February 25th, 2026 Episode Recap

The February 25th, 2026, episode of MC Fireside Chats, hosted by Brian Searl, focuses on the cutting edge of AI and technology within the outdoor hospitality industry. The panel features recurring guests Matt Whitermore (Climb Capital), Kurtis Wilkins (RJourney), and Cara Csizmadia (CCRVA), alongside special guest Patrick Mullen (AffinityX). The discussion centers on the rapid evolution of AI, moving beyond simple chatbots to sophisticated agent-based systems that are reshaping business operations and marketing strategies for campgrounds and RV parks.

Kurtis Wilkins opens the session by introducing the "Dark Factory" pattern, a sophisticated AI agent workflow. He describes a system where a core agent manages a team of sub-agents—such as specialized legal, finance, and operations personas—to complete complex tasks like rewriting an entire HR handbook in parallel. This approach allows for massive organizational efficiency and flexibility, enabling businesses to rewrite operational foundations in hours rather than months.

Brian Searl provides context for this technological leap, noting that while the industry is discussing advanced agents, global adoption remains low. He cites data showing that only 0.3% of the world's population pays for premium AI tools like ChatGPT or Claude, and 84% have never touched AI at all. Despite this, he emphasizes that for those in the industry, steering conversations with specific "expert" personas is already standard practice and essential for staying competitive.

Patrick Mullen shifts the focus to digital marketing, explaining how his team at Influence Outdoor Hospitality uses AI agents to analyze complex Google Ads campaigns. He notes that while Google provides its own AI suggestions, specialized agents can look at specific signals like assisted conversions and attribution more effectively than a team of humans. He highlights the emergence of "point solutions"—AI wrappers trained on specific industry data—that offer more accurate results than general models.

Cara Csizmadia raises concerns regarding the shift in search behavior, specifically the introduction of travel ad formats within Google’s AI Overviews. She questions how independent campground operators can compete with large Online Travel Agencies (OTAs) that may dominate these new AI-driven spaces. The panel discusses the "zero-click" search phenomenon, where users get all the information they need from an AI snippet without ever visiting a campground's website.

Patrick Mullen responds by emphasizing that foundational SEO remains critical. He explains that for campgrounds to appear in AI results, they must focus on "answering the questions" through robust FAQ sections and properly marked-up content that addresses specific guest needs, such as being pet-friendly or having specific amenities. He suggests that localized businesses still have a significant opportunity to win in intent-based searches.

Matt Whitermore shares his experience transitioning from standard AI interfaces to "Claude Code," a local developer tool. Though not a developer, Matt explains how using this technical environment allowed him to replicate and optimize months of strategy work in just two hours. He highlights the shift toward "vibe coding," where non-technical operators can use AI to build custom software and departmental strategies by simply describing their business processes.

Kurtis Wilkins adds that this accessibility to custom software is revolutionary but warns of the risks, noting that while AI is solving impossible mathematical and coding problems at an exponential rate, it still requires human commitment to refinement. He points out that AI has moved from high school-level capability to the cutting edge of mathematics in just one year, underscoring the speed of change operators must navigate.

Brian Searl demonstrates the sheer speed of modern AI by showing an application called "Chat Jimmy," which generates the full code for a comprehensive campground iOS app in 0.18 seconds. He predicts that in the near future, software will be written in real-time as users open their computers, essentially eliminating the traditional "moat" that software providers currently rely on for competitive advantage.

Cara Csizmadia reflects on the personal and emotional shift in AI interaction, mentioning how some individuals now use voice mode to "process" their lives and business problems. The panel discusses the efficiency of voice-to-text for brainstorming and the implications of AI systems "learning" an individual's personality over time, leading to both massive productivity gains and new privacy considerations.

In closing, the participants share their contact information, including Cara Csizmadia for the Canadian Camping and RV Association, Matt Whitermore for Climb Capital and Unhitched Management, Kurtis Wilkins for RJourney, and Patrick Mullen for Influence Outdoor Hospitality. Brian Searl concludes by inviting listeners to the upcoming data and analytics episode and promoting the newly released 2026 Insider Perks pricing report.

Shorter summary: Host Brian Searl led a discussion on the rapid evolution of AI agents and "dark factory" patterns with experts Matt Whitermore, Kurtis Wilkins, Cara Csizmadia, and Patrick Mullen. The panel explored how these technologies are revolutionizing campground operations and digital marketing, emphasizing that even non-technical operators can now leverage high-speed AI tools to build custom software and optimize guest acquisition strategies.

Event Date: February 25th, 2026

Event Type: MC Fireside Chats

Organizer: Modern Campground

Interviewer: Brian Searl

Participant(s) Details:

Participant 1:

Matt Whitermore

Director of Market Expansion

Company: Climb Capital

Participant 2:

Name: Kurtis Wilkins

Position: Private Equity Analyst

Company: RJourney

Participant 3:

Name: Cara Csizmadia

Position: President

Company: CCRVA

Participant 4:

Name: Patrick Mullen

Position: SVP

Company: Digital Operations of Affinityx

[00:00:24] Kurtis Wilkins: I go by both Brian. Professionally I go by Kurtis just because our CFO goes by Kurt.

[00:00:39] Kurtis Wilkins: I can have a new name on here that's fine.

[00:00:47] Matt Whitermore: Hey everyone, Matt Whitermore here, director of market expansion at Unhitched Management and Climb Capital. We're an owner and operator of RV parks and campgrounds, growing portfolio across the country. I [00:01:00] love to join on this call and learn about marketing and AI. So I'm excited to chat today.

[00:01:12] Matt Whitermore: I'm huddled over a tiny little table in a conference room. I just joined a coworking space in downtown Syracuse.

[00:01:19] Matt Whitermore: I was getting a little antsy working from home. I've got an awesome office set up, but just like to get out of the house. And so it's this beautiful brand new office building with all kinds of events and stuff in downtown Syracuse. So it's been a lot of fun.

[00:01:54] Kurtis Wilkins: Kurtis Wilkins with Our Journey, RV Resorts and Advanced Outdoor Management. I guess [00:02:00] I am I call myself the homie in our office that stands for Head of Monetization, Income and E-commerce. Matt, I think you should adopt that title. We do a lot of the same stuff.

[00:02:15] Brian Searl: Go sorry, sorry, please. I thought you

[00:02:26] Brian Searl: Alright. Cara, go ahead.

[00:02:57] Brian Searl: It's a good thing AI is moving really slow. You didn't miss much.

[00:03:01] Brian Searl: And Patrick, our special guest. Last but not least.

[00:03:25] Brian Searl: For sure, yeah thanks for being here Patrick. So typically how we start this is we'll go to our regular panelists, Kurtis, Matt, Cara, I know Kurtis is excited so maybe we'll just like first we'll ask Matt and Cara so they have a chance to talk. Do you guys have anything that has come across your desk in the last month that you feel like we should be talking about before Kurtis hijacks our show? He's like looking at me swaying on the show.

[00:03:56] Brian Searl: Alright, I'm I'm interested. But Matt, Cara, you have anything [00:04:00] before Kurtis gets into it?

[00:04:05] Brian Searl: You timed that really well. You got back in.

[00:04:10] Matt Whitermore: Kurtis, I'm ready for you to blow my mind.

[00:04:15] Kurtis Wilkins: Okay so I oh sorry Brian I should ask first before I dive right into this like

[00:04:25] Kurtis Wilkins: So I left the last show and I always leave with great ideas too. So I wanted everybody to know that and everybody was talking about Claude and Opus 3.0 or 3.5 and I was like okay I'm a little bit behind on that. I was a big Grok fan. I was a ChatGPT guy right? I was like okay I'm gonna go dip my toe into Claude real quick and see how that's going. And dipping my toe, I ended up dropping into an ocean of capacity and I this has kept me up at night Brian like how awesome this is. And it's called the Dark Factory pattern. And I if any for those of you [00:05:00] who don't know what that means and I'm this is gonna get a little bit deep into AI but imagine a team of AI agents that work together to create something outside of your regular context window. For example, I want an HR team to review my handbook. This person is going to have this perspective. They're going to think about it from this way. They're you know we're going to be looking at it from safety and compliance and OSHA. And then I want this other agent to be looking at it from the time card and the clock. And then I want to introduce a finance person who's going to be looking at is like how much does this cost? And then you can assign a very large handbook run 70 agents in parallel and have them build sub agents with the team that you've assigned and it will absolutely it will overwrite [00:06:00] the whole thing, spit it out and shove it out to you as a completely full functional document. I was I just thought it was amazing. And that also goes for like Ops teams. It goes for development teams, construction teams, like anybody that you want to have take a look at this and you want to have an entire

You want to have the best attorney team in the world. Just go say hey I want the best attorney team in the world to your agent have them list out what those job titles look like say now assign those tasks into small tasks and it's changed. It's I mean it's absolutely changed like organizational structure. And I was just blown away by it in the last two three weeks.

And then there's only something like I don't know 8 or 9% of people who actually have free subscriptions. There's still like 6.8 billion people in the world who have never touched AI to ask a conversation.

So I thought that was interesting. But like we're getting into the what you're talking about is getting into the weeds with agents, right? So

[00:07:39] Brian Searl: I'm gonna I'm gonna try to say it how I would say it and then you can see if you agree like a little bit with what you're talking about, right?

So we've done something similar with the Open WebUI that I have Matt and I were talking about before the show that I have installed on my little Mac mini. But either way the purpose is agents, right? And so the earliest stages of this were if you knew how to prompt in ChatGPT when it came out, you could say go in and act as a lawyer. And what that does is it basically takes the whole corpus of ChatGPT's information like it knows everything about everything on the internet, right? And when you ask a question like write me a contract that is good for waivers at a campground for my swimming pool. It will pull from all that knowledge and somewhere in there will be lawyers that are refined a little bit, right? But it's if you if you go into that same conversation you say act as the best lawyer who has 20 years experience writing waivers for campgrounds with swimming pools. All of a sudden it narrows down that knowledge and cuts out all the other garbage, not garbage, but all the other information that it could pull in and doesn't need for this conversation. And then it writes you a much better waiver. So that was the very early that wasn't an agent, but that was the very beginning of steering a conversation into something you specifically want. Act as a lawyer, act as an accountant, act as a digital marketer, act as a whatever, right? And so we were doing this two years ago. But still so many people have never experienced even that. And then what you're talking about with agents is the same thing. So it's and I don't I am not familiar with what you just talked about with the factory. I've never heard that thing before. But but agents I have. So you can explain what that is and how you found it in a second if you want to. To for everybody else.

That is a lawyer that like the deeper example is that you get an agent that is actually trained on Our Journey. And it and all your SOPs and all the things and all the ways that you right that you interact and like to write legal contracts and a history of all that stuff so it's the lawyer for Our Journey literally. And then you can have an accountant for Our Journey and a marketer for Our Journey and fuck you could have a team you're talking about 70. I don't know if I can say the F word on this podcast but I just did. Sorry for the family-friendly children who are probably listening to us banter about this stuff. Then you can spin up like 70 agents but you can have a team of lawyers. You could have a lawyer that specializes in liability at the campground and a lawyer that specializes in contract law and a lawyer that specializes in HR and then these can be permanent like this is what we're talking about when we talk about the chance what's coming with the employment replacement.

[00:10:59] Brian Searl: Thought that we want it to happen but it is happening and whether that's only 0.3% of people or not like it's gonna be ugly.

[00:11:57] Brian Searl: No I think you're right like an SOP to instruct the agents how to create the thing with the team.

[00:12:21] Brian Searl: when you say that I think of Mel Gibson and Conspiracy Theory movie.

[00:12:38] Brian Searl: Okay now that you say that there are actual dark factories in China with the robots running them that have no lights on.

[00:12:53] Brian Searl: Yeah they're in China so I just didn't put it together with what he was saying but yeah.

[00:13:06] Brian Searl: For sure we got you though we're following.

[00:14:22] Brian Searl: Yeah the AB testing thing is crazy. There's a I was doing I think it was even an Indian or a Chinese startup company with AI that I saw a few weeks ago and I don't have I didn't have a chance to log into it but it was a platform built on the whole like we will tell you what's going to happen with your marketing for example right? And so then you can put in like a and you can do this with agents too right the way you're talking about it Kurtis right? But you can say like I want to run this Google Ads headline here's the park that I want to run it for what will be the output of this camp- or the impact or the results of this campaign? And then you literally have 150 different AI agents quote unquote responses that take on the persona of your audience so there's a single mom of two kids and a retiree who's traveling across the country and whatever your buyer personas are at your park. You have maybe 10 of those and then you get 10 versions of each 10 of those and then you ask them like would you click on this Google ad if you saw it? Yes, no, why, not, whatever. And so you can AB test in 10 seconds and figure out like what's the perfect Google ad campaign I can run. Same thing with blog writing and Facebook posting and not saying you should do this every single time right but it's really interesting.

[00:15:41] Brian Searl: How do you know that? How do you know it's human made? That's going to be the problem.

[00:16:04] Brian Searl: I mean I think that was unique to the United States I think the French have always had good bread but anyway continue.

[00:16:12] Brian Searl: Patrick, what are you guys doing with AI at influence?

[00:18:56] Brian Searl: Yeah, it's interesting to me like you mentioned the human in the loop thing. Because I've seen this as I build out Google Ads campaigns using the agents and stuff like that is that there will get to a certain point where it will make a suggestion, it will say to do something and you'll be like, no, I know better than you. But I think that only lasts as long as the agent doesn't have, Kurtis, you were talking about context and front of a context window, but I'm gonna talk about context in like information that it has about your organization. I think the only reason that I'm, I don't even want to say smarter, I just have a more complete picture to refine that AI output now is because I have been working for the client for x number of years and I know everything about the property and I have weekly meetings with them and I know what their pain points are and what their weak spots are and their strengths, weaknesses, opportunities, whatever threats. And so as soon as the AI gets all that plugged in, I think the human becomes a weakness. It's like we were, who is it? Is it Jeff? I can't remember his name. Jeff Cavins is Outdoorsy? I think the CEO of Outdoorsy. I know it's Jeff. I can't remember his last name. It's Cavins I think. Anyway, we were supposed to stand linked in yesterday about insurance. And I responded to him and I said what why are you gonna need insurance in five years when everything on the road is not everything but like a significant portion of cars on the road are robo taxis and self-driving cars? You won't need to have your Tesla Cybercab insured at the same way you would a human driver because the weakness is the human driver. The self-driving car is going to be a hundred times more reliable. So yes right now there's insurance from companies like Lemonade and stuff for full self-driving, but long term like you're gonna get to a point where it's so saturated, it'll take 10, 15, 20 years, right? But you're gonna get to a point where robot axis and everything else are so saturated in the market that it's going to become a liability issue for the government to even issue a human a driver's license. And that'll come in 10 or 15 years. And then it'll be illegal for humans to drive on certain roads. Because you'll just kill people and the AI won't. And so it's crazy.

[00:23:00] Brian Searl: Oh I agree with you, yeah. It's definitely necessary now.

[00:23:13] Brian Searl: And then we'll all retire on the beach and not have to do any work every day.

[00:23:21] Brian Searl: It's not my dream. I'm gonna be miserable if I have to retire and not do anything.

[00:24:15] Cara Csizmadia: Definitely.

[00:25:08] Cara Csizmadia: Interested to hear from you marketing guys about this. I saw Google is already testing travel ad formats inside their AI overviews. Are you guys seeing your clients already prepared or at least starting to prepare for that? Like obviously classic SEO and pay-per-click and all that stuff is totally shifted over. Are they moving in that direction? Your clients, are they doing it already?

[00:25:36] Brian Searl: Yeah, for sure. As long as it's not Kurtis. The stage is yours.

[00:25:38] Kurtis Wilkins: Yeah, to me the interesting thing about what Google's testing and they said that this was like specifically for the local and like destination search categories, and I thought the craziest part about the whole thing really is the way theyre displaying their ads inside the local overview context is that it's an ad from an OTA. So I'm going out, my search context is what are the good local properties with amenities with pools in Syracuse? Google then builds this context and drops an ad for an OTA that fits the context of your AI model. Not a paid ad specifically for what they used to do which was the exact property or whatever it was that specific keyword. It's trying to match the context with the ad and that's usually an OTA. And that changes for a business owner specifically who might not be buying on an OTA from their own marketing perspective. They're trying to win with direct bookings on brand search right? The real question is how to win inside that AI context with direct bookings directly from your PMS because those integrations don't exist inside of whatever AIO currently has available inside to Google. So right now I wonder, is there a way around that other than joining OTAs who are paying hundreds of millions of dollars to Google currently? Is that Patrick, to your point, does PMax allow you to enter those overviews specifically when someone uses like a generic destination search? My assumption right now is not to the top, it would push you down if they don't have enough ads in their inventory specifically for those OTAs and then it would be a specific brand search potentially. But I find this I find that particular problem with this kind of the scariest and I love I generally love a lot of this stuff, but that from a business level, that's what keeps me up the most at night in terms of direct booking strategies right now for hospitality specifically.

[00:25:40] Patrick Mullen: That's a great question. So we do it a couple of different ways, right? There are ways that you can go ahead and get even if it wasn't a hotel, you could push certain room nights that weren't hotel related. So we've had some capabilities like that. So we've been fortunate. Look, we all think that's critically important. A lot of that actually has to do with SEO and OTAs, right? Like for branded search, if you're not winning and a separate conversation but if you're not winning and an OTA is winning or in this case Google has better placement the question is you're paying commission on things that you should have owned. But with that said, we don't have any partners right now pushing in that direction. But certainly we've been doing it for a while. We think that with a mix of PMax has always worked really well as long as, the inventory is able to be pushed. So it's critically important, but it should be for backfill. And I think a lot of folks don't look at it as backfill.

[00:26:34] Patrick Mullen: Yeah, sorry, just to close it out. Like completely agree, right? As a marketer, OTAs provide tons of value. The Billboard effect is completely real right? As long as you have good basic SEO down right? They will find your brand eventually out off of the OTA but nobody, most folks unless it's independent property look at it as primary. And when they're doing that, they're paying commission up front on everything that they're selling. So I think it's always a strategy that they should consider, but I know that's not necessarily the AI conversation. But Google's got great AI and they're going to continue to make sure that they're leveraging the paid side of things. I think it's important that there's always a good balance between what you own and what you can own versus just taking the easy way and paying the commission if you will, if a [00:27:00] click cost is a commission, ultimately it is.

[00:27:48] Patrick Mullen: So if I could take a quick crack at it, what I would say is, SEO is SEO and a good foundation in SEO is going to get you a reasonable placement inside of those AI overviews or AI snippets.

[00:28:14] Patrick Mullen: Yeah, absolutely. So good SEO to influence is certainly making sure that when we're talking with a customer. There are always keywords and strategies that they want to employ right that they think are critically important but when we do the research if we understand what people are searching for and what some of those more active keywords may be, we want to make sure that we're showing up top 1, 2 or 3 organically for our partners in those spaces. Now having said that it changes a lot with you know the AI responses right whether we call it AIO or GEO that's associated. Really what it boils down to is people are asking questions quite often and you know we don don't see many campgrounds or even other businesses right that talk about the problems they solve for folks. And how can you do that? You can do that by having really robust FAQ sections making sure that you know you're marking it up properly and I know that's a slightly different topic but it all...

[00:29:15] Patrick Mullen: It all ties back to AI because look at the end of the day, if you've got great FAQ content, you're talking about problems that you solve so for a campground if you know you make sure you're talking about the fact that you know you're pet friendly and that you've got certain amenities so when folks are asking the search engine and getting an AI response and they're saying you know I'm traveling such and such a location I'm looking for a camp ground or an RV park that is pet friendly. If you've got that content and it's marked up and it's proper in your SEO that's gonna help you actually show up in those results. Now I think it's tough because now you're competing against content publishers. Right so now you've got local brands vying for some of the same things that national brands can win on a regular basis but if we actually tie it all the way back to this community and it's campgrounds and RV parks they can win and they can win at a very high rate. They're not gonna be able to go in and do the things on Reddit that you read about, they're not gonna be able to go in and do some of the things that larger businesses do because they've got swarms and swarms of people, maybe agents can help out with that over time. But it all comes back down to consistency proper content and understanding what people are searching for. And I think the big thing with AIO or GEO is location and it's about asking answering the questions that logical folks who are in a buying cycle will ask. And if your property is structured accordingly you're going to show up at some point.

[00:30:49] Matt Whitermore: Yeah you know I'm taking it all in. I am not a digital marketing expert so it this is all really educational stuff to me.

[00:31:01] Matt Whitermore: You know it is. And I honestly I have to just separately I have to thank I have to thank you Brian because some of the partners at Climb Capital caught the episode from last month. And now because I got fired up and I was talking about how great Claude is and they saw how passionate I am about AI. I am now charged with rolling out our AI policy and strategy and enterprise implementation across the organization. So thank you. Thank you Brian for some new responsibilities at work. But in it's a joke at all.

[00:31:42] Matt Whitermore: I mean you know I have not yet really employed anything a gentic other than right Claude code is a gentic in and of itself. But that's right I'm not I haven't set up an agent to do anything automated over a prolonged period of time yet.

[00:33:22] Matt Whitermore: It's crazy. And so a month ago we were on this same version of the podcast here and I was talking about Claude skills and Claude projects and felt like I was, humming along, doing some cool stuff, still just scratching the surface. And I'm fortunate to have you know some friends, some buddies that are more tech savvy than I am and they both separately were giving me the same feedback if you got to get on Claude code. Just get on Claude code. And I'm like, I don't have anything to code. Like why would I get on Claude code? I'm not I had this mental block that let me think of an app I want to build first and then I'll get on Claude code. And then you know I set it up on Claude desktop and I click over and I'm like I gotta set up this weird file path thing and like I don't even know what this is and like I've got some skills to work over here let me just get back to work with my skills. And it took a final push last week and I what I did was I took my 18 skills that I've that I built, I downloaded all of them. I got over the hump of setting up my file path workspace. And I dumped my 18 skills into Claude code and said this is my first time using Claude code. I don't know anything about automation or orchestration layers or coding. I'm a novice at this but I feel like I've built a pretty good system of Claude projects and Claude skills. I want you to [00:35:00] treat me like I'm a fifth grader doing all this stuff for the first time and tell me all the optimization opportunities I'm missing out on how I can replicate this in Claude code and just let's strategize around this, show me what I'm missing here. And I spent the entire day replicating my entire Claude skills and Claude project setup transferring all that context from one side to the other. And I couldn't even really articulate it but I was having some pain points with Claude projects and Claude skills. Namely and I had no idea that Claude projects limits the context window on whatever model youre using back down to 200,000 tokens as opposed to what Opus and what Pro use natively. And the other issue I was having with Claude projects is the context isn't preserved sequentially from project to project. I had to figure out a way like if I wrote something over here and needed it to reference it in project B and C I had to figure out a way manually myself to preserve that knowledge and send it across the projects. And so Claude code just immediately the obvious solution to this was just using Claude code instead. Was the right you have it's a sandbox environment that you're in a ChatGPT or in a Claude

[00:35:58] Matt Whitermore: And so and then the other piece is that the artifacts right? Yeah you get these cool artifacts or in ChatGPT it's called a canvas. And yeah it's like a it's like a document that's separate from your chat but there's no architecture there. There's no file or folder system, there's no organization. It just goes and creates new artifacts. You know sometimes it does sometimes it edits the old artifacts, sometimes it creates a new artifact. So I was fighting it like I created a project management skill and went deep on okay how do I optimize my project? How do I optimize my project instructions? How do I optimize my personal preferences and all these things. And it got incrementally better. But as soon as I jumped over to Claude code that you know just got over that mental block like I still have no real intention of coding. I use it to create things mainly markdown files right? That become strategy documents or content or right, you know I'm still in that mode of like I'm creating, I'm building out departmental strategies. I went through and for Climb Capital like you know created a Claude MD for really each department that I touch. We're looking we're doing a lot of automation on the acquisition side on the capital raising side and just like still just scratching the surface. But that day I felt like you know I've had these mini epiphanies with AI over the last 2 years. And that day 1000% more like my life just changed radically. I feel like I'm a 100 times more productive. What I did in the last week with Claude code on even on Claude skills and projects would have taken me two months. And I was like mad at myself at first because I had spent probably two and a half months building out this Claude skills and Claude project setup and I literally replicated it in two hours and made it 100 times more effective in Claude code. And I was just like this is amazing that I made this improvement but imagine if I just listened to my buddies that told me to get into Claude code two months ago. I'd be light years ahead. I'd be the two months if I had spent that in Claude code I'd be a year ahead. I'd be two years ahead.

[00:39:53] Kurtis Wilkins: And Yeah I have to add I wanna can I add to that real quick? That's what I'm talking about right there in terms of that dark factory [00:40:00] pattern is you said two devs but why limit it to two devs? That's Why not a thousand? Why not right exactly.

[00:40:07] Kurtis Wilkins: Yeah. And I'm like and so I look at it so you know we own our own PMS like we're right there with software and I look at it as it's a it's an interesting perspective for business owners because form fit software just became really accessible. All you have to know is your business process very well and you're going to have to be committed to refinement. And all of a sudden now you have access now there's a lot to that. So I hope anyone listening to this podcast knows that there's a lot of professionals that have right? We're not quite there. There's a lot of bugs that'll still happen. We call it vibe coding for a reason. But

[00:40:52] Patrick Mullen: Yeah but it Can I add something to that Kurtis? So look I totally agree and I think you both nailed it. When you know your business and what you're trying to solve that you're going to be able to use AI over time and probably sooner rather than later to actually build something custom. But taking that down the road is what I wonder, right? And the thought for me anyway is that's the scary side of this. It is a real risk. And I certainly hope that folks who use it really use it for POC first and one of the ways that we find it to be valuable is instead of going through wireframes and sharing with you know your teams and your executive leaders how things are going to be built and how they may work once we develop it, you can vibe code something that actually works today and then take it to your dev team and your security team and your database team and say this is the functionality that we're looking for, let's reverse engineer the full spec. And now you have a way that you can actually move through your dev process much more efficiently but still employ all of the necessary controls around enterprise software.

[00:50:42] Cara Csizmadia: I'm just thinking I never have to build another PowerPoint from scratch ever again.

[00:51:05] Kurtis Wilkins: I want to talk a little bit about so all these skills and all these tools like one of the things and Cara this was you that I think it was like five months ago, six months ago is a long time. You told me that you were talking to your AI and I now talk like a crazy person to my AI for 30 minutes every morning while I'm on my walk. And I just talk about alright what are whether hey what are the new features that I should be including talking about AI or maybe it's something we're working on in the company and this is to Matt's point and Patrick's point. It's AI teaches you AI just when you're working out just talk to it. You look crazy like I recommend walking at 5 AM, there's usually nobody out at that time. It's interesting.

[00:51:51] Brian Searl: Crazy would be not talking to it.

[00:52:20] Brian Searl: The privacy is interesting too. And like I don't want to take a dark turn here but I just want to say real quick did you read Cara, you and I are up here in Canada with the shooting that happened in BC. Did you read about ChatGPT? And how they shut down the guy's account like two months before the shooting because he was talking about violence and they never reported it to the authorities.

[00:52:42] Brian Searl: Yeah. Like it's really interesting the things that it will know about us and whether that's minority report-ish or like helpful, nobody really knows yet but continue. Sorry.

[00:53:10] Brian Searl: In the speed with which is going to run. So Matt you were talking about coding and I know you haven't done Curtis and Patrick have done coding too. But you know how long it takes like you have to do you ever get annoyed by it you're like it's going to code this whole thing for me and I gotta sit there and watch it? And it takes for a while, right? But this is where people are not understanding of where it's going to go. Jessica, share my screen for a second over here. She's gonna make this big. This is something there's a couple examples of this right? But this is just an easy free example I could pull up. It's called Chat Jimmy, you can go to it at chatjimmy.ai. It is not like a v- it's not the world's best coder to be clear, okay? It's not the world's best anything. But I want you to pay attention to this real quick. So I'm gonna so you guys can see this big enough. So down here in the bottom I have typed vibe code me an MVP so this is a minimum viable product version of an app for a campground on iOS. Include every detail I need, make it super comprehensive, give me the full code. All right so you know how coding works on Claude Code right Matt?

[00:54:06] Brian Searl: Okay, watch this. Don't blink okay? There's your code.

[00:54:14] Brian Searl: It just did all of that... I'm scrolling down through it all. It did all of that in .181 seconds, 15,300 tokens per second.

[00:54:29] Brian Searl: This is where AI is going. So you saw OpenAI had a partnership with a company called Cerebras which is a big company that does super fast inference chips. A couple of months ago, a month ago or something you saw Nvidia they didn't buy Grok [G R O Q] but they basically gutted the whole company for $20 billion and took all their talent. It was the same play. This is the inference speed. So when people think like it's gonna oh you're never gonna vibe code software cause that's gonna take like people are gonna have to know terminal and all that other stuff. This is where the future is. It's gonna literally spin up apps on your computer in real time in half a second and you'll never know. Like you'll click the email button and it'll write the email application and by the time you're used to it opening on your computer, it'll be written and it'll be your email application. So it's gonna get really crazy guys. In case you were just bored with where it's at now.

[00:55:37] Brian Searl: I think yeah, mine calls me an idiot I think pretty frequently.

[00:55:42] Brian Searl: I mean I'm definitely dumber than the AIs are. We all are. I don't know what to tell you.

[00:55:55] Brian Searl: I didn't listen to the whole speech but I heard anecdotally about it.

[00:56:44] Brian Searl: Hasn't even started to rev as fast.

[00:58:01] Cara Csizmadia: Oh gosh, ccrva.ca or on any of our social channels. We are Camp in Canada. Final thoughts, I mean I feel like I say this every time but I'm always just shocked. A month from now to connect with you guys again at all the crazy things we'll be considering in such a short period of time, so I can't wait.

[00:58:25] Matt Whitermore: Thanks for the time everybody. As enjoyable chat. You can find us at climbcapital.com or unhitchedmgmt.com and unhitchedrv.com. To quickly fix my Claude Code issue, if you have advice for me before the consumer side of things, I'm gonna go uninstall and reinstall Claude and if that doesn't work I'm just gonna go buy a Mac. So I appreciate this the chat.

Kurtis, final thoughts?

[00:59:16] Brian Searl: Awesome, thanks for being here Kurtis. Last but not least, Mr. Patrick.

[00:59:46] Brian Searl: Awesome. Thank you all for joining us for another episode of MC Fireside Chats again. If you're not bored and sick and tired of hearing me, I will be on Outwired live with Scott Bahr in about 2 hours, we're going to be digging into some pricing data we released a big huge report. The Insider Perks pricing report 2026 came out today. If you haven't seen that, it's really good details like 70 pages long, really good stuff. We're going to dive into that on the MC Fireside Chats episode next week which is our data and analytics one. But other than that, we're going to be talking about data, AI, and more stuff on Outwired. So if we see you there, great. If not, then we'll see you next week on another episode of MC Fireside Chats. Thanks guys. I appreciate you.

[01:00:18] Cara Csizmadia: Thank you.