0:00

Two to three times the human traffic you've got on your website is actually AI. So if you look at it, you've got one third humans, two thirds AI. That's humans sending their AI agents out to understand your website. If the models can code, then they can code themselves, and then you just see a growth cycle of, of the models just self-improving until they eventually get to a, a really higher accuracy point. We're in one of the greatest times to start a business, and it's even better for those who have expertise because the value of expertise has never been greater. If we have one life, I would hate for at the end of that life to look back and go, "I never achieved the things I wanted to achieve." So just do it Welcome to The Business Game, the show that breaks business down into real levels from zero to over fifty billion dollars so founders can see exactly where they are and what comes next. Today's guest is playing at level one, which sits in the zero to one million dollar bracket, the founder startup and survival stage. At this level, you're proving the model works. You're validating demand, refining positioning, and fighting for consistent revenue. The challenges? Doing almost everything yourself, hiring your first team member, and stepping from practitioner into CEO. The goal? Turn expertise into traction and build momentum that's repeatable, not accidental. Welcome back to The Business Game. Our guest today is Huxley Peckham, an AI whisperer and tech communicator from the Sunshine Coast. He started out with Web3, made the jump to AI, and now runs Tech Horizon Labs with a product called Are You Found by AI, a business that shows companies what the AI tools are telling their customers about them. Right now, Huxley's playing the game at level one. Huxley, welcome to the show. It's great to be involved. So maybe let's just start with a little bit about yourself and, um, how you got to where you are. So if I was to start from the beginning, uh, I think when I was very young, it was always curiosity that drove me to experiment with, uh, the computers that were around me- Yeah um, and understand them. I took apart plenty of consoles, um, and definitely mo- lawnmowers as well. Yeah. Different machinery, uh, electronics, and playing around with it, and yeah, I think that there was a point where I started to see certain, certain bottlenecks, you know, in that interaction. One of them definitely became a bottleneck around value flow, which led me into Web3. Okay. But, you know, the way that money was being transferred at that time, I felt could be faster, you know, and could be cheaper. And then the same sort of thing happened with AI, where I felt that the keyboard and the mouse was our final limiting factor. Yeah. Um, I came from a background in writing code. So I, you know, write Python code, and that was, um, a- one of the most painful experiences of my life. Yeah. Um, you're just, like, constantly battling the code, um, to get it to work. And then, yeah, it... Then as I was experimenting with- you know, the code and then seeing how AI could write the code when ChatGPT first came out. Um, I was experimenting the day ChatGPT launched and was just kind of blown away. Um, I think if I went back to that point now, I would be less blown away because the models and how they've advanced now. Yeah, yeah. Um, but it, the experience of like seeing something that feels like it responds to you and you could kind of talk to something, yeah, it's, uh, it, it blew me away, and so that was... You know, I saw a heap of things to be, you know, a heap of areas in the business that I was working in that could be optimized by that. And, um, so we went about automating the business, and then, and training people on how to do some of that themselves. Yeah. And then at one point, you know, there was, there became a clear divide on, you know, w- where, where should you automate, um, where does the, you know, where does the human need less and less involvement in? And then, then it became more clear where the human needed more involvement- Yeah where it actually needed to be involved in the decision-making and the review. Um, and that's what sort of, that's what urged me to get more involved. I think I went down that rabbit hole. I've been in, you know, obviously experimenting with AI models of some form for- Yeah over a decade now, so it's not- Yeah new to me to understand something that could be slowly brought to a higher level of accuracy with, um, continuous review and more data being exposed to it. But at one point, um, it was clear that there's also a workflow that you need to build around the AIs so that you can continue them, like continue the accuracy- Yeah um, improvements. So that became pretty clear to me that we needed, we needed to do, you know, more training. We needed to become, um, embedded in the team and embedded in the organization, and then I branched out from the company I was on at the mo- at that point- Yeah um, to start my own, uh, business. It was about two years ago. Yeah. Um, and I, I, I really thought about how to position this because, you know- I spent a long time thinking that this space is really fast Yeah You, you can't jump in too quick now without it sort of maturing, but I need to jump in now. So, you know, I know that I need to be in this space- Yeah and in the frontier and close to the frontier, as close as I can, and I've eventually now gotten really close to the frontier- Yeah um, where we're partnering with the Frontier Labs. Then going, how do we position ourselves from the start so we could s- take that journey, um, without being disrupted, you know? Yeah. And that's, that's, that's proved to be v- been really successful, but it took a long time before anyone realized what we were doing was the right approach. So, so what made you, what made you think to yourself, "Hey, look, I'm in Web3, AI's kicking off- Mm-hmm ChatGPT was starting." I think for most people they were kinda like, "Is this gonna be..." You know, there was lots of people talking about it. It's the ne- it's the next internet, it's big, blah, blah, blah. Mm. What, what gave you the confidence to know that you'll make that move? Yeah. It's, technology has such a hype cycle where, you know, and we experienced that with Web3 when we first got in. Um, you know, I could see it being really useful for global, uh, global payments. Yeah. Right? But then everyone sort of went out to every other use case they could think of that could be possible in the next 50 years- Yeah and maybe half of them didn't have the infrastructure there to, to, to fulfill most of those use cases. Um, when I look at the architecture of a system, and that's, you know, something that I've become, um, I was very good at when I was a kid, and then now I, I've, I've really honed in on that and started to look at the bigger picture of, do we have all the moving pieces for this- Yeah to move forward, or is this tied to one piece that is gonna be our bottleneck? Is it, are we able to have like a modular system, a modular architecture so we can move the pieces and keep moving forwards? And what I found with Web3 was that there was certain use cases that were just, you know, never, they were never gonna go away. Yeah. Um, global payments, global, like cross-border payments, um, you know, holding of, of assets in like escrow, like those sort of things that they haven't gone away. Yeah, yeah. But we've, we've gone to the hype, we've gone through the hype cycle of like let's explore all the other things that could be possible and let, let's pitch that idea and maybe it'll happen. Yeah. Same thing happened with AI where I saw it early days evolving in computer vision models, and that was becoming, you know, we're getting a high level of accuracy there. Yeah. Um, but then you started to see- Like, like what really brought me into AI and, and like looking at the f- future of it was w- when you dial it back to the fundamentals and the first principles of it, um, you've gotta understand that it is, you know, it is non-deterministic, so it's, it varies. And so you've gotta build deterministic systems around it. Yeah. And the, the real use cases there that started to emerge were, um, were clear to me that- If we get that right potentially in the next year, um, r- particularly if we were to use it to do automation in a business- Yeah we would need it to code. Yep. Right? So that was a, that was a very simple idea, and also the sim- the other simple idea that we thought is that not only does it need to code, but it needs to code with a very high level of accuracy, and almost n- everyone needs to be able to do it. Yeah. Everyone needs to be able to access that, and if you can get to that point, then the models also self-improve. Yeah. So if the models can code, then they can code themselves, and then you just see a, a h- a growth cycle of, of, um, of the models just i- self-improving until they eventually get to a, a really high accurate, high accuracy point. So I was very early on the coding agent- Yeah um, stuff. Yeah. And I was preparing businesses for that because they weren't, they weren't, um, sure that coding agents were gonna be a thing, so you know, they were doing all these other automations at the time, like, let's do this, uh, this workflow mapping, and let's map where the data goes from one site to the other. That's still kind of useful, but the technology underneath it's changed. Yeah, right. And so that's when I was preparing for coding agents to become, you know, more widespread- Yeah like months before it even happened. Yeah. And I was telling e- all my clients, like, "Don't pay me to do an automation if it's doesn't make sense for the business." Yeah. Like, if it's gonna cost too much and it's not gonna deliver anything within three to six months, I would hold out. Yeah. Like if the ROI- Yeah is not there within that timeframe. So, um, that was something that if we, if we had those foundations right, we could build an, like we, we could build these in powerful companies with their own custom software that was personalized to their needs, to their use cases. Yeah. And that for me felt really powerful, and I, I thought, you know, that's, that, that's something that we need to, to race towards. Um, and then since then I've actually built a product that plugs into coding agents- Yeah as a tool. So let's, let's talk about the product. Mm. Yeah. So, um, so that's, that's the, the... Is that y- your only product that you have inside your business, or is that your new launch? I find it hard not to just stay up at night and build new products. Yeah. Um, there's one that I focus on, and that's the, the main one I dedicate all my energy into. Yeah. Um, it's very hard not to spend two hours and suddenly spin up another product, because that's all that it takes sometimes- Yeah to get the first version out. But then to roll it out and get users, completely different- Yeah um, process. Like we've solved, uh, the execution of product. We haven't solved the execution of marketing yet. Yeah. Well, well, welcome to the club. I think that's a- Yeah. that's a, a big thing for any business. Yeah. Doesn't matter what size business you are, it's, uh- Yeah if you can get the marketing engine- Yeah going, then, uh, that's something itself. So- Yeah so, well let's talk about, uh, Tech Horizon Labs then, and what, what- You know, what would be, how would you explain what your, what your business does? So the best way to explain it is we are... Is it- I'll, I'll get to a more technical version of it- Sure that I will try to break down first. Yeah. But the best way to explain it is we are an AI implementation partner for businesses- Mm and we offer with that, uh, training and, uh, software services. Okay. So we are, we have productized the training into packages, and, uh, we've made sure that that training, um, you know, helps build internal capacity for a business so that they can, uh, start to experiment with these tools. And then when they want us to be involved more in the, the s- the scale of that- Yeah then we, we partner with them and help them roll it out to the rest of the team. Okay. Um- So what's an example, what's an example of what those, what, what that would look like? So you'd come into a, what, what size business are you dealing with? Typically, the businesses that we really partner with right at this early days were up to 20 employees. Okay. After that, we are working with enterprise now with, that are 50-plus employees- Yeah in a different capacity. So it's more of an, um, we're doing, we do come in and do a sprint, and we go, "What are the things we need to fix to improve your website- Yeah and your visibility?" Yeah. Um, and so we plug them into our data sets, and then we figure out what the roadmap will be for the next month. Yeah. And what's the highest priority fixes that we need to get started, and we, we go about pushing for that change. So wherever we need access, uh, we'll get access to their website, we'll get access to their analytics. Yeah. Um, initially we'll make sure we've got all the analytics there so that we can make, uh, informed decisions on where the highest impact, um, options are for them. Yeah. And then we go about fixing them wherever we can. So if there's a point where, um, you know, we just need to consolidate the, you know, the entity and how they talk about themselves- Yeah which is one of the big problems we're seeing at the moment, is that people love to use three to four different names, and they spell their business name differently- Yeah every time they talk about them. Yeah, yeah. Um, and so that's something that you've gotta get consistent across the website copy. The other side of that is then embedding machine-readable texts into the website that also consolidates that entity. Yeah. And then that's a process of what is your tech stack? Like, is your tech stack actually able to provide us with the access to do that in a reasonable time without it blowing out this 30-day sprint we're trying to do? And if it doesn't, then we have a conversation around, um, like, you know, obviously I've s- had situations where people go, "Oh, just rip out the tech stack now, and let's just bite the bullet and do it." Yeah. But sometimes it's, you know, it's good to have some hands-on experience before you, you do that, and you, you back it up with, "No, actually, this took- Yeah considerable amount of time with your specific website editor." Everyone has a different setup. Yeah. Everyone uses website editors differently. They have different tools they plug into it. But with that spec- with each spec- specific business that we work with, we would look at- This is how long it took to even get access to it, and this is how long it took for us to then modify the settings so that you did have a machine-readable version of your website. And by that it means can agents go to your site and confidently understand what you do- Mm-hmm across all your products and services, and then r- include your name in an answer that someone might be requesting on, across, you know, across any of the seven major engines? Yeah. Um, can it confidently include you in an answer that might be a broader query or one that's more specific around, like, your business if they're doing due diligence- Yeah, yeah in your business? So we're optimizing websites so that, yeah, um, you know, they look nice for humans, and we're also optimizing them so that they look nice for AIs. Yeah. They can quickly understand what you do- Yeah and, uh, they can quickly navigate and even use your website. Yeah, 'cause I think that's, I think that's a big issue for a lot of businesses too, right? Mm. I mean, we're, we've, n- you know, one of our businesses has been guilty of that as well. We have different, uh, different names out in the market- Mm and different products and different spelling and- Yeah and all that sort of thing. So, um, so essentially what you're doing is taking that, making sure it's consistent- Mm-hmm making sure that AI can read, you know, in, in layman's terms what it needs to find. Mm-hmm. Um, so- Exactly so, so how much of your business is, is that piece, and then how much of it is, you know, your new product? Yeah, we, we started out with training and we realized that with ChatGPT, Claude, they were out there trying to understand the business. Yeah. So you, it was either a process of, um, well, we decided then that with TechRising Labs, we're that partner, and we really do the done-for-you side of the, you know, website building, systems building, software training, you know, that side- Yeah of it. We take that off people's plate. Yeah. Then we went, "Well, w- we're actually finding that, um, we've built this agent that does an audit of their website." And so, um, you know, and we f- find the issues because f- find these issues in their website that weren't exposing the right information even to their own agents internally. Right. And, you know, you can do, uh, voice memos, you can do recordings. Um, you know, you can get that context from people's heads and get them to talk about that. Yeah. And that's one part, but then putting that on the website and making sure that everyone on the team also has access to that, and everyone, even their clients, have access to that, that's when we realized that there was a huge market for making, uh, you know, going through the web and making it readable for agents- Yeah and usable for agents, things that pe- agents actually want to use. And that, then we built an agent through that process that did that, and then from building out the architecture of that agent, um, we actually distilled it down into a, a, a smaller version- Yeah that could be exposed into a website. And then that became now, um, it's a, it became a large part of our business because it's not only a service we can just offer at a much lower cost. It's actually free to start the website, to start using the website- Yeah and connecting your Google Search data into it. Okay. Um, but then, you know, it became a, a process where people would use the free version, and then they'd come to us and say, "Can you just do it for us?" Yeah, right. And so it also became a part of our, it was already a part of our higher touch point funnel, but then it became a part of a broader funnel that we then, uh, were able to give people some value upfront and- Yeah um, show them a bit about w- what we do. And a lot of people actually just use the product and plug it into their coding agents, and it just optimizes their website for them. Yeah, right. And so they love- They love that process. Um, but yeah, in terms of how much of our business it is right now, we're still mostly on the productized services side. Yeah. And I think we always will have an element of that. We'll just take a, we, we just have a very small, uh, group of clients that we deliver that higher touchpoint experience to. Yeah. Okay. Yeah. I, I, I think, I think what you're doing is a really good thing, right? And, and it has a massive need. Like, we have many, many businesses, but, um, you know, I think for most people in, in the past, they put together a website- Mm-hmm looks good, and they think it's all working, you know? But reality is- Mm is, you know, there's a whole SEO sorta- Yeah set up in the background that needs to be happening, and you can have a crappy looking website, great SEO- Yeah and be converting- Yeah just with a simple landing page, right? Yeah. Um, so what you're doing is you're, you're basically, you know, you've got sites like, uh, Revvit and Lovable- Mm and all these guys pumping out websites like there's no tomorrow. Yeah. But there's no depth to those websites, so- Mm I guess what you're offering is the, uh, the depth behind all that sorta stuff, right? Yeah. It's... I, I mean, I love Lovable and Replit to some degree. I think they're, they're, they're fun. Um, I used Replit well before it became an AI agent platform- Yeah um, 'cause it was a great way to host code on, on the web, um, and just make it very simple to execute that code. But what happened was is that we actually noticed that you don't have any control again over the depth that that server would have, um, on the, in the settings that you could adjust on that server. Right. Okay. So, so if we wanted to make it faster, you, you weren't sure if you were gonna have the fastest server speed. It may just dial you down when it gets volume. Right. And then the SEO tooling that they built into those, uh, are pretty average. Yeah. It's not the greatest SEO agent out there, and you really want full control over the website to do that properly. Yeah, yeah. Fair enough. All right. So, so let's, um, so you've been running the business for two years. Mm-hmm. Um, what, what's the, what's the growth been like for your business over the last two years? I think, uh, the initial few months were a d- like, it was more about getting out there and educating the market- Yeah and educating people and offering training and workshops. Um, so we offered, uh, the AI workshops for all the local chambers of commerce and the tourism bodies. Yeah. Um, we partnered with the local council to offer one-on-one support to businesses, and, uh, that was most of where our, our focus was on, was educating the market. Yeah. Um, and we also found that n- not many people even knew what, what, what we did, what the value of AI was, and there was a lot of, um, a lot of, you know, discussions of, "Does it do anything? What, what do we need to do- Yeah with it?" Um, recently it's become- really hot topic and everyone's, you know, I've been having a lot of people coming up to me and talking to me about where do you see, um, where do you see it going? How can we, how can we just have that conversation? It, we're just getting a whole heap of volume of people coming up to me- Yeah and, and asking me about it, reaching out to me. It could've been I've just talked to them in the, in the first, you know, uh, six months of starting the business, and then now after two years they're coming back and saying, "All right, let's do it." Yeah. And I'm like, "Great." Yeah. Um, we should've done it then, and then you would've been in it early, and it's actually a real advantage to be early, like, you know, in the SEO area. Yeah. You, you, it was the real advantage to be early. Absolutely. You know? Absolutely. So same thing's happening now. We're, we're doing so well now. We've just hit, you know, we just hit 20K MRR. Yeah. So that's- Wow we're pretty happy with that. Yeah. Um, and it's been a process of- Then productizing further so we can, uh, reach more people. Yeah. And now we're on a, we're at a global scale. So we work with, you know, universities in the US, we work with, uh, univ- local universities. Um, we're working with, uh, a whole, um, plethora of broker-style businesses right now- Yeah that are feeling the heat. Yeah. Because, uh, you know, there are certain styles of businesses that do feel like there's a lot of disruption, but also with that disruption, there's a lot of opportunity. Yeah. And, and so in that, I guess, in that scale- Mm that, it, it's funny you mention that, that AI's, you know, people, people didn't know about it two years ago. Mm. And now they're all coming back to you. I always joked with everyone about, uh, QR codes. Mm. QR codes were around for years, and then all of a sudden COVID happens- Yeah and the government starts enforcing QR codes for everything, and now everyone knows how, what a QR code is- Yeah and how to use it, you know? Yeah. But prior to that, you couldn't, like, you could do some magic stuff with QR codes, right? Still can. It's a- Yeah I love QR codes. Yeah. No, absolutely. But, but, you know, I think AI's moving in the, in the same space where people are starting to take it real serious now and starting to understand that it's impacting every part of your business- Mm-hmm every part of your personal life, and, and everything that goes with it. So, you know, reaching out to somebody like yourself or your company to try and help is, is a, is a need. Mm-hmm. I guess in, in that, so in that, you know, you've got people coming to you now, and how are you managing that scale? Like, is there a impact personally to you or p- to the business? Are you having- Mm to employ and put job ads out? Where, where are you at? Yeah. Personally, it is a constant balancing act of, um, you know, of onboarding, um, of, of the marketing side and then the delivery side. And I think every business, that was a conversation we had with every business is, "Do you want us to help with customer growth?" Yeah. "Or do you want us to help with the delivery to those customers?" Um, for myself, it's been very interesting 'cause I've had to reinvent, um, myself. I've also had to reinvent the systems that I, I put myself into almost on a, like, quite a regular basis- Yeah because new things come out, and they, if I, if we don't shift over to that, then we won't see the growth. Yeah. So we ended up in the last, excuse me, two weeks- Really focusing on a new system, an agentic workspace, that really allowed us to hack and build our own software inside it Yeah. Okay and host all that on our own machines. So that really has unlocked a whole heap of growth for us. So I'm able to, uh, really just converse with a team of, you know, of 10 agents that then delegate to a, a team of up to 50 sub-agents underneath them. Uh, and we could just r- I can just go in and dump what we need to do, the scope of the work, and then brainstorm that, build it out, and, uh, go and ensure that we've got all angles covered- Yeah when we, when we del- develop the plan, and then execute on that, and then review the, the body of work that they've produced. That's been a massive unlock, and it feels like weeks have been condensed into days- Yeah. Right just from that one transition I still think, though, that, you know, from that process we've, um, I've involved more of my co-founder, who's more on the commercial side. Mm-hmm. But she's obviously still involved in, in almost every aspect of the business as well. Yeah. Um, and that's been really helpful for me as well. So I can focus on the systems- Yeah and the architecture, and setting up the tools for the agents, 'cause they're constantly asking me for new tools, so I've gotta- Yeah. stay on, um- Yeah stay on, on the ball with them. And, um, and then she can focus more on the, on the commercial side, and that really, um, that's had a massive impact as well. Yeah. So being able to bring in expertise into that kinda system, um, really amplifies also the agents' capabilities. Um, I can spend more time in the areas that I enjoy, and then she can spend more time in, in those, those parts that I was just kind of half doing. Yeah, yeah. Like, you know, I wasn't quite updating the website all the time, and she can focus more on that. Even though, you know, she, she comes from, uh, maybe not so much the, the, you know, technical background, the system is, is enabling, like- Yeah uh, you know, we've got each other doing things that we never thought we, we would've done before. Yeah, yeah. Um, so that's been incredible. Um, on my personal life, I think it has been massively focused on the business, and it's probably suffered a little bit- Yeah in that process. Maybe sleep as well. Yeah, yeah. Yeah, people always talk about mental health, right? Yeah. It's, uh, as a- Yeah as an entrepreneur starting a business, that's- Yeah that's the biggest impact I think there is. Absolutely. So yeah, I mean, I, I, I try to, you know, get to the gym as much as possible. Yeah. Um, stay fit, stay healthy, eat good food. What, what about, like, so, so as a business owner at level one- Mm-hmm been running your business for two years. Scaling- Mm-hmm it's, it's now sounds like it's gonna move into a, a high growth space. You're very lucky that you're in the AI space 'cause you can put on agents- Mm as opposed to, you know, more costs in staff, I guess. Mm-hmm. Um, you're paying tokens instead of staff. Yeah. Tokens per intelligence is something I've also weighed up. Sometimes it's cheaper to hire a person. Yeah, right. Okay. Um, so we are, we have got roles that we are looking for. Yeah, yeah. Well, how did, how does that process work? Work, work, walk me through that. So- Hmm for someone like yourself in, i- you know, that's heavily into AI- Mm-hmm what, what does that look like? Do you get to a point and say, "Hey, look, we need a, a, a marketing girl- Mm-hmm or a marketing guy." Um- You know what? I'll, I'll spin up an agent. Mm-hmm. Or are you, are you looking at that and saying, "No, look, I need a, I need a person," but obviously that person's gonna have to be, you know, uh, very literate in AI at the same time- Mm right? Mm. What, what does that look like for you? Marketing's one that we actually spent a lot of time looking at. Yeah. I think really look, if you look at marketing, and we don't wanna, we don't wanna amplify slop. We don't want to... I don't wanna put an agent in there that just amplifies, um, you know, mediocrity. Yeah. I don't- that's not what we're trying to do with that. We're trying to amplify expertise. Yeah. I think it's the best ti- um, we're, we're in one of the greatest times to start a business, and it's even, um, even better for those who have expertise because the value of expertise has never been greater. Yeah. Right? Yeah. So if I was to look at the next role, I would look at, yeah, absolutely, someone who can go through with a fine comb or, and review possibly some agents that would augment that role of marketing. Yeah. So it wouldn't exactly, um, do all the marketing, but it would probably do some research around their roles. It would help them with some copy creation or refinement. Um, but they would still, at the end of the day, be responsible for the review of that copy that was put out- Yeah into the internet. So it needs to still sound like the brand. It still needs to still sound like a human, um, and still cut through the thousands of posts you see on LinkedIn that are probably AI slop at the moment. Absolutely, yeah. So Well, it's hard to tell what's AI now and what's not- Yeah half the time, you know? And that's a big problem in itself because if it doesn't clearly sound like it's not AI, then mo- even if you aren't using AI- Yeah then people will think that you are. Yeah, yeah. So you have to, you have to be so contrarian, you know? Yeah. You have to- You have to say something that probably an AI wouldn't say 'cause it's agreeable. 'Cause- I was, I was talking to a- Yeah I was talking to my business partner about this just today actually, and, and said, you know, like maybe my, my poor pronunciation of words, and he goes, "Yeah, but AI will eventually just be able to do that too." I've had some good banter with some AI agents before. Yeah. I, I don't know, I still have these, these discussions with AIs that go, "All right, well, if that's the case, then here's another, like, approach to that." Like, I still find angles that the AI's missed. Um, and I still s- take a very, a contrarian take because I have a, a real strong opinion on, on what I'm- Yeah on that work that I'm doing. I don't know if... It'll be interesting to see how, how well they get at, at making jokes. Yeah. You know, that's a good one. I- Yeah I think, I think we're still, comedians are still in a job for a long time. Yeah, yeah. Absolutely. Yeah. Do, do you, do you think, just on the, do you think that given you're a- in AI- Mm and some of our business is playing that space, and like, like I was saying, everything personal to business is- Mm-hmm AI right now. But I, I do wonder sometimes to myself if we're in a bubble right now- Mm where the outside world still isn't aware of how deep it is. You know, a plumber or an electrician or, you know, you go to local networking events and that sort of stuff, like- Yeah some of these guys that are running really, really good businesses just have no clue. They're not- Mm it's, it's not impacting their world as much as it is. So I guess the question I have for you is, with that in, I mean, they've all got websites- Mm-hmm but with that in thought of mind, does marketing for you look like, you know, uh, online, or is it old school marketing where you're just getting out and talking to people and saying, "Hey, look, you know, this, this world of AI is crazy over here." Mm. "But, you know, I gotta get down to your level and talk to you about the impacts of that for your future and your business- Mm and your family," et cetera, et cetera. Or is it, you know, AI-generated marketing on Facebook? You know, what does, what does that look like? I have a, I talk to a lot of clients who said that they hired an AI guy, and the first thing they did was generated all their marketing copy for them. And I looked at all the marketing copy and I went, "Well, it's generating absolutely nothing." Yeah. Like, in terms of, like, likes. And then you're also using a social media scheduler that completely ruins your algorithmic- Yeah uh, opportunity there. So, um, I think there's, like, there's a different approaches to, to marketing. When you look at, uh, like, someone who's so far away from... Well, like, right now we're building brains, right? So that's gonna obviously impact knowledge work. Yeah. Uh, well, the next step is building hands. Yeah. And those hands are gonna be involved in the construction for sure. Yeah. We're even seeing exoskeletons come out, and that's been a ma- that's gonna be massive for- Yeah construction guys. Yeah. You know, um, I know a lot of them who ruin their bodies- Yeah and they don't wanna do that. Yeah. And they would, they don't, they, that's the last thing they wanna do. And I all, I'm all for that. Um- Relaying it to the everyday person who probably is getting quite a lot of referrals already. Um, you know, referrals, paid acquisition, that's a short-term strategy, and it can sometimes turn off. Yeah. Like, you can suddenly lose that lever, um, either because everyone's referred every, all the work, and you're kind of drying up that circle that you've got. Um, and then you need to have another lever to pull. Yeah. And that could be more of a sh- a longer term lever, and that longer term lever takes a few months, maybe a year to build up. Yeah. So you've gotta start before one of those taps turns off. Yeah. Um, when it becomes a conversation for a, a someone who has a website, not all construction guys have a website- Yeah 'cause they're looking at maybe higher ticket tenders, and they don't even need a website 'cause, you know, they've got a mate who refers- Yeah them the tender, and then they just apply and they get it. Yeah. But there's certainly parts of that business that could be involved with AI. Yeah. Um, like the, the tender itself- That's important. Like, you wanna get a r- sometimes they have 300-page tenders. Yeah, yeah. They're very, v- quite large, and you've got to have almost your whole business, like all your processes, your policies, the team members, everything documented to quickly fill that out- Yeah um, in a format that then the person who's providing the tender can read. So that's one side of the business. If you were looking at the higher ticket construction space, that is really, um, really has the opportunity for automation on the tender process for document processing. Then you go into invoicing, um, time sheets, like, you know, uh, then not just time sheets, but also automating the time sheets so all they need to do is click a button as they enter the site and click a button as they exit. Yeah. It s- needs to be so involved in their workflow. I've seen a lot of people who build time sheet software. He's a construction guy, he's out there, he's vibe coding with Lovable, and then they build it, but then it's not actually ingrained in the workflow. Yeah. So everyone ends up having to open another app and then enter all the details in and then do it, so they never end up using it. Yeah, yeah. So, you know, with the clients that I've worked with, we did, we, we started with what we could f- at that point- Yeah to automate everything we could, and then mapping out the workflow outside of that to go, "How do we go about, um, automating more of the time sheet?" So as they enter the job site, it's a big sign at the front that has a QR code. Yep. And they scan it, they click a button, it fills in all their details from a database and authentication software- Yeah in the back end, which most people with vibe code, they, they don't see that technical complexity behind the scenes. They- Yeah and then with that technical complexity behind the scenes creates the simplicity for the user- Yeah, yeah and the reliability for the system for future audits. Yeah. And so when you then go to apply for a tender and you can say to them, "I can tell you exactly how many people are gonna be on site at any time," the person who's providing that, that tender application, who is looking at your application, goes, "That ticks all my compliance boxes." Yeah, yeah. So that, um, that's a really important for inc- not just streamlining the tender process, but improving the quality of the tender through the systems that the business has built. Mm. And so that's, it's completely separate to a website in most situations, but in some situations you can actually build a very simple website and then if you've already got that in the right foundations, you can then build those systems off the back of it as well. Yeah, yeah. You know, coding agents are very capable of spinning up a subdomain and starting to build your ERP systems out the back. Yeah. If you've got the right architecture in place, it's, it's, you know, it's an incredible experience to do. So- If I was to, you know, say that a website is just the first step that we take, um, for, if I asked a construction guy, "Did you know that, you know, roughly about two to three times of the traffic, or two to three times the human traffic you've got on your website-" uh, is actually AI. So if you look at it, you've got one third humans- Yeah two thirds AI. Yeah. And that's humans sending their AI agents out to understand your website. Now, if it's a, a, you know, electrician and they've got a website or even maybe even like a, you know, a painter- Yeah um, and they don't know that you're a world, like award-winning painter because you've got a badge there, but it doesn't actually show it in plain text. Yeah, right. Then no one knows you're an award-winning painter. Even when the AI agents go out there, they go, "Sh- which is the best painter in the region?" You're not being shown up because it's not in a machine-readable format, even though I could tell you, you know, first of all, this, like you're- you've got incredible business you see. Yeah. But what the AI see is not that. Yeah. And that, for them, they go, "Right. Well, what does that mean for me?" And I said, "Well, if you, if you fix that, then you're being recommended in the top three spots of every AI query." Yeah. And you're... I would be surprised if next month we didn't get you five more clients. Yeah. Yeah. And, and they go, "Okay. Well, that makes a bit more sense." Most of the time they understand SEO, so it's the same sort of process. Yeah. It's getting you in the top. Yeah, yeah. Yeah. I, I think, um, I think that's an interesting thing, and I don't, I certainly don't know what the stats are on it, but I mean, people are making queries now, and I mean, I know I certainly do. Mm-hmm. I'll use Claude or ChatGPT to ask questions instead of Google. Mm-hmm. I find Google almost like, you know- It's a bit, bit hard now bit old school. Yeah. Bit old school. Yeah. Yeah. So you, y- do you, are you finding the same thing with clients that are, uh, people are using- Mm-hmm ChatGPT and the, you know, Grok and all the, all the k- kind of mainstream AIs to, as their search engine now? Absolutely. Everyone is using it. Yeah. And even when they do a Google search now, the top 30% of their screen is, is AI. Right. So e- almost every query has an AI overview. Yeah. Not only that, but Google has filled the last section, uh, below the AI overview with a Google Maps listing. Yeah. Often, often you'll get five Google Maps listings before you even get- Yeah the 10 blue links. And then not only that, but underneath that, that's usually three to five paid ad positions. Yeah. So then you're competing right down. So the stats around it are before AI, you would get 15%, uh, click-through on average for, uh, being at number one. Yeah. Now number one gets you 8% click-through. Oh, wow. So it's about half. Yeah, yeah. Um, you're, you're... If you are in that AI overview, you, the customers that are referred and click through, I think roughly three times more likely- To, uh, use your business. Yeah. So the moment they are on your site, they have already got the intent to buy. They're ready to use you because they've done all their research before they've hit the site. They don't even half the time even read the site- Yeah because the AI's done the reading for them. And that's what I found when I was doing the workshops, is that people were teaching AI about their whole business and then going, "Okay, now we've got a problem. I need to, you know, ask a expert, like a lawyer or an accountant, um, about this problem that I'm having. Um, we've gotten pretty much there with the solution, but I just need someone else to check." Yeah, yeah. And then they'd go, "Right, here's the top three accountants in your region- Yeah to go and do that." Mm. And then you can pick based on, um, wh- uh, the top three, bes- based on the best for the budget, best for expertise and, and longstanding history. You can pick based on different factors and have it instantly compared in front of you. The Business Game is proud to partner with HighLevel, the all-in-one CRM and marketing platform built for growing businesses. With CRM funnel, website, and course builders, marketing automations, email builders, and so much more, HighLevel software will drive your customer retention and your profits through the stratosphere. And HighLevel is fully white label, which means you can brand the platform and mobile app with your logo and company colors and then resell it to your customers for a monthly subscription. Ready to go HighLevel? Click on the link below to begin your 14-day free trial. Yeah, yeah. It sounds like what you're doing's pretty solid. I mean, I, I see lots of, uh, you know, social media posts and lots of people out there spinning up websites and sp- you know, out there calling the local plumber and saying, "Hey, I can do your SEO," or, you know, that sort of thing. But, um, I think there's, I think there's a real art to putting that together and making it a solid offer- Mm and not just a, you know, we alwa- we got that stuff back in the day with, uh, websites as well, right? Mm. You know, when the internet popped off and everyone was a web designer. Yeah. And, uh, yeah, that sort of thing came along. Yeah. I still find that even these days is a lot of people have, like, learnt a website editor. Yeah. And, um, they're still stuck with it. Yeah. And I've even had people say that AI's gonna go away, and they're building websites or they're building IT systems, so they're saying it's fine now, web- AI's gonna go away. And I'm like, "I don't think it's going away." That's I, I don't think it is. They, they think it's gonna get, it's the cost is gonna become too much for people. Yeah. And I go, "Well, okay, if even if it costs them $100 in AI tokens, it costs you, they cost them to pay you $1,000- Yeah to do that. So if it's $100 and it does it almost instantly, I think they're still gonna go for that. Um, it depends. There are still, like, there's still a lot of areas where the cost is one factor, then ex- like the quality as well is another factor. Like, is it actually good quality? If you're a very early stage business, just whip up a website, get something out there. Yeah. But then at one point, you know, your lovable site is not really doing all the, the search stuff that you need. Yeah. You need to start optimizing to get that organic traffic coming through. Yeah, yeah. Um, so yeah, it's... I, I still see everyone using these, like, pre-built website editors that integrate into the CRMs, and they have like a whole s- all-in-one suite of tools. Um- I, I st- if we're shifting most interfaces, like I see that the interface generally is dying for most industries. Yeah. And they're shifting into the AI agent doing the interactions. And I, I hate using interfaces now. I almost... It annoys me. Yeah, yeah. Because I just say to the AI, "Go out and use that interface," or like- Yeah have direct access into the back end of that to use that website. Yeah. So if you've got a booking calendar on your website, you should be exposing that as a small tool for the agent to use and read, so it's not only quicker for an agent to book a meeting with your potential client, but also more secure for them. Yeah. So they don't have to like navigate the interface and potentially be exposed to prompt injection attacks. Yeah. So there's a number of other factors that come into that. But also, you know, you just as a business wanna be something that agents u- want to use- Yeah because, um, and, and because it's, it's gonna make you more money, but it's also, it's also gonna help you in so many other areas. Like I, I could go, I could go into like the level of customer service that you can like reduce because you've provided enough documentation on your website. Yeah, yeah. I won't even go into insurance companies and like the whole process of them providing documentation on- Yeah, yeah on that. But, um, there's so many industries as well that, that are affected by a lack of documentation on how they operate. Yeah. And I think of that because coming from a background where, you know, where we're building agents, I always know that the f- the, it's, uh, it's garbage in, garbage out. So if you don't have, uh, good quality data to be fed into that agent, then it's just gonna do really... It's gonna be subpar. Like- Yeah the outpa- output's gonna be subpar. Yeah. And it's the same with a, like a, you know, you'll hire a junior researcher straight out of uni. Like, you've gotta give them good processes, you've gotta give them good tools, you've gotta train them well. It's the same process as that. Does take a little bit more of a technical understanding of the, of the system as a whole, and we spend more time on the architecture, um, than we do executing the individual components, and that's something I've become, yeah, really, really good at. Yeah. Yeah, okay. Yeah. Let's, let's shift gears for a second and just think about, uh, your business and scaling. What, what, you know, with all the things that are happening with AI- Mm moving faster, you mentioned people coming to you more often. What, what breaks first for you in being able to manage to scale your business to the next level here? Like, as in the s- the business as a whole that could... Like, what parts of the business would break if I was to see that, like, looking forward? Yeah. Like, I mean- Yeah every, every business, uh, you know, e- through every stage has some challenges- Mm as they scale. So I guess where you sit right now in your business and your setup and systems- Mm-hmm what, what would be, what do you think as the influx of business comes to you, people start to understand AI more- Mm-hmm where, where's the pressures of your business start to crack? Like, what cracks first? Mm. Is it your system? You know, I mean, you're a systems guy, so hopefully it's not your system. You know. They're pretty reliable. Yeah. The systems are pretty reliable. Um, I'm very proud of the systems that we've got in place. Yeah. Um, it doesn't mean it's perfect. You know, there's alwa- because we are always uh, moving and adapting. There's a certain level of experimentation that comes with new use cases. Yeah. And I'm always open to improving my systems as well. Like, I, I, I always open that up. I've actually open sourced the, you know, the methodology behind how we score a business on visibility. Okay. And we open that up to other people to, to, to contribute to. Yeah. Um, we've worked with marketing agencies to build the systems out so they're more suited towards their needs- Yeah um, and oth- and different businesses, the way that they work. What breaks most of the time is when a business is quite rigid, or there's bottlenecks in that business that they don't want to remove. Right, okay. And so when we go to plug our systems in and to work with them, we find that we've then gotta, um, spend a whole heap of resources to go get around that- Yeah uh, process that's, that's really stopping us. Um, it could be a form that's not being, uh, made in a readable format. Um, it could be ge- unstructured data that's not, um, that hasn't got the right labeling around it. Or a booking calendar or Or a booking calendar. Yeah. Yeah. Booking calendars are, uh, actually- surprisingly broken systems. Right. Yeah. Um, and, and not, like, also sending links to people is another one that if, you know, if they, if they, their email systems don't trust those links, then you're gonna have problems there. Gets lost, yeah. Deliverability of emails, that's a big one that breaks because you want the data around, uh, the email, whether it's been opened- Yeah have they read it, have they clicked the link, and you can't get that without links. Yeah. So you need to have built trust with their email service before you do that. That's one that breaks, and I've spent a lot of time building agents that can use reliable email systems- Yeah, right to do that. Um, now I'm shifting people more into, well, our agent workspace is like a Slack interface. Yeah. And that's where we're shifting people into, so we can provide a public space for everyone in our community. We're about to launch that so everyone can come into our workspace- Yeah and experience the future of work and- Yeah, okay collaborate. And the partnerships we build in that workspace, when we d- create a joint channel, I've never seen anything like it. It's- Yeah, right like the level of, uh, s- of execution that you get to when you can collaborate with someone and then talk to their agents, and I'm even getting my agents to talk to their agents. And then you go, "All right. Now what are the questions I need to ask that they haven't answered?" And then you- Yeah go get on a call, quickly go through, like, what's the questions, get the context out of them. It's so fast and, um- It's really exciting. And, you know, we, we work with some of the biggest, um, agent tooling companies in, in the world that are right on the frontier of that, you know, like Firecrawl. Yeah. Like, they've just raised 78 million or something to build, like, the super intelligent intelligence connector into everyone's agent- Yeah um, platforms. So that's been a really ex- interesting experience collaborating with them. And then we're working with another team called Raft that's out of San Francisco, and, um, they only launched, like, a month ago. Okay. And we got onto them, like, as they launched, and then worked with their team to build out the system so that we wanted, like, so that we were comfortable using it, and now we're bringing other people into that. So what breaks in that process is, you know, it's very early. These tools are very early. Yeah. So we now need to build out the runway, the infrastructure with that team. So we're launching apps in their marketplace. We're rolling out, um, systems that plug into their systems, and we're building out on top of our, our systems we built use Cloudflare- Yeah and building out context layers so that we can plug into their systems and help people do it even faster. That, like, pretty much take our path, but do it much faster. Yeah. Um, so I think the only thing that breaks then, like, I would, I would see, I would see myself bringing in another, um, another engineer for sure. Yeah. And, but I would also see as- myself bringing in someone- Like, I want an engineer, someone who's just right in the weeds of it all- Yeah and really can go through with a fine comb. Like, you know, leave no stone unturned when it comes to the systems, it comes to the data, and make sure that they're, um, just constantly in that space. Yeah. And, and, like, allowed, given the permission to take ownership of that space. Um, and then the other side of that, and I think it is kind of an engineer in a way, we look at, like, a marketing person- Mm-hmm. Yeah that does have some engineering capacity or the ability to, um, you know, build a quick tool and a coding agent- Yeah you know, that sort of thing. Like, or interact with a coding agent in a way that, um, that they can describe the system at least from a high level of what they wanna build- Yeah but still go through the copy really, uh, r- really well. Like, they've really gotta be able to communicate an, a level of authenticity, a let- level of expertise also in that space that we'll, we'll need to replicate from how- Yeah we're doing it right now. Okay. So those are the two things that I think would possibly, um, possibly show how are we gonna get someone with, like, that level of expertise in a market that has a talent war going like crazy right now. So... It really is, yeah. Yeah, so. Well, well, on that, I, I guess in itself, what, what do you think... What, what's your view on what AI will look like in the next three years? How fast are we moving? Hmm. We're moving very fast. Yeah. Um, we're moving light speeds. Like, in, in, in terms of I think it's, it's coming down to a day-to-day thing, where every day we've got a new model, every day we've got a new harness for that model, a new tool for that model, and it's just getting better and better really quick. Mm. And the, the models that are coming out now, they're really expensive, but they're suddenly getting a lot cheaper. Yeah. So we've just seen, like, Fable 5.1 come out, and then suddenly Opus 5.5 came out, and then Sonnet 5.5 the next day. Yeah. So it was like a week after or two weeks after Fable was released, we saw models that are 40, 50% cheaper with the same level of intelligence. Yeah. And, um, w- almost the same level intelligence. How do you, how do you keep up with that? Like, how do you... Like, just as a, you know, I'm a business owner, right? Mm-hmm. How do I keep up with what, which one to use- Hmm what's next? I see, like, lots of thought leaders, you know, they'll, they'll be out and they'll be like, "Claude, it's all Claude code." And then next they're like, "No, ChatGPT just upgraded. Forget about Claude." You know, like- Yeah you know, and, and like what you just mentioned, three or four big players into the market- Yeah upgrading. Like, w- how do you keep up? What do you, what do you... What does the world... Is, are we just in this bubble right now where it's gonna, the ball's gonna slow down? Or is it one or two break free and you just hedge your bets at the moment? I think by 2027 it's gonna be pretty incredible the level of intelligence we're gonna get out of those models. I couldn't tell you if it's going to be, um, a logarithmic pattern, if we're gonna start- Yeah to trail off. Yeah. From what I can see right now, we are seeing some bottlenecks in the systems, but then all of a sudden new systems come out and new models that are just... They're not even text-based models, they're like determin- they're almost comp- almost deterministic models- Yeah um, where they just operate in code, and they're much smaller and cheaper, and like you embed them into the broader system, so then you become... It's more about having multiple agents than one big agent that does everything. So How do you stay up to date? Well, you certainly have to be using AI to stay up to date- Yeah to a, like, really high level. You know, I've optimized all of my feeds to just staying up to date. But also, you know, there's newsletters, there's, uh, there's, there's YouTube channels I look at. There's also people in the space that I stay, I, I stay up to date with, and we share information, and it comes straight out of, you know, straight out of Silicon Valley or say that- Yeah we actually got a really good digital scene, digital hub scene in, in, in this area. Yeah. A lot of great talent. Um, some of the best engineers in the world come here. So we're, we're in a great spot for that. Um, we've got... You know, I've got group chats that I've built so that people can just openly share intelligence, and that's what- Yeah keeps me, like, seeing what, uh, like, other people are doing as well. And I'm in, uh, multiple group chats that do that. Um, I'm always looking at the, like, the harnesses, and I stay up to date with the harnesses, and even talking to the teams who are building the harnesses before they've even released products. Right. I'm talking to them about the product rollout, and then all the way from, like, brainstorming the product to what's the roadmap like, to how are we gonna get VCs- Yeah involved. Yeah. That's how I'm currently staying up to date, and it's a, there's no one channel, because if you rely on one channel and it goes, then, you know, it doesn't keep you up to date, then you've, you've lost that. But at the same time, like, I could tell you what happened this morning, but then in terms of testing it out in your business- and actually applying it to your business, you might still be, like, a few weeks off. Yeah. Um, there's certain things that are important to keep up to date with, like a new model launch can be really helpful. Like Opus 5.5 has been really helpful for people to launch inside Claude. If you're, you know, if you're a business and you've got ChatGPT and Claude, I would encourage you to tr- like, allow both. Like, I wouldn't say that any one of them's gonna be the winner. Yeah. I think we're gonna have a lot of winners. Yeah. Um, and then you wanna start to secure your data so that you can shift that data around whoever the leading model is at that time. So it's, it's inevitable that we're going to have a constant battle between two, maybe three, maybe a cluster of people at the frontier- Mm-hmm who will keep winning in each domain. Yeah. And you need to be able to shift your data and your context through each one of those models as you need. So to do that- You actually need a, we've... Our agent workspace system we build on, Raft, which we've launched an app into, they own the context layer, so we host that on our machines. Okay. And they've built a system for us to, uh, train the context layer, and then each agent can use its own subscription and its own models. And we have the ability to switch the model out and then still maintain the, the context- Yeah of that business. Yeah. So you don't feel like the moment you've, you've set up ChatGPT, and all your data's inside ChatGPT, and you can't shift to Claude because it's gonna take you a few days to do that. We just do it the morning of. Yeah, right. You know? Yeah. So a strategy you can take in a business is trying to secure your data- Yeah at, at some point, 'cause sometimes when you're starting, you just kinda need to get started with a, with a model, and you can take the slow process of migrating that context out. Yeah. Um, unfortunately, ChatGPT has made it a little bit harder to do that. Eventually, you wanna get to a point where your context layer is stored on your machine, and you can plug that into whatever you want. So that becomes, like, the brain that you hold access to. Yeah. And you can plug that in wherever you want. Um, you could even give it hands if you want. Yeah, yeah. And so that's been, you know, get- having a little server, um, like a Mac Mini. Yeah. I always encourage, like, some kind of s- uh, easy system for you to use, and, you know, these models are gonna be optimized for Mac Minis 'cause OpenAI has just bought 10,000 of them- Right just to optimize their models for the interface inside- Wow macOS. Yeah, yeah. Um, doesn't sa- I'm not saying that, like, Linux isn't gonna be optimized at one point. It's just, uh, like- It's leading the way at the moment, Mac Mini Mac, Mac Mini's the- Yeah the leader at the moment. Mac Studio is awesome, but not everyone's got that kind of budget. Mm-hmm. But a Mac Mini base model M6, great setup. You can plug in a hard drive. You can plug in a camera. You can plug in, uh, microphones and, uh, like, custom keyboards with displays and just say, "Set it up for me," and it will load its own custom firmware into all of it, and you'll have your own system. Yeah, yeah. Um, so you're giving it all the hands to do that, to train, like, on document processing, to doing, like, just voice transcription and just teaching it about, like, your s- your business and how you work and just sort of giving it direction. It's much easier to do that. Yeah. Is that going to, like, change... I, I think for a lot of people, they have to take some responsibility, I guess, around owning a Mac Mini and then having someone who can come in when they get stuck. Mm-hmm. Um, but that is a little bit different to how typical IT would normally work, where we would normally be fully managed- Yeah by the IT company. Yeah. Um, so there's, there's a new space emerging from that, um, and how you manage IT services 'cause, you know, hosting your own brain is, is incredibly powerful. With, um, just a question around this as well, if you were- Right now, speaking to our audience- Mm-hmm and saying, "Hey, guys, y- you know, there's lots of people out there that wanna kick off their own business, do their own thing in the AI space," what, w- given where you're at, you're, you're ahead of everybody else in this space, what advice would you give somebody if, if you had to give them three tips on how to get started, what they should do, what they should- Mm focus on? There's the first step, which is, yeah, build out that context. Like- Yeah build, secure your data and build it out. So that would be, uh, using a voice transcription tool, like, uh, voice memos on your phone- Yeah or WhisperFlow is incredibly powerful for replacing the keyboard, which is like I was saying at the start, like, you know, we've got this bottleneck here, which is the keyboard. Yeah, yeah. We need to remove that. Yeah, absolutely. Um- I've, I've started doing that myself, to be- Yeah honest. I just, I love it. Yeah. I just talk to it and have conversations, send off all my emails. Yeah. Done half my job by the time I get to the office. There you go. Yeah. That's what you wanna be, and everyone should be- Yeah at that level- Yeah pretty quick. Um, if you're still typing emails, uh, to, like, you can still edit the email at the end, but if you're still spending a long time typing things out- Yeah um, a lot of people are in offices, so I understand that you can't always be talking in an office- Yeah about everything, and I can, I can get that. I guess try and get into a booth where you can talk and open that up and really just, you know, ex- explore every angle about- Yeah your business or every angle of that product or that problem, um, that you're looking at, and go, "You know, what have I missed?" Or, like, try to come at it from a very neutral perspective. "This is what I think, but what are the different angles?" Yeah. "What have I missed about the way that I've, I've, I've currently assessed this?" Um, so there's a, there's a slight approach to how you would talk about your business as well in a more neutral way, but yeah, just drafting emails is gonna be huge. Yeah, yeah. Um, so yeah, capturing the data and securing it would be the, one of the first steps, like building out that context layer. Um- The other one is, you know, b- building out your, the culture in your team to allow for, uh, like, uh, for them to use the tools that they need for, for the job that they're trying to, trying to do. Yeah. Um, it could be any tool. They could come to you and say, "We've actually started using this tool a lot, and it's created a huge amount of impact." Um, encouraging the team to come to you and say, "We've been using this tool." For most people, they go, "We don't wanna hear about AI. We don't wanna hear they're using anything outside of the approved IT list." Yeah, yeah. Yeah. So you need to also allow them to come to you pretty early on in that experimentation process to open up that workflow and have a culture of, you know, of, of sharing that. Yeah. Whether that's in a group chat, or whether that's just one-on-ones where you just open it up and go, you know, "What have you been using? Let's get it on the approved IT list," or, "Let's look at it to see if there's another tool that's similar that could be- Yeah easily approved by the I- uh, by our IT guys," or whoever it is that's gonna do the approval. Um, that would be another one, like opening up this, opening up sharing. Yeah, yeah. I think outside of, of that- You also wanna have certain parts of your system that aren't being shared, that are locked down. And by that I mean Get the data that is quite structured or, or unstructured, and then you give it some structure. As in, get your financial records, and obviously, you know, any, any advice you take from an AI, you feed your financial records. I encourage you to do your own due diligence on it. And, you know, I'm not, I'm not saying go and do what it says with your finances. Um, but get some of those things like your finance records, your close meetings with your, your directors or your founders. Um, build that brain out so that you can have a real conversation about the future of your business- Mm in a way that you're open to sharing with the model as well, a little bit more context. Yeah. 'Cause you may have parts of your system that you go, "Look, we may not share, like, exact finance records to everyone on the team," but at least for myself or for the core, um, directors or founders, you need to have a, a free space to share all of the context and for everyone to look at financial projections- Yeah um, look at projections of sales across the team. Look at bottlenecks for, like, where you're spending your money on, where, where are the bottlenecks in, um, if you're getting customer feedback, like, getting all that into a, you know, a brain that you can go, "Where are people stopping?" Yeah. Getting all that data on the telemetry of your business across your website, across the systems that people interact with internally, and querying that, even if some of it's quite, like... You know, you need to consider the, the sensitivity of it as well, and there's different guidelines out there from, um, you know, there's, there's a, there's the voluntary AI standard from Australia, which just looks at a, a traffic light system when you- Yeah when you do that. And you can go and look at those sort of standards when you, when you go to share that kind of context. But building out a, a context layer that, that does really know your business from those voice memos, um, f- and documenting all your processes, that you could then say, "All right. What's the next..." You know, if you plug in a really big model, even a team of really big models, like the big frontier models, and you have each of them, uh, brainstorm different angles of your business, then they can come to you and say, "Here's the different things we see you doing in the next few years in terms of your roadmap for your business, in terms of the, you know, efficiencies, um, new products, new markets- Yeah new value you can create." That is a, a really powerful thing for businesses once they've built all of those systems up before that, that we talked about. Yeah, yeah. Like, you know, the ingestion of, of the context of your business- Yeah the openness to sharing and providing some of that. That's, that's, that's the hard one, right? Mm. For people to get comfortable with connecting their financials- Mm and connecting, you know, all their sensitive documentation. Yeah. 'Cause right now I, I, I think, I think almost AI's in this space where everyone was scared to tell it things at the start. Mm. But I'm pretty sure it knows everything about everybody by now. Like, people are very comfortable with telling it anything. Yeah. But when you get to, you know, do I wanna connect my QuickBooks account to it- Mm maybe that's a different conversation. Yeah. Yeah, it's super powerful when you do that. Yeah. The QuickBooks and Xero and Stripe, they've all got some great connections into it. Um, there's different levels of connections you can get as well. You can get, uh, the point at which you're actually integrating straight into the back end of those systems, not just, you know, the lighter- Yeah lighter connections you can do inside Claude. Yeah. Um, you can start to download a, a database of all your financial transactions onto a machine- Yeah and then have a very small model, um, hosted on that machine, go through and, uh, remove any sensitive information as well. Yeah. There's processes to do that. But yeah, uh, to getting the... Like, understanding AI is probably the, the bottleneck to doing that, 'cause most people think that any kind of AI coming in will probably go and share it with, um- Yeah, other, yeah the wrong people, yeah. Yeah. Um- Yeah it probably knows a lot about you already. Absolutely. Yeah. And I think you've gotta be also considering whether that is actually sensitive data. 'Cause most people will go, "Oh, you know, the, um... Like, my purchases of software products, I don't want anyone to know what, what software we're buying." Yeah, yeah. It's like, I think they probably know. They know. That's right. I think they know. That's right. I, I think you just need to now look at your subscriptions and start looking at how many subscriptions you have. And just, I think the gain of you for the next one month optimizing your software subscriptions is, is bigger than the risk that possibly in six months' time they're gonna train a model around your software subscriptions. Yeah. And then everyone knowing about it, and then maybe getting the right, your exact systems if they prompt it hard enough, is, like, a smaller risk. And even still, yeah, I think the gain there is you're probably gonna change anyway in that time. So just, you know- Yeah think about, there's different, uh, different approaches there, and different models, and different security layers for each model, different levels you can go to with that. Yeah. So we- we're, we're, we're nearly time up- Mm-hmm and gonna wrap up, but, um, one of the questions I wanted to ask you is, what, what do you know now that you wish you had known 10 years ago? I would say that I would just be encouraging myself. Um, and m- many people say this, but just encouraging yourself to just, just start. Like, have a bit more confidence in starting. Yeah. It come, it comes out a lot. We hear that, we hear it a lot on the podcast- Mm is, is people just say, "I wish I had have just started earlier." I mean, even for me, I'm, I'm someone that gets things done pretty fast. Mm-hmm. But I, I always feel like it's still too slow, so. Yeah. You always- I feel you on that you always go, "Oh, my..." Is it, like there's imposter syndrome, there's- Yeah there's so many different elements that come around that. At the end of the day, if you just start and figure out some space that you can start in that allows you to improve yourself without exposing, like, too much, ex- yourself too much. Yeah. Like, w- we started with, um, education initially to learn and, and train people. We'd learn things and just immediately relay that to other people. Yeah. That allowed us to start and learn- Yeah and then share that learning. Yeah. And, uh, that was just a process we're very open about as well. Yeah. And we're open that we're all just learning. Yeah. And that really helps, helped me start. I mean, it's different in the knowledge working space. I'm sure there's different- Yeah like, there's different spaces out there that it might be different. But just starting somewhere that allows yourself to just experiment and, and do what you wanna do. Yeah. Um, because, you know, I could never see myself, um, you know, going back to what I was doing before- Yeah because I, I love what I'm doing now- Yeah so much, and it's the, it's where I need to be. Yeah. And I, I could see that. And, uh, I just never, like I, I don't wanna ever go back to, you know, being the, the lesser version of myself because I'm not allowing myself to do that. And I think if we have one life, it's gonna sound very inspirational, but, um, I would hate for, at the end of that life, to look back and go, "I never achieved the things I wanted to achieve." Yeah. So just do it. Totally agree. Yeah. Totally agree. Do you, uh, do you have a, a quote that you live by or a quote of your own? I just say, "Let's do it." Let's do it. Yeah. I mean, Nike said, "Just do it," right? So. Just do it. Just do it. I say let's do it because I just like saying it. Yeah. 'Cause, like, if we wanna make something happen, let's just do it. Absolutely. Yeah. Totally agree. And I, I agree, Nike's got something there. Other than that, like, you know- Um, uh, another quote that I, I like to live by is, um, "You often never rise to the level of your goals, but you'll always fall to the levels of your systems." Yeah. And so you can have big, hairy goals, and that's great, but often the- at the end of the day, what's gonna catch you is having reliable systems in place. Yeah. Absolutely. So. All right. Last, last little piece is our, our quick fire, which is just a bunch of questions. Mm-hmm. Uh, try and answer in a sentence or less if you can. I'm gonna read them off the screen though. Mm-hmm. Um, so with a business, would you build... rather build a business or buy a business? Mm. Just quickly, I would really l- like, I would just buy, uh, build a business right now. Depends on which stage you're at, but I, uh, building a business right now is really, um, inspiring. It's a great time to build business, uh, build a business and, um, it depends, maybe if the business is already established well, but you often get much more satisfaction out of building. Yeah. What about, uh, revenue or profit are you chasing at the moment? Revenue. Revenue? Yeah. Um, profit, obviously we're in a very, you know, interesting space with profit. It's... But at the end of the day, if you're, if you're starting, you need to just chase revenue because it's just about getting into the market and, um, it may not be that, like, you may not make that much money, and when you actually look at the bigger picture, it's not that much profit. But- Yeah. Absolutely. Yeah. Yeah. Growth or stability? It's a tough one too. That's a tough one. I like stability. I like stability, but that doesn't mean that you, you can't, you can't chase growth. So I mean, yeah, I like stable systems, but I guess in the broader business, you wanna be chasing some level of growth that's acceptable without sacrificing stability. Mm-hmm. What about, uh, hire fast or hire slow? Hire slow. Hire slow. Take the money, so, you know, investment- Mm-hmm VC, or keep control? Keep control as long as you can. Bootstrapped or venture-backed? We're bootstrapped, but, you know, we're chasing a very competitive market, so, um, if you need that growth, then try to get to some point where you're not sacrificing too much equity- Mm-hmm to get to that next level. But I can understand everyone's in different stages, but I've seen a lot of startups that chased equity too quick before they had actually validated product market fit, so. Yeah. Yeah. Okay. Uh, big idea or great execution? Great execution. Perfect product or first to market? You could be, like, second to market and you'd probably be fine. I agree with that, yeah. Get started with a product, and then you will learn. Like, make it a good product, but you will learn how to improve that, even if you're not the first one. We chased a very competitive market. There's probably 50 other people doing it. We still looked at the angle at what they weren't doing well, and it's not perfect, but it's m- we're providing much more value than they are, and people are shifting over those, off those products to go for us, so. Yeah. We weren't first, and we don't have a perfect product, but I think that what we chased was agent tooling, so just providing that data to the agents, where most of them weren't. Yeah. And so we built an agent context layer for AI visibility. You just query that, and you do what you want with it, and that was, I think, perfect for agents. So may not be perfect for people. I mean, AI is a perfect example of that, isn't it? Like- Yeah look at the race right now. Yeah. It's crazy. Yeah. What's one thing that you've done in business and you've made a mistake that you'd do again? Like, I would do the mistake again? Mm-hmm. Like, I would happily keep making that mistake? Um, I'd happily offer things for free again- Okay and make no money. Yeah. And I'd happily keep doing that if I could, and I probably will in the end, probably get to a point where I'm just, you know, philanthropic, like, just offering things for free. But, um, to start, you've definitely gotta just get started, and sometimes that's for free, sometimes that's for very little profit, and you're just scraping by. Um, but that compounds really quick. Yeah. So before I ask you the last question, where can, uh, where can our viewers and listeners, uh, find you? You can find me on, uh, Google, so you can search my name. Hopefully we find you. Yeah. I've done a, I've done a quity- pretty good job of indexing myself on Google as someone who helps people index on Google. So you search my name, Huxley Peckham. Um, you can also search, uh, for our product, areyoufoundbyai.com, and, uh, you can check out our product there, but I've got a whole, you know, whole... I'm on every channel, so you can find me. Perfect. So last question, Huxley Peckham, if business truly is a game, what's your definition of winning? Just playing the game. Just being on the court. Being in it. Being in the court. If you can be on the court, you're already winning, I think. Even on the sideline, like, if you're sitting on the sideline and you're on the, on the bench, and that's something I learned from a young age from being in, in sports a lot, is even just being ready for when the opportunity arises and you're sitting on the bench, that's the best you can do, even if it never happens. Be ready for when the opportunity arises. Love it. Mm. Thanks again. No worries.