use of it and all that sort of stuff.

Don 33:39

That's where it 22 point seven billion is a big chunk of cash for our economy.

Dean 33:45

So yeah yeah and how do you make proper use of that? Yeah um what's the outcome?

Lucas 33:50

What's the outcome for society? That's what I'd like to know.

Magnus 33:52

I

Why Projects Fail With Five Whys

Magnus 33:53

mentioned in the intro 70% of all digital transformations fail. Why?

Dean 33:59

Uh all right so across that it's typically because they're asking for the wrong thing or they don't even know what they're asking for in that sort of space. Uh mostly leaders are asking for outcomes that are not particularly measurable or not easily measurable in that sort of space. They might say I just want to make it efficient I want more efficiency. That's a great sort of concept but it's pretty hard to know what to target there um to know what actually needs to change and for many I guess we we don't see that sort of um that translate across there's not a direct one-to-one I want more efficiency in what where do we start all that sort of stuff and they can't normally pinpoint that um and that leads to I guess projects being started incorrectly or with the wrong sort of goals in mind.

Magnus 34:46

So I'm I'm thinking so it's potentially somebody who sticks their hand up and said I want I want AI automation um I want to introduce it into my business but they really don't know where or how and just by throwing some money at a digital solution they don't really fully understand what their outcome is that's what you're saying Lucas yeah they're not asking the five whys so this is like something that um so what are the five whys the five whys this is what everyone else asks what are the five whys what are the five whys yeah so wives or whys two different yeah they're they're two different things let's go with five whys um so when like to Dean's point when you start a project or you want to implement AI or you want to do any type of project and you go well why what is the outcome we're trying to achieve and you go so perfect example um I want to I want better quoting why do you want better quoting?

Lucas 35:40

Because I want to um you know the reason I want better quoting is because I want to win more deals why do you want to win more deals? I want to win more deals because I want to uh scale my business why do you want to scale your business? Why do you it's like you keep you keep going deeper and deeper and deeper until you get the it could be acquisition it could be sell I want to sell the company I want to remove myself from the from the from that role I want to people don't go deep enough.

Magnus 36:03

So you what you're saying is you've got to ask the why question five times until you realize that you just want enough money that you just want to retire.

Dean 36:10

Yeah and change that why that that why scenario you're right in you asked the question if I if I'll pick on the the quote one again um why why do we need to focus on your quotes? Why is the quotes the problem? And whatever the answer is um because they're too slow we're missing out on deals because we don't respond quick enough why aren't you being able to why aren't you able to respond quick enough? Because it's hard to recall the information it's hard to get that information out of Jim who knows all this stuff. And why is it all just with Jim so that's it's very pressing down there. Lucas went down the path of why is the the mindset side of things I guess when I go through the those sort of whys it's let's press on where the problem is as opposed to the the feeling behind it. So Luke is very good at setting up the motivation tip behind it through that sort of why scenario. And then I will normally come into the the very problem specific um drill down on that.

Don 37:06

I'm gonna ask you some specific examples here, right? To just to give a bit of clarification on this. So um you've mentioned before that um the most common mistakes that that companies make is that jumping to solutions before defining the problem and I think that's one of the things that you just brought up. And then they they decide they need a new system runs vendor demos and and somewhere in that process the question of what do we actually need to do to get this quietly replaced by which of these platforms looks closest. Can you walk us through how that actually unfolds in a real Australian business? We don't need any specific examples.

Lucas 37:45

Yeah so when most organizations come to us they know that they they they're they want to go on this AI journey but they don't know where to start right and so that they might go well actually the biggest problem in my in my business today is quoting sales report writing and delivering a project whatever that biggest problem is right and so if they want to they they typically come with that scenario in three months from now and six months from now I want to be able to deliver XYZ faster better better outcomes it's up to us to figure out how to get that they don't know businesses don't know business leaders don't know and so how it's up to us to figure that out for them and to do that we have to dive deep into the process with the staff members that are in those organizations that are in the dude doing the doing the actual task and then pull it out so what Dean mentioned before and asking the five whys to the individual staff members that's the detail you need to understand because typically a leader is not let's call it the in the in the uh in the trenches they're not in the weeds they're not in the trenches right and so what they think how a process runs is completely opposite of what actually happens in a business usually and so you know you out there doing the the the building and the selling and so forth versus how an actual labourer does the building it's completely chalk and cheese sometimes and so getting into that part of that construction industry is you got the pie eaters on site and you got the biscuit eaters in the office.

Don 39:14

Talking about biscuit eaters mate um we need a bet better reporting system isn't a requirement right we need to know it is a requirement that's what they want that's what they tell you well here's the problem so I'm asking you for you we need to know the profitability of every project within 48 hours of completion is the real problem.

Dean 39:34

That's yes and nobody ever comes to me with that as the exactly so why do we need better reporting it's like why do you think it is so hard to get the Australian leadership teams to do it is because you still think the problem is we need a better reporting system.

Lucas 39:50

Because like because they can't they can't they don't know they've not thought about it enough they're not they've not had this honest conversation with themselves personally and they've not had their honest conversation with their team is it because the people who are building it itself is not being at the core face of it and and talking to these leaders or more it's an education gap. We have to do this because it's the only way we get to understand then what the actual true problem is to know what to solve. What most times happens is they would go to their internal tech person and go, I need a better reporting system. The tech person then makes loads of assumptions goes and builds this reporting mechanism and goes here you go and then the CEO goes that's not what I want you ask for a better reporting system they'll say it looks great.

Dean 40:36

Yeah that looks great but it's not solving the problem I've still got questions. Okay well that's that's where the problem lies and asking the the question the way you did before Don that's that's perfect. You need to know within 48 hours I might ask the question as to why do you need to know profitability 48 hours after it completes but whatever the reason is behind the business so be it. But at least now we have a very structured measurable target. I guess to your question before um Magnus around some success stories and this this can this might help to answer that question a little bit as well I'll pick on one that's that's uh in the timber supply we'll call it that um so bespoke timber solutions um they identified that if they can get answers to their customer on the spot while the salesperson's on site they can get the answer there and then for them whether it's for say fire ratings, um wind, whatever I don't know everything about their industry, all that sort of stuff, even pictures and uh samples and all that sort of stuff and imagery, then they will get close that deal more often than not. So it's like okay that's great we've we've identified that so now how do we bring that knowledge forward? So in that sort of an instance just using something like Microsoft Teams, they're already Microsoft House, to be able to bring that information forward straight to the salesperson, they can just ask in plain language I need uh Western Cedar, XYZ, whatever it is, what's the fire rating on that? To be able to give to that customer while they're there in front of them. That gets the deal closed. That's just how that particular part of the industry worked. So that's bringing that information consolidating everything being able to use AI to just ask in normal language. In the past they had to go back to a sales team that was back at head office would then spend half an hour trying to look that up get that information text it back to the person and hopefully they're still on site with the whoever's leading that out.

Don 42:24

Yeah so that was a loaded question right the reason why I asked was because I wanted to understand are we actually asking the right questions? It doesn't seem like we are and is it because in this modern society or modern future of work the decision making process is that different to what it was before or do we need to learn to ask different types of questions? Like those questions natively for construction or building industries wouldn't be something that's normal. So how do we fix that problem? How do we get the right people to the table so we can solve problems quicker and build better efficiencies.

Dean 43:04

I do feel like there's a business coaching piece in there that I feel like sometimes I'm feeling and I I feel very much out of my shoes in that sort of scenario because you're right that the construction space will ask for I need better reporting in my business. I want to know how my business is performing and to me other than them looking to their financial records and then that just brings up more questions for them why was this not profitable?

Magnus 43:26

Why where did this cost come from I agree I see it being a a a business coaching problem because I think you pointed out Lucas you've you've got the CEO and they go right this is a problem this is what I want to solve but it's that level of communication right through the organisation and then getting the buy in of all those people downstream to be able to utilize any new technology back to the mindset piece if they're not prepared to do that and the CEO hasn't got the skills to be able to convince the rest of the organisation to get On board, then it's doomed to fail.

Lucas 44:02

I think the CEO or the C-level suite definitely has the right capability to do it, but they don't have the time constraint. And so they don't have the time and effort to be able to make that happen. So we're we're coming up with a new terminology in this space called as a uh called an accountability partner. That could be a business coach, it could be an internal resource, but an accountability partner to hold that person, that team, that department, accountable to achieve the outcome within that three, six month, twelve-month period. Um, because right now what we find is you go, we want to solve this reporting. I don't know why we're picking on reporting today. Well, let's pick on reporting. Everyone wants reporting. And it used to be always being dashboards, and now the question is why are you on dashboard? It's like, well, if you just asked the question and got the answer, then why do you need a dashboard? Because you can just have a natural language conversation now, whole different subject. Um but if you had someone in the middle, whoever this person is, this accountability partner, and they were the ones holding the internal team accountable, the internal department accountable, external consultants accountable, and they just got shit done, then I think everyone wins. That's been missing for 20, 30, 40 years. That accountability partner has not been there, and it's usually gone to the business coach or gone to another exec leader that doesn't have any time and they got one hour towards it a week or something. And I think successful projects, doesn't matter what industry, will be more successful having that accountability partner there.

Don 45:21

Yeah, the other thing as well, I've I've seen this a lot being in the software vendor side is that most businesses when they're investing into capital investments, they have a lot more scrutinity than they would in a in a piece of you know $100,000, $200,000 of software. Does that still exist?

Lucas 45:38

This is like the 1% model, right? What you put 1% into marketing, you put 1% into IT, and that's just the line item and you forget about it.

Don 45:45

But is that sustainable in the current marketplace?

Lucas 45:47

I think we need to shift our mindset completely. Yeah. It's no longer a one, it's no longer this this percent line item. It's a if you're not embedding AI into every aspect of your entire business, you're not winning the race.

Don 45:58

So do they need to now start looking at um software in the same way they would look at a piece of capital?

Lucas 46:05

Yeah, well, I I I like to use the analogy that's only just come to light uh with these big firms like OpenAI, Amazon, Microsoft, Anthropic, they're all doing it. Essentially, if they're hiring a new new engineer and that new engineer is on $750,000 a year, their expectation is that engineer should also consume $750,000 worth of tokens, consumption of AI. So essentially that that role is now $1.5 million. If they're not consuming their same value of their salary, they're not productive. That's I don't know where they get the figures from, but that's that that's their mindset in this conversation.

Magnus 46:40

So just explain to me where when they're spending that token amount, what are you referring to?

Dean 46:45

So you imagine that the number of words every time you ask a question of ChatGPT, those get turned into numbers or tokens. That's that consumption. That's how they measure it. That's how they measure it. And it's that's how they eventually everybody's moving to that sort of charging model. Chat GPT, you pay a fixed fee, but you also hit certain limits, particularly in Claude, I see that probably more so. Um there's other ways that other companies use it, but when it comes to business use, it's very much a consumption. You pay for API usage, token usage. That's how that side of it gets meant.

Don 47:18

Software as a service is also trying to move towards that direction, right? Instead of per seat, they're looking at tokenized sort of thing.

Dean 47:24

Yeah, that's where I think that that whole SaaS world is getting turned upside down at the moment, given that most businesses want to it used to be all about seat count, yeah. Whereas now it's like, why do I have to pay for all of these seats when AI can just query your system, get the results? I only need one seat. So that'll probably turn to a consumption model.

Magnus 47:43

This is Molly. If you hear a bit of barking in the background, this is why let's embrace it.

Lucas 47:47

Molly's being needy. So Molly's embraced AI to the extreme. Haven't you, Mollsy?

Magnus 47:53

I've

Discovery Design And Getting Buy In

Magnus 47:54

heard have you guys structured some process that's called discovery and design?

Lucas 47:59

This is what we've understood to try and solve these problems that we're talking about today, right?

Magnus 48:03

So maybe can you explain that for us?

Lucas 48:05

Yeah, so does the discovery and design process, or now we've just introduced a new one in the last six weeks, which is Discovery and Design Pro. Um, how do you take everything we've been discussing to get a real world outcome the business is looking for, right? And so, as an example, if you were to dive into say you come to us and say you want to solve this, you know, you want a renewal reporting system, or you want a new quoting system, or you want a new scheduling or resourcing um uh better process, whatever, whatever frustration you have in your business, our job is to dive really deep into that frustration, analyze it, and then figure out how to solve it using AI. I'll let Dean get into the weeds because I'll talk about the from a high-level point of view. The high-level point of view is the usually the C level suite, the business owner, wants an outcome. So I have a really good strategy conversation around the outcome and really dive into why do you want even want that outcome to get to the nitty-gritty because usually they don't know. Once we understand that outcome, then essentially I handball it over to Dean and the and the developments, development team, and then they actually figure out the real root cause of that outcome and then build the problem. Build sorry, build the solution to solve the problem.

Dean 49:16

Solve the problem. So I guess it's it's a very structured approach to those why questions we were asking before. So being able to sit with the individuals that are uh actually doing the work or involved in the work that's required. Um, so not just taking the on the face of it what the executive sort of management might have said the problem is, to go, okay, well, what's it look like at the ground level? What actually comes in from a customer? What are we, what do we have to deal with day-to-day? Not perfect world. We need to be able to cover all of that because that's often where things can fall off the rails. And often that's where a little bit of the the hype, if I call it that, in the AI space is AI can solve this for you. It can if it's perfectly consistent and exactly received in an expected way, but business doesn't run like that. Yeah. So this um discovery and design process is about understanding well, what's everything that you've got to go through and how do you deal with stuff falling off the rails now? How do we deal with the exceptions? Um, and it's really about documenting a lot of that, bringing that forward and mapping it out, making it quite visual. We've go through a process of making that very visual for both management and staff. Going through that actually brings those staff on board. This is this is everybody from management right down to ground level staff because they understand then that hey, this AI, it's not here to take my job. It's to take away this bit of frustration that annoys the hell out of me every week. Because I have to do this repetitive thing where I've got to follow up the client 15 times, et cetera, et cetera, because it's missing this bit of detail. So we that's really what we're trying to target. Um, and as part of that, we often find out and kind of educate, I guess, the staff around what are they really like in their job and what do they want to hold on to? Let's not take that away with AI. Instead, let's just foster that.

Don 51:03

So, on that, um, you know, there's this perception mainly with the the accounts teams or CFOs that with AI coming, everything should be now cheaper. How much of a reality is that in your experience and in your opinion?

Dean 51:19

Everything should uh okay. So, yes, that side of it, that that world is changing quite a bit on our side, and we had to uh a few weeks back um essentially rethink our business model a little bit around that. Previously, developing out anything, AI automation, anything that required development was really about hours, development hours, and that's how it would be priced, that's where the revenue came from. Now, because a large portion of that can be done by AI, it's no longer a 16-week give it to a dev, they sit in a dark corner and punch out code, but it changes to, well, now we need to understand how do we get AI to be consistent in this space, creating better AI prompts. That's where some of the prompt engineering sort of scenario and that terminology has come into play. So it's just a shift in where the cost goes. It has come down overall, but it's not a complete just cut that line entirely, and it's raw savings off that.

Lucas 52:16

But what we are seeing is a lot of companies are now wanting to bring staff members in their business to do a lot of this uh development work. So rather than going external all the time, can we now bring a staff member internally and then can they hire us to be advisors and guide that staff member to get the outcome rather than just uh always going external? And that was never a thing until two months ago with barcoding. Because now staff members can do it themselves.

Magnus 52:39

So

On Site AI Voice Vision Privacy

Magnus 52:40

the guys actually on the ground, on the site, the site manager, the various tradespeople, how does any of this AI technology help them?

Lucas 52:49

Uh we're not seeing it on the ground at this point in time, we're seeing it in the back office right now. Well, that there is a huge opportunity in that space, if I can say it out loud. The unions are stopping a lot of it. So as soon as we talk about as soon as we talk about uh on the site, people video, photos, unions get involved. They don't want the privacy.

Magnus 53:13

So efficiency, inviciencies that involve less people, that's an industrial relations issue.

Lucas 53:19

Yeah. Okay. Yeah. And so I think that will shift. I think there's been a big movement, especially with the Olympics coming here in Queensland, and essentially they're saying unions have to stay out of it. So I think we'll see a big shift here. Um, but that's what's held back from the ground level perspective until now, until recently, anyway. Um what we're seeing is how can they use not even AI? AI, uh sorry, um, it could be AI in the background, but essentially applications on their mobile to do a lot of things. So could they do handover documents or could they take take site photos and upload those site photos into the back end of the system? This is not AI, this is just technology. So it's like, well, the AI, the buzzword, is now enabling the conversation to be a technology conversation to go, well, why are we still doing paper? Because a lot of construction sites on they still do can use paper.

Don 54:09

But it can it can automate a lot of things, right? In the past that it couldn't do. Yeah. Where now with AI on the back end, it just allows you to do it. And also the the maintenance of those apps and the build of those apps becomes a lot easier. Correct. But when you say it's not in the ground yet, I was quite surprised because that's what at the end of the day that has the biggest efficiency gains. I think it would be it'll change soon.

Dean 54:31

I think um I guess from some of the more let's call it practical projects that we've done for people on the ground, it's been largely around the admin stuff that they generally suck at. Yeah. Call it that. Taking photos aside, proper photos, uh, good photos. Doing time sheets, for example. Like there's a lot of time sheeting sort of stuff out there, but why not why do you have to open an app and punch in numbers, fat fingers, build construction fingers on a phone? That's a bit of fun. Why not just hit the button and just tell it, hey, throw in a I I started here at 6.55 and going home at 2.05. I'm at whatever job I'm at. The end. Let AI do the rest in the background. That's to me, for most construction people, that's going to be a lot more natural. They're not going to push back on that so much. And it didn't, it's not as frustrating for them in the in most cases.

Don 55:21

But even having conversations, like you can record a lot of conversations now and then turn them into actionable items where that's I think where you know I just say unions and and a little bit of privacy as well that comes into play then. But but when you look at things that these industries are not good at is the administration side of it, which is you can have all those conversations, you can do all of that, then could be building actionable items out of that. It's not just a construction industry problem, it's a problem with everyone, right? Yeah. And if AI can make that more efficient, then it's just a it's just a legislation.

Lucas 55:56

Right. So as an example, is um let's call it timesheading, scheduling and timeshed. The huge opportunity I see in the in the especially the tradey industry. Where I say a huge opportunity in this space with I guess the construction industry is with the tradies, the tradies that are always on site. So we we work with a lot of plumbers or sparkies or um, you know, the they're always in their youths, they're always traveling from site to site. Where there's a lack of technology is with these, the actual people on the ground. They're still doing it the old school way. Maybe they might be using an app here and there, but what I call it the old school way, where we can automate the entire time tracking scenario now. These apps already have what we call open APIs, where we can integrate, say, AI, we can pull data from these apps, we can pull GPS locations. We know when a trader goes to a site, they're clocking on for that site. When they leave that site, they're clocking off for that site, that time in the between. And then we can also integrate AI so when they're leaving site, they can have that quick conversation. I use this particular material on site, I use this equipment, whatever it might be. Because the number one thing we hear all the time from any tradie, timesheets never get done. We don't know what we're invoicing, and we also are not invoicing materials that are being used on site. That's all lost revenue. That happens at every trade organization, nationally, globally. That's like such a quick win in my mind. Um, I don't know why they're not embracing uh embracing that that opportunity.

Magnus 57:21

Yeah, well it's interesting. There's a there's an app, I won't mention which one it is, but but my wife and I both use, and that's understanding where each of us are at any given time. And that just stops us having to pick up the phone and ringing each other up, going, Are you still at work or are you here, are you here? Um, a lot of my mates don't like that theory, but I don't have anything to hide as far as where I am. But I think but what I'm getting at is that that app has even got that technology that I spent this and I look on there and it says my wife's been at this thing for eight hours and twenty five minutes. So, and then it's can does your travel time that okay, you've driven from there to there. That's almost you give that with your employees that some admin girl sitting in the office and you know that he was on this site from that time to that time because he stopped, then was traveling for this.

Don 58:09

It's all done. But I think that's where the industrial relation issue comes in, right? Like everything from collecting voice data to collecting location data, then that that becomes a privacy issue and a lot of flags kind of warrants that off.

Magnus 58:25

I was going to say, well, there's there's the majority, the vast majority in Queensland, for example, are non-union. So you've got the whole residential sector, none of those guys are involved in the unions. We're dealing with those top-tier firms. Yeah.

Don 58:37

But I think now we're talking about the downside of the social impact of how AI could have an impact as well. So it I think it is a can of worms, but um there there is a lot of technology that can bring efficiency that may cross that boundary off or push that boundary off that gray area of the internet.

Lucas 58:56

Oh, I say we're a technology company. We'll tell you what's possible from a technology point of view. Yeah. It's up to you from a legal compliance, risk unions. It's it's up to you to figure that that landscape out. We'll say this is what's possible from that point of view. We'll give you insights of what we've heard in the industry, of course. Um but you know, it's a gray area in most cases, so push the boundaries if you can.

Don 59:17

Dean, what do you see as the most effective AI piece of AI or automation that's coming?

Dean 59:27

That's that's coming. Okay.

Don 59:28

So even Well let me let me let me let me ask two questions. What's out there right now and then what's coming.

Dean 59:39

I guess I'll I'll keep it focused to the construction space because I think every business will have uh different levers that they would want to pull on in that sort of area. I would think anything to do with voice for me feels a lot more natural in the construction space because you're not having to worry about fat fingers, as I said before, that feels a lot more natural. And the more that you can do out in the field just by talking to something and getting um speech back. So both ways, not just talking to something, but having it talk back in a language that makes sense to a construction person, the instructions, whatever questions that might come up, to me, I think as of right now, that is quite strong.

Don 1:00:19

Yeah.

Dean 1:00:19

Um, I think what will probably come up in future, and it's still a little ways off, will be the vision element of that. And I'm gonna put the whole EBA stuff aside, but um, to be able to say, for example, take a photo off-site or show the video off-site and to understand that, yeah, the scaffold's been put up in the right way, or no, there's no safety concerns across that. That's that's really elevating safety, that's really elevating um efficiency for the team to make sure they're building the right things in the right spots or whatever the case may be. And I think it'll feed into another real major pain point, which is swims that need to happen across a lot of sites. That's a that's going to be a big thing for them as we start to see that flesh out.

Lucas 1:01:00

Yeah, think about robotics, right? Think about like having a robotic dog, or I say dog, but whatever item that might be, with cameras like a rover going around your construction site with live video footage that's bringing back to your head office and it's been analyzed on everything Dean just touched on. That could be 24-7. You don't have humans walking around trying to do checklists. Technology can do it. You talk about drones, but talk about like robotic devices uh doing that. It's a huge opportunity in this space, but I think yeah, us as Queenslanders or even at Aussies, we're pretty far behind when it comes to compliance and regulation and risks uh associated with it.

Magnus 1:01:35

If we get a construction business owner listening to this episode, what's one habit do you think they could change that would be the biggest game changer for them?

Lucas 1:01:43

Can I say that out loud? Yeah, change your mindset. Change your mindset what's possible. Just because you've used that spreadsheet for 30 years doesn't mean you still have to. If you change your mindset to go, if I was starting this this company today, what would I do differently? Ignore the legacy, what I'm what I've what of the the heartache that I've gone through. If I start this company again today, how would I do it differently? And the only thing I think what we miss as humans, we always forget to ask the question from the customer. What does the customer want? What the what outcome are they looking for? If that's the outcome, how can we get to that outcome faster, better, cheaper, more reliable, more consistency?

Magnus 1:02:18

So I listened to a podcast last week and you just reminded me of of it then. Is for the people you mentioned law firm earlier. And if you were you've got that law firm, you've got that codec camera, but you've got somebody young who's just starting up a business where the focus is the business is built around AI, not built around an existing uh structure that's not prone to change and just how they're going to just overtake the industry. Where I'm going with that is I'm wondering from a construction perspective, if uh how that level of technology, being construction such a doing on-site type activity, that there'd be some advice out there to some younger guy at 25 who maybe just finished his apprenticeship and and can actually wanting to build his business based on AI. What would you say to him?

Lucas 1:03:09

We're seeing already 25, 30 year olds where they they they think differently, right? They don't have that legacy of just because we've always done it that way. Um because there's a lot of compliance in the construction industry, they're going, Well, I don't, I'm here to be a builder, I'm not here to do compliance. So, what you normally find now is you normally have all these compliance officers, and then you have all the HR team, then you have all the project managers, and then you've got some people on site building. But all the back end offices actually just ugh, you don't need it, right? Because all that can be AI agents. So if you go, well, I'm here to build you a house, a townhouse, a story building, a complex, uh, whatever you or whatever you're there to build, you're there to build. So any back-end administrative work, which is all administrative work, everything we just talked about, is now AI. You don't need any back-end people at all ever again in life. That's how the new generation are thinking. The other older school businesses will get there eventually, but it's gonna be a three-year journey to have that mindset change.

Jobs Agents And The Future

Don 1:04:09

So, on that, let me ask this question from both of you. Which industries and which jobs do you see displaced in the next year and the next five years?

Lucas 1:04:21

Well, I'm definitely not leaving school or coming out of university wanting to be an admin clerk, that's for sure.

Don 1:04:26

Or a lawyer. Oh, well, I think lawyers are still gonna have a place. Someone's gonna I think they're gonna have a I think they'll have much more productive. I think this is where the question I'm asking is mainly because I think there's still gonna be professions where AI is gonna assist them to be more productive. I'm asking the ones that are gonna be displaced. So I'm I'm saying a radiologist is is going to be, I don't think they're ever gonna be displaced, but they are going to be able to do more work and give us much better results. Same with a lawyer, right? Like you're gonna have to have that human element still there. They're probably gonna be able to read a whole bunch of stuff. Like, you know, we the way you see a lawyer is you've got a whole bunch of interns that are reading all that's gonna be a thing of the past. But where is where are you where do you guys think the true displacement where a full AI agent takes over that role?

Dean 1:05:19

So you think it was admin, you're putting towards that very administrative clerk role. I would I'm if I'm thinking industry specific, I think marketing's gonna be one of the big yeah it's anything that is heavily digital, um digital delivery in particular, um, comes away from it. I think they're gonna get hit the hardest. Um in terms of legal and that, I did want to I guess the thing that's they there is ripe for disruption there. I think the biggest hurdle, and whilst a lot of lawyers and people that own those sort of firms like the idea of these uh models that are coming out, the legal ones, they haven't really thought, well, how Think about how they bill us. It's time. So what happens when we reduce the amount of time that it takes to now do a particular job?

Don 1:06:17

And it's the same for consultants, right? Like you can do deliver the same consultancy project in in a weekend that used to take me a weekend, I can do it in two hours.

Lucas 1:06:25

I'll answer your question and I'll let you jump in, Magnus. The answer to that question there, Dawn, is anyone who's not embracing the AI tool set, whatever that that tool set is, will lose the race. There will not be a job for them. But the new generation coming in, they're creating new jobs, new responsibilities, new ways of working in whatever field, whatever industry, because they're embracing the technology. Um, so I wouldn't think it's I don't think it's a an industry, it's definitely job roles, 100% are being displaced. But entire sectors are going to have to rethink how they even um perform that particular activity because that will be gone.

Magnus 1:07:03

What I was gonna say was when I look at I look at it as a hybrid type model. So when you look at law, and I completely get what you're saying as far as the time goes, I get a friend of mine that does a lot of tutoring, right? And they say the teachers are never going to be redundant because you need that personnel, you need that one-on-one. Not every parent can afford 80 bucks an hour for their child, but when you've got one hybrid model where that one teacher is doing three kids, or you've got the AI to be integrating, helping one kid with their homework, you as the tutor, instead of them going off doing AI, you charge 10 bucks an hour and you build an AI agent that specializes in year seven mathematics, you can then charge that out at 10 bucks an hour. You're still charging out the individual. So same with the law firm. You've got some issues with go, you'll take this through one of our AI agents that the law firm builds, they charge that AI agent out of 50 bucks an hour and not 700 bucks an hour for the lead legal guy. So you need you need both.

Don 1:08:05

It still becomes really interesting how you price that, right? Because it is gonna, yeah. I think that's gonna be a big industry challenge. Yep.

Magnus 1:08:12

But I think I think as a law firm or as a tutor, you need to have both. You need to be able to offer both. If you're not embracing any of the AI, you're gonna get law firms that are just gonna be set up as AI where you go to them and you're gonna get this price. I I to me would be able to.

Don 1:08:27

I don't know if that's as easy as you think, though. How do you change a mindset of a law for law firm that charges 300 bucks an hour for the cheapest uh uh person in their portfolio now saying, oh well it's gonna be fifty bucks for a huge drop in revenue. That's right.

Lucas 1:08:42

And that's where it's gonna be value-based pricing.

Don 1:08:44

Yeah. I don't think it's it's it need it can be an early uh based and that's gonna be the fundamental.

Magnus 1:08:52

In my experience with law, it's just legal, but yeah, it's a but it's a sample. It's a difficult one for a result, because the result is is isn't as simple as it's it's just this to be able to get there because there's so many hoops. And you mentioned compliance, like the construction industry, I don't think people realize, is the most heavily compliant industry out there. It's more compliant than medical. And that the more automation you have with that, the better.

Don 1:09:17

That story around the value of consultancy, right? There's this story that I heard years ago where you know they hired a uh at a rig, they hired a consultant. He came in, he looked at the problem, and he said, Yeah, hit the nail on there, and then he left. Yeah. And send a bill for you know 1,500 bucks, right? And they said, 10 grand, let's say, right? It's like you spent 10 minutes, how do you justify $10,000? He said, Well, if anyone else came, they would have spent three days looking for that nail. I saw that nail and that's experience. Yeah, I think that's the mindset that I think that's gonna need. It's going I don't think it's the mindset of the service provider, it's a consumer that's gonna have to change because someone who gives you an answer within five minutes and charging you ten thousand dollars. So that's gonna be the trick. I think we're gonna run out of time, but I do want to ask your thoughts a very quick answer on agentic AI, where agents are valuable right now and and where you think they're gonna be valuable in the future.

Dean 1:10:23

So that agentic AI, I guess for that's really the practical implementation. We're we're moving away from the toys that is just talking back and forth, yeah, to it doing something for you on your behalf out in the real world. Um and they can be very small, um, they can be very big in terms of what they can do. But typically what we're seeing more of in that and where the real value comes from is when you have multiple agents all working together to get uh a bigger sort of outcome. Um and something that's often overlooked is having agents that can check on other agents and things like that. That's they're all that's where I think the that's where the bigger play is coming into to the market now. How far are we from that? Well, you can kind of see it now with the likes of say OpenClaw and Clawed Co-work. And I know uh OpenAI has one, and I can't remember the name of it, but it doesn't matter. Um they're kind of already in play now, but they're very much that sort of retail-facing version of this. Yeah. Um, you can certainly build your own and you can use the AI tools to build their own agents. So that's where some of this is is getting really exciting.

Lucas 1:11:31

And Microsoft Copilot has the Copilot agents, but then they also have Copilot Co-Work, which will be released in the next few months.

Don 1:11:39

In your opinion, what has been the most exciting um technology implementation over the last 12 years for you guys?

Lucas 1:11:49

So I couldn't I wouldn't be able to pick one particular probe, but I'll give you some examples on a few, where staff members uh start actively using the AI that has been custom-built for their business with their own data set and getting real live outcomes. So the example Dean gave before, we released this one for this um this timber company maybe September, October last year. For four months, crickets, hardly any use. You look at the results, you look at the thing, hardly any uptake, right? Fast forward to February this year, and the uptake just went from hardly anything to like massive spikes, and you talk to the the individual users, the the salespeople, the individual people using these uh the technology, they're like, I'm giving the answers to the the builders or the architects on the spot. Well, yeah, that's what it was built for. It's like like their eyes just light up and they're just like, oh, this is a thing. It's like the customers are now really happy. I'm getting deals, I'm winning more deals. Like it's those moments where it's a bit of a takes a while for it to happen. But I think once it does happen and you see that light bulb moment go off, you're like, okay, now I see that's like that's the true value. And you go, Oh, you feel so good that you're like, oh, finally, people are embracing this. But uh, society is really far behind in embracing technology, right? So um uh yeah, we've solved some really cool, crazy things over the last um 12 years, uh, but it's not about the cool crazy things that we've solved, it's the use of the technology and what the people are actually using it for and on on the ground is there. That's it's pretty exciting.

unknown 1:13:20

Nice.

Quickfire Takeaways And Closing

Magnus 1:13:21

Dean Dean's got to go. It's been an absolute pleasure. We're just gonna wrap it up with the quickfire questions for for Lucas. Thank you.

Don 1:13:27

Thanks for joining us, Dean. Um, he's got to go ahead and solve problems. Yes, I do. Straight into it, indeed.

Magnus 1:13:32

So quickfire question time, Don. Go for it, mate. Lucas, you ready? The most expensive lesson you've learned the hard way.

Lucas 1:13:43

Uh I'm always too early. I don't know if I can say that out loud. Um uh we've been I've been too early in so many industries where um invested too early, down in the wrong industry too early, try to launch too early, the the world of society is three, five years away.

Magnus 1:14:03

What technology on the market today do you think genuinely is overhyped?

Lucas 1:14:08

I'll say AI is pretty over overhyped. Uh underhy is robotics.

Magnus 1:14:14

I was gonna that was my next question. What's what's underestimated right now at the moment?

Lucas 1:14:18

100% robotics because people are very much arm length of it right now.

Don 1:14:23

You really think so? You you think we have the you don't think we're looking at robotics in the wrong way?

Lucas 1:14:29

Trying to solve a problem with a human uh you got a humanoid, yeah, you got obviously got the humanoid robotics, but I think what else can it do outside of the humanoid experience? And there's so much opportunity because I I I I travel around the country, I speak to um business owners nationally. The biggest problem most manufacturers or warehouse uh organizations have logistics, they can't find staff to do the warehouse jobs. So if you can't find staff where you need something, you need someone to do it, robotics can do it for you.

Magnus 1:14:59

The worst piece of advice you've ever heard a vendor give a client.

Lucas 1:15:04

I hear it all the time. I heard this last week. I said, Oh, can I see your roadmap? This is to a vendor, I won't name names. Can I see your AI roadmap? Oh no, we've got 20 years of uh manufacturing experience. Uh, you don't need to see our AI roadmap because we know exactly what what how to solve the problem in that manufacturing plan.

Magnus 1:15:20

A book or podcast that shifted your thinking in the last 12 months.

Lucas 1:15:27

Last 12 months? That's a that's a topical one. So my main podcast that I go to every single day is the all-in podcast, but it it has become very political um the last little while.

Don 1:15:38

I agree. Um tough one to listen to these days.

Lucas 1:15:42

Yeah, it has become very one-sided.

Magnus 1:15:45

Let's let's take it back. What about a book that's had the biggest impact of your in your life that you've read? Rich dad, poor dad. The one thing you wish every director listening to this would actually do tomorrow.

Lucas 1:16:00

Take responsibility and stop delegating.

Magnus 1:16:04

Final question finish this sentence mastery is not just knowing, but doing.

Lucas 1:16:12

Well said.

Magnus 1:16:13

It's been a pleasure, mate. It's nice to have you on the opposite side of the table for a change. Really I think that what you're doing in this AI space is very, very exciting. And I don't think you're too early on this one, mate. I think you've time to think you were too early back in 2014, but I think you're on the money now, and um people just got to start using it.

Don 1:16:34

Sam mate, any any entrepreneur you need one one to work, and I think you're you're on to it this time.

Lucas 1:16:39

So um thanks for being your second favourite podcast host.

Don 1:16:44

Until next time, Don. Thank you. Thank you, Magnus and Lucas.

Magnus 1:16:48

Been a pleasure. Thanks, boys. Hope you enjoyed this exciting episode of the Mastering Podcast. If you got value from today's conversation, hit that subscribe button now and share this episode with a friend. Until next time.