An Ode To The Luddites

AI is everywhere, but the shed industry doesn’t get points for being first. Cord Koch goes solo for a candid, ground-level take on why slow adoption can be smart, especially when the hype cycle is louder than the results. If you’ve felt pressure to bolt an AI “solution” onto shed sales, customer service, or marketing just because everyone else is doing it, this is your permission to pause and think.

We unpack the stats that should make any portable building business owner stop and re-check the math: widespread AI implementation across US organizations, but a much smaller share reporting real bottom-line impact. Cord connects that to what many of us see daily in shed leads and conversions: customers still want clarity, speed, and a real person who can listen, ask the right questions, and make a judgment call. That’s where automated voice agents and generic chatbot flows can backfire, especially for first-time buyers who are still deciding whether to trust you.

From there, we get practical about what large language models are actually good for in the shed industry: brainstorming campaign ideas, drafting and editing website copy, organizing messy thoughts into clean structure, summarizing information, and speeding up back-office work. We also get honest about the limits, including “strategy slop,” shallow advice that sounds polished, and the reality that AI is only as accurate as the data it can learn from. If shed standards, craftsmanship details, RTO explanations, and product differentiation aren’t well documented, AI tends to remix the same generic content.

If you want to grow with modern tools without losing the human edge that closes deals, hit play, then share this with a shed owner or sales manager who needs it. Subscribe, leave a review, and tell us where you’re using AI successfully and where you refuse to use it. What belongs with a human every time?

For more information or to know more about the Shed Geek Podcast visit us at our website.

Would you like to receive our weekly newsletter?  Sign up on our website: shedgeek.com

Follow us on Twitter, Instagram, Facebook, or YouTube at the handle @shedgeekpodcast.

To be a guest on the Shed Geek Podcast visit our website and fill out the "Contact Us" form.

To suggest show topics or ask questions you want answered email us at info@shedgeek.com.

This episodes Sponsors:
Studio Sponsor: Shed Pro

IFAB
Identigrow
Solar Blaster
Cardinal Manufacturing
Velocity 360

Transcript

INTRO 0:00

Hello and welcome back to the Shed Geek Podcast. Here's a message from our studio sponsor. Let's be real. Running a shed business today isn't just about building great sheds. The industry is changing fast. We're all feeling the squeeze, competing for fewer buyers, while expectations keep climbing. And yet I hear from many of you that you are still juggling spreadsheets, clunky software, or disconnected systems. You're spending more time managing chaos than actually growing your business. That's why I want to talk to you about our studio sponsor, ShedPro. If you're not already using them, I really think you should check them out. ShedPro combines your 3D configurator, point of sale, RTO contracts, inventory, deliveries, and dealer tools all in one platform. They even integrate cleanly into our Shed Geek marketing solutions. From website leads, to final delivery, you can quote, contract, collect payment, and schedule delivery in one clean workflow. No more double entries, no more back and forth chaos. Quoting is faster, orders are cleaner. And instead of chasing down paperwork, you're actually running your business. And if you mention Shed Geek, you'll get 25% off all setup fees. Check it out at shedpro.co/ShedGeek. Thank you, ShedPro, for being our studio sponsor and honestly for building something that helps the industry.

Cord Goes Solo With A Theme

Cord 1:34

Welcome back to another episode of the Shed Geek Podcast. I'm your host, Cord Koch, coming to you from a sunny, temperate, beautiful day here in Metropolis, Illinois. And today I am joined by me. This is going to be my first attempt at a monologue. So, please bear with me. I know that Shannon does these a couple times a year. Um, and they're always so enthralling, his thoughts and just kind of a live look into what's going on in the industry. And so, I'm going to carry on that tradition today with a look at, I think, what I will call an ode to the Luddites or an ode to those um who are not quick to adopt. Um, you know, so this will be in the particular um case or the particular topic of AI and what the possibilities are, uh, but how there has been some wisdom in um actually being slow to adopt some of those technologies as they bear themselves out. So, if you'll stick with me um and give me a shot here, I think uh I think it should be a good episode. Uh I'm gonna do my best to be entertaining for 30 or 40 minutes.

Contact Lines And Community

Cord 2:50

So, before we get rolling on that, just a few ways for you all to stay plugged in with the podcast. Of course, our call in line, Shannon Cell phone 618-309-3648. Our email info@ shedgeek.com if anyone wants to reach out via email. Of course, we do have the contact forms on the website, uh, the website itself, www.shedgeek.com. We always try and encourage participation um on the Facebook page, both our own and the group pages in the industry, shed sales professionals in particular, seems to be a really great resource that continues to um just offer a ton of support and a ton of insight from real industry professionals on what we are seeing day to day on the ground in the shed industry. Uh, and of course, if you have any friends in the playing community uh or anyone who would like to listen to the podcast via landline via telephone, that call in line, that call-in landline number is 330-997-3055. So pass that on to them.

Velocity360 CRM Partner Shoutout

Cord 4:06

And again, um before we get rolling here, I would also just like to shout out our exclusive CRM provider partners uh over at Velocity360, the exclusive CRM provider of Shed Geek Podcast and the Shed Geek Media Network. Um these guys just speak shed. I probably talk to uh Brandon and Joe once a week um just about what's going on in the shed industry, what's working, how they are getting their results. They've got a fantastic relationship going on right now with Shed Pro, where they're putting out some of those white papers that really discuss some of those results and go through um uh how they're able to deliver for you. They know what works, they know how to close, and they know how to help your sales team convert more leads. That's the bottom line. And not just convert more leads, but do it in less time. Uh you know, I think it's fairly common in sales in general, and particularly uh in our industry that many people worked for years off of uh yellow legal pads and spreadsheets and kind of had homemade CRMs, if you will. Um, but you know, Velocity 360 is really offering a turnkey solution, something that is easy, straightforward, simple. They do all the training, they make sure that you are up to speed on how everything works. And this is fully configured, right out of the box. Workflows, automations, reporting, um, that are all proven to work in this industry. Again, back to uh all of the relationships and partnerships that they have built uh over the last 18 months or so. They are just absolutely getting the results, and that is what that is what matters, and they do all of it with what they call their true white glove experience, meaning that um there is no there is no hassle. They take the handoff, they move your data, uh, they move all of your client information, your sales information over. You don't have to touch anything. A true white glove concierge service. Uh, so we appreciate those guys over there for that. Get in touch with them directly, tell them that Shed Geek sent you, uh, or you can reach out to Shed Geek, uh click the link in the newsletter, or give us a call, and we will be happy to make that connection for you. So

Ode To The Luddites On AI

Cord 6:35

having said those things, um an ode to the Luddites of as you all know, Shannon and myself and the Shed Geek brand, everything that we do, we have a tendency to always be pushing forward, pushing for the adoption of new technologies, of new ways of doing business. You know, I think that this bears itself out. Um, you know, CRM is actually a good example of it. Uh it bears itself out in um SEO and um Google business profile optimization in uh review strategy, like we've talked about with Gerald Rhodes before. Um uh in campaigns, right? How are you structuring your um Facebook campaigns versus Google campaigns? Are you running campaigns or are you reliant on simply posting uh you know posting your sheds onto Facebook Marketplace? I mean, all of these things have been um covered thoroughly and on podcasts over the last gosh, you know, two or three years even, um, going back to really the inception of the podcast um with Shannon and Kyle already six years ago now. We're in year six. And the march, the drum beat of Shed Geek has always been to say, let's try these things. This is modern business, let's catch up a little bit with the modern era. You know, this is uh an old school industry, construction. Um, you know, it's a blue-collar industry. Um, it's an industry, industry that um is still very much based on the inventory model. Um, you know, maybe that model is being updated and maybe it is being um viewed as just in time inventory and the most valuable SKUs. And we've talked about some of those things with uh CEOs of companies and production managers and how that side is changing. But today um I really want to give props to those people out there who push back and who might say no. Um Shannon commonly says that uh in the shed industry, if you say you've got a good idea before the person you're talking to even hears the concept, they already say it'll never work. No, no reason to try it, no reason to try it. And you know, of course, usually we're kind of saying that in a teasing way and saying, come on, guys, get on board here, um, walk with us, um, let us kind of let us show you, let us educate, um, let us inform, let us move those things forward. But specifically today, I want to talk about um AI and how um wise it is seeming um that much of the industry um has kind of pumped the brakes on that and has not been necessarily first adopters, and I think that's to their credit. Um you know AI has done so much for um the way that that um we do business here. You know, if you are in the content side of business, if you are in the um even to some degree, to a pretty good degree, the um if you're analyzing data, if you are looking at data sets, if you are um trying to discern uh some straightforward patterns that have good data on them, AI does really good. Um but there are also downsides and these things are bearing themselves out in the broader market as we speak. You see uh a lot of a lot of hesitancy. You see a lot of people talking about the AI bubble bursting, um, because frankly, all a lot of the capital that was raised for this AI expansion, for the data center data centers, for the GPUs, um, you know, paying NVIDIA to you know produce all of these chips that are a new generation of chips that can handle this type of really data heavy, text heavy um LLM models. And so, you know, I think it is it's natural for business to business buyers to be trend followers. It's very natural for it it's a it's the safe decision to make, I guess is the way maybe I should put it. If you are a uh if your business uh operates in a B2B space, um so you know, if that is those of us who are interacting with um you know uh software as a service, SaaS models, um, those of us who are interacting with professional services, those of us who are um, you know, having websites built, those of us who are engaging in uh SEO, all of those kind of non non-consumer activities that that our companies undertake. Uh the safe bet, you know, just sociology-wise or economically is to follow the trends. Um because what you don't want to be is left behind. And that's that is where uh sometimes you know our criticisms, I think, fairly come in and say, hey, this is a proven method, it gets results, you need to adopt it. Um in the case of CRM, you know, that allows your salespeople to handle, gosh, three, four, five, six, ten times the leads possibly, you know, depending on what the quality of leads are and how responsive and all that, but it is the safe decision to follow trends because you know it's rare that a person loses their job because they did the same thing the competitors are doing. Right? It's rare that if the whole crowd jumps on whatever the newest thing is, that anyone is really going to pay uh a heavy consequence for having jumped on, other than the time and money and you know the hours that were spent implementing whatever that solution might have been. But you know, it's just the nature of people, it's the nature of sociology that no one wants to no one wants to be the person who put up the stop sign um and said, no, we will not do that, we will not do as our competition does. We will purposefully allow for that trend to play itself out, and then we will decide whether or not it's valuable. Um sometimes that can really come back to bite you because if it is uh a market-changing event that's happening, um, companies can grab up market share and can move really fast.

AD-FIRST CHOICE METALS 14:28

When it comes to building something that lasts, every detail matters. At First Choice Metals, we proudly support the shed, carport, and pole barn industries with durable, high-quality steel roof panels, siding, and trim accessories built to perform and made to last. Every product we offer is American-made. With craftsmanship and finish you can see in every panel. More builders and homeowners are choosing metal roofing and siding for a reason. It offers superior durability, a longer lifespan, and improved energy efficiency compared to traditional materials. At First Choice Metals, our mission is simple. We provide exceptional products for your building projects and deliver service you can count on. Because being your metal building component supplier means more than just making a sale, it means being there every step of the way. First Choice Metals. Built for strength, backed by service.

Why Shed Knowledge Is Hard For AI

Cord 15:54

I think, you know, the modern sales tactics will wind up being video, but I guess to circle back to the kind of original point here, across the country, at this point in time, 98% of all organizations in the USA have implemented some kind of AI solution within their business. These are uh statistics that came out about two weeks ago, and I was I was looking them over and picked out a couple of the most interesting. 98% of all organizations have implemented some kind of AI solution within their company. Only 5% report that they have seen an effect on their bottom line. Now, I'm not saying that some of these things are not longer-term ROIs, but the way that it was sold, the way that it was pitched a few years ago, and just for full transparency, I was a chat GPT user on day number two. Um, I was following a podcast at the time, um, the all-in podcast, and they talked about this, they talked about the launch, and you know, I don't know exactly where I fell, but I would think I'm, you know, my account is probably in the first, I don't know, 100,000 or so people who started to use this technology on a regular basis. So this is not coming from a place of being a hater uh of AI technology or more precisely LLM technology. Um, I was an early adopter, and for the type of work that I was doing at the time, and even to a degree, the type of work that we still undertake here at Shed Geek and Shed Geek Marketing uh regularly, it's great. It produces lots of text, it produces lots of options, it produces lots of content. Um, I would say that for the shed industry, because there is not a whole lot of data out there in the world um that is particular to sheds, there's a lot of shed content, right? There is there's a lot of shed information in the broad sense, but as we all know, most of the websites, most of the brochures, um, most of the video content that is produced by the industry in the industry, um, it's all kind of the same. There, there, there are very few um places, and I've actually experienced this firsthand lately, um, as we undertake a project to start to put into place some shed standards um where we can reasonably grade, you know, an A-rated shop versus a B-rated shop. What does quality mean? You know, how does the how should a what is the craftsmanship and the materials that are required um to create the floor system, you know, a high quality floor system that won't sag, won't buckle, um, you know, that is that is uh appropriate for transport, all these things. Same with wall systems, uh roof systems, ventilation systems, uh, door and window uh framing, right? And I've experienced this firsthand because there are not really repositories of that kind of shed knowledge, um, even whenever it comes to more consumer-facing things, uh, RTO and finance and payment options, you know, really there are some surface level explorations of what that means, but deep dive comparisons, um, you know, uh cost calculators, you know, tools that tools that actually kind of help the consumer directly are few and far between. And so, you know, even in even in what we do uh around here, Shed Geek Marketing and Um Shed Geek Podcast and the media that we put out in the newsletter and everything else, AI can be helpful to take your thoughts and put them in a structure that is well organized and makes a good argument if you're making a good argument, tells a good story if you're telling a narrative, but that they require heavy editing because very simply the knowledge base to truly uh for that that LLM model um to truly understand or be able to grasp or have enough data points really um to put those sentences together in a way that is uh both truthful and insightful and rich in in its content, rich in its context, um it just doesn't those things are not readily available um as data points um in very or if they are, and I'm sure someone out there is saying, Well, gosh, we've got so much information on our website, and you probably do. I'm not saying that you don't, but whenever that uh LLM is scrawling the internet for the correct answers, for the answers that make the most sense, and frankly use those words in that order. Um, it's not really thinking, it's kind of like smooshing uh words together in a way that seems to make the most sense as it understands you know text based language. And maybe this is a good spot to just circle back a little bit and talk about what we have been referring to as a AI really is. And it's another acronym. It is uh they are currently LLMs, they are large language models, they 100% uh rely on data input, they rely on um both text, uh numeric-based um inputs. Uh if you don't have the data, if you don't have the richness of uh information, uh you really don't have anything with an LLM. Now, of course, the big ones, the ones that we're all using, um, you know, the Clauds and the Chat GPTs and all of the kind of very popular models, they've effectively been trained on, you know, most of the open source internet. Um, you know, so they do have lots and lots and lots of knowledge. It's not to say that they are insufficient in many things, um, and in well-covered topics, and particularly well-covered topics that have been well-covered since the advent of the internet, you know, um, you know, they're very good. Uh, and even in, you know, more historical contexts, as long as those individual events are well covered uh in on the internet, then they're actually very fairly nuanced. If they have 10,000 sources that talk about, oh gosh, I don't know, uh the Battle of Antietam, uh, which, you know, there's probably 10 million sources that talk about things like that, then they're gonna have really good insight. Uh, and they're going to be able to compile that knowledge in such a quicker way. I mean, you know, what would take a person, you know, weeks, months, years of their life to formulate uh a thesis about you know what to learn or what to take away from Antietam. Um, an LLM could do that and could even propose or modify thesis that it's already had as inputs um easily and can make it sound great. And part of that is a little bit of the trick of AI right now. It is that it can take very mundane thoughts and they can and it can make it sound like a term paper, uh, a true, a true uh, you know, a doctoral thesis, um, because it is really good at language, it is really good at presentation, it is really good at narrative structure. Um, and all those things are so valuable. Um it's good at organizing thought, it's good at it's good at taking the disorganization, and I'm having some of this right now, right? I mean, this podcast would probably be better if I would record it once and then shoot all that text uh into an LLM and then just read a script, right? And organize my own thoughts in a better narrative direction than me just giving my thoughts, you know, um or presenting my thoughts as they come to my mind with a brief uh bulleted uh layout here that I have next to me, right? So, it is good for so many things, but I think whenever we really start to get down to what are the core values of these AI systems, these LLMs, um, and how is that changing the sociology, the psychology of what the actual customer, consumer, the person we want to get in front of, because that is the core of what we're trying to do here, right? We are trying to sell sheds as an industry. We're all out here, whether we're talking about SEO and CRM and Google ads and new websites and uh mobile optimization for websites and uh race to face and um you know RTO and well, what is how about the premiums on that? What are we what are we getting back as a manufacturer? All these things just add up to selling the product and how to sell it the best and how to sell it the most efficiently. And so when it comes to how does AI play into this, two of these statistics really jumped out to

People Want People In Sales

Cord 26:42

me. How is it being in implemented and implemented effectively? So, the first statistic that just grabbed right a hold of me is client-facing roles in the United States, roles that deal directly with the client. The thing that AI is supposed to do, right? The thing that it claims to do, the thing that you know uh the advent of true AGI, artificial general intelligence, might be able to do at some point. But as of today, client-facing roles, real people talking to real clients is up 25%.

AD-THE CHALLENGER 27:27

Introducing the Challenger, the revolutionary new shed mover that's taking the industry by storm. The Challenger features a touchless design, ensuring it only makes contact at the end of the skids and the bottom of the floor joists. So rest easy knowing your gable walls will remain pristine. Worried about reliability? Fear not, the challenger is always up for the challenge and will not overturn or scuff your shed. Harness the power of innovative wireless technology. Remotely control the challenger via your smartphone. No cell service needed. And guess what? The cell phone remote comes standard with every Challenger at no extra cost. Featuring floating forks, a four-directional tilting mast, and sliding weight transfer, move the shed weight precisely where you need it to get ultimate traction. With self-loading capability, straight onto your truck. Say goodbye to the need for a hydraulic carriage. The Challenger makes your job easy and more efficient. Choose from two machine sizes, 10,000 pounds and 15,000 pounds. Perfect for any job. Visit www.shedchallenger.com or email us at info at shedchallenger.com or call or text 828-220-3508 for more information. Experience the future of Shed Moving with the Challenger.

Where AI Helps Without Annoying

Cord 29:10

Right? Across the country, across the country, the last three and a half or four years of AI implementation has led to more salespeople, effectively, client-facing roles. Now that could also be uh, you know, on the backside in customer service, that could be in uh, you know, in technical service or those type of helplines and things like that. But client-facing roles are up 25%. I think that really says something. I think it says something about not only the psychology of business owners, but it says something about the psychology of the consumer. You know, and it says something about the way that that business owners start to really suss out and figure out what is the consumer looking for. You know, because there are boundless, boundless um companies out there who will today put an AI, you know, facing, customer-facing um agent in between your customer and your company. Uh I wouldn't even want to guess. I there are probably 10,000 agencies, more than that, 100,000 maybe, out there in the world, um, you know, everywhere from you know, kind of outsourced uh Indian tech companies to thousands and thousands right here in the U.S. Who would be happy to place an agent of some kind, uh a voice agent, in between your customer and your company and start to do that that filtering process, start to answer what people would call level one questions, right? Uh who would you like to talk to? What are you calling about? I've noticed, and actually I kind of like this idea. I'm personally a Samsung user. I'm not an iPhone user, and I get uh I get teased plenty for still having not uh moved over to the iPhone, but the whole ecosystem, uh you know, I think it's in it intimidated me 15 years ago whenever I was initially making these decisions, and it felt strange to buy into you know the device that I use to talk and text on, then is going to almost require me to use those same branded devices in every other facet of life. So um, so I I've have never made that jump to Apple and to iPhone, but a great feature that that is now out there for Apple, they're using that very brief, um, it's almost like uh an answering service. It asks you who you are, and whenever you call someone, you're almost like leaving a short little message. And I think interpersonally, that's almost charming, right? If I am trying to um get a hold of Shannon, you all know that he and I wind up having so many conversations, so many phone calls throughout the week from every side of the industry, um, you know, from uh kind of industry insiders to all of the vendors, um, to the very tops of some of the uh manufacturing and RTO companies that that just want to know what's going on out there, which is a reason why, frankly, we started doing shed geek consulting uh is because it just makes sense to kind of have a beat on everything and be able to um advise on what the best incremental moves, uh moves for growth, moves for strategy. You know, we really have such a wide uh array of relationships and information that's coming in all the time, but that's uh that's a whole nother topic. But if I was trying to get a hold of Shannon, which he doesn't have this set up on his phone, but I do think it's a good idea in an interpersonal sense, you know, two peers, two colleagues, two people who know each other, have an existing relationship, you have each other's cell phone numbers. Actually, I think psychologically that's charming. Um, to be able to say, you know, I know that everyone in the shed industry right now is so very busy in this spring season. I'm just gonna quickly tell this AI agent um, you know, that I'm just uh checking in on whether or not we got that proposal over to XYZ customer, um, you know, give a short little thing, and then it gives him that summary so he knows whether or not to jump over, switch what he's doing, uh, jump on the call. I actually think that's kind of charming, but you know, the broader, the broader psychology here is that people are starting to get um very sensitive, you know, in the beginning, uh in the beginning, gosh, this was only a couple years ago, um, you know, that all these kind of things have started. Um, but I think that there was a lot of confusion and consternation. And um, you know, even to this day, all of us who have, you know, older parents or grandparents um who continue to kind of get tricked by um you know kind of AI generated reporting or AI generated images online or you know, whatever that might be, um, you know, all of us have had that experience still to this day, but in general, the psychology and the sociology of that AI in between, when you're thinking about a customer and a company, a customer and a product, a customer and a service, people don't like it. I mean, it's just that simple. It's one thing if it's convenient, if it allows for like I'm like I was saying, if you are in a position where you are not in a hurry, you have an existing relationship with that company, you know that you're gonna use them, um, you like the sales guys or the technicians, and you are just doing the formality of talking to this AI in between just to say, hey, this is Cord, uh just calling to check on what time the guys would be out to uh you know, service the HVAC system, you know, before it really starts heating up. Good, great. You know, like that that is not in any way ruining my experience, but for that first-time customer who does not know you, who does not have a personal relationship, who has not already had a positive experience. Now, if they're calling you, very likely they have, you know, done a little bit of research, they've entered a good Google query, they've started that process, you have some amount of validation online. So, you've got some reviews, um, you know, your website gives good information, blah, blah, blah. So maybe they've warmed up to the idea, obviously. They're calling you. Um, but you know, depending on just how relaxed that person is about their experience, uh, you know, this is all starting to backfire. Um, you know, cold calls um worked really well for a long time, right? I mean, you're talking about um 60s, 70s, 80s, that sort of classic uh family sits around the dinner table. The first few hundred companies that that start enlisting teams of people to go through notebooks by hand and call and catch someone right as they're sitting down to dinner and have a five or ten minute conversation with mom or dad about whatever the cold call of the day was. Uh, door knocking, you know, of course, Joe Ignis always tells the story of um you know door knocking and selling uh books and Bibles while he was in college. And um, and you know, that was effective. And in small, you know, niche kind of ways, that can still be effective today. But my point is in general, the sociology of the consumer started to say, hey, look, this is no longer novel. This is no longer something I enjoy. I no longer enjoy just having the in that case, maybe it's the attention of a company who wants to sell you something and would consider you a good customer. Um, or in the in the case of AI, and particularly AI voice agents, it's no longer novel to people. We've all heard them, we've all talked to them at this point, we've all had some kind of a throughput interaction where we are trying to get a solution, trying to get somewhere, and it just feels like frustration. So um that brings me to the second statistic, first statistic, client-facing roles up 25%. People want people. Companies are hiring people to talk directly to their customers. Non-client-facing roles, right? Roles that are um data entry, roles that are um design, roles that are copywriting, um, roles that depend on, you know, frankly, some back-end creativity and some back-end um, you know, data entry analysis, um, the straightforward roles that I think we can kind of label as boring, not to say that design is always boring. The creative process of design uh is fun, right? But then the designers themselves who then have to pull you know 10 different things around uh or not 10, you know, a hundred, uh 500 sometimes, right? Who are making a design actually look right in whatever their photo or video editing software is, it's boring work. You know, it is it's work that you can throw on your headphones and really zone into, but um it's work that is done at you know a certain percent of quality. I'm not no one's gonna say that that AI in every circumstance is producing the quality of work that a designer would by just putting their hours in. But what it allows for is for uh a company or an agency um or um you know a data entry um office, you know, inside of a company or division inside of a company, whatever that might be. What it allows for is instead of having four designers or four data entry clerks, right? You wind up having a designer who is able to work faster, work at a higher volume, who is able to tweak when it comes to the copy side of things and all the text that's being generated, same thing. Like you're good copy writers, start to be good copy editors, you know, you start to be able to produce a much bigger amount of content at a much faster pace, and then you just need the watchful eye that is making sure that we are whatever it might be, staying on brand, not being offensive, um, saying things truthfully and honestly, you know, all of the all of the things that you know an editor already does anyway, you're just effectively taking away the most tedious parts of those processes and allowing them to be overseen um by fewer people at a higher volume, at a better, uh uh a more efficient scale. So, you know, when it comes to um when it comes to the shed industry, when it comes to implementation, I think that the text-based um the res the auto replies that are happening in a text-based function that are limited and that are well trained, which of course depends on how much data, you know, how much text um this is this is one of these places where um you know every company that had that had been using a CRM or you know a hub spot already had some kind of call recording or voice-to-text system that was just being data-based the whole time is probably ahead of the game. Although, as we know, things change. Um and those calls from four years ago where you're offering solutions that no longer exist, better be careful, right? Um, or offering solutions that are now outdated, you've now in-house that, or you've now um decided that that's going to be a contracted service. Um, you know, again, heavy editing, right? It requires, it still requires a human mind, human discernment, um, you know, and human decision making um to be able to say, no, this really truly does represent our company.

AD-LUXGUARD 43:13

Hello, shed sellers. Let's take a moment to discuss the shed customer and meeting their expectations. I remember growing up in the neighborhood where a certain percentage of the houses had well-manicured lawns and well-manicured homes, these were the type of individuals who felt it was important to purchase a well-constructed home or vehicle or maybe equipment to help maintain the quality of the item. As shed manufacturers, we seek to provide a well-built quality shed. We want the customer to feel satisfied that their hard earned money has been well spent on a product that will last. At LuxGuard, we believe adding high quality rubber flooring to your line of sheds makes sense to the customer and adds value the customer can appreciate. With each year, sheds are becoming more complex. The customizations we are seeing are virtually endless. LuxGuard not only gives a complimentary aesthetic appearance for their shed, but also protects the floor from spills and keeps cleanup simple. Offer your customer the customer service they seek with LuxGuard. At LuxGuard, we are committed to delivering exceptional customer service and innovative products to help our customers achieve their goals. We strive to meet the evolving needs of the customers. To speak with one of our ready-to-serve customer product specialists, simply call 336-468-4311. To see our product and view an installation video, just visit our website at LuxGuard.com. LuxGuard, the floor that lasts a lifetime.

Harvard Study And Strategy Slop

Cord 44:42

But you know, in in those limited use cases, you should do it today. In a use case where you need to produce um brand copy. You know, you need to uh we're dealing with several customers right now, and um, you know, when you need to produce ideas, when you need to brainstorm, when you need to take ideas that you already have um and have them put into a better context, but what these what these LLMs can't do is actually use discernment and decision making. Uh Harvard study uh came out saying that basically um the business advice that is always given, you know, if you're asking these LLMs for strategy, for discernment, for um effectiveness when it comes to like true strategic thinking, they're just mimicking it, right? Um it turns into uh oh gosh, they had a word for it. Um oh now I'm not gonna think of it, but it had the word slop in it. So it was similar to how people call you know the AI videos and stuff, you know, AI slop. Um they had coined a term for that, but uh it's basically doing that in a corporate speak version, right? It's it talks about partnerships, it talks about collaboration, uh, it advises whatever it might be, um, you know, whatever that that sort of uh you know vertical integration or you know, it says all the right things, but there's no discernment going on there. Uh, whenever they ran a mass survey, you know, 10,000 different uh different business um scenarios, it always landed on the same strategies, which are just pop culture strategies, right? It is just what's being talked about the most. It is not actually discerning what the advantage is um to owning that side of the project, uh, to um you know, owning a company that is in line with your distribution model. It's just it it's not there. Um, you know, do we ever get to artificial general intelligence? And you know, what does that what does that mean for for people and and what does that mean for our thoughts about what it means to sort of be conscious? And you know, there are a lot of implications there, but the long and the short is nothing out there as of today is generally intelligent, nothing can show true discernment, true uh consciousness, obviously, um, true strategic thinking. Um so we I just want to give a round of applause to everyone in the industry who said let's hold up, right? Let's let's pump the brakes on this. You know, obviously this is a big trend and it is it is moving the economy, it is promising big things. But let's just let's just see, let's be conservative, right? Let's let's let's let our nature win out on this. And and I think that by and large, those people are being

AI Bubble Signs And Copyright Risk

Cord 48:20

proven right. Um Sora, the video uh you know production app that was so popular there for a while and overwhelmed our news feeds on Facebook and the scroll on Instagram and all the apps that we like. Uh, even YouTube, you know, had gotten to where it was all faceless, all faceless, all um AI, you know, voiceover. Um, and if you'll notice all those things are are moving back. Sora got shut down, you know, copyright trademark issues. Um, you know, it's using uh IP um that you can prove that it's using because it's directly sourcing it uh you know from the internet from those from those um you know open sources that that contain IP within them. Um but if you reproduce that IP, obviously uh you could have lawsuits on your hands. So, uh you know Sora is out of business. The outlook for AI um that was so I mean that raised trillions and trillions of dollars and has trillions of dollars invested into uh the market cap of all of these publicly traded AI companies uh that promised all these data centers and you know to this point, you know, not a single gigacenter, one of these huge, you know, um um acres, you know, acres and acres of AI processing power that have been proposed and have been um money's been raised for. I mean, the the capital has changed hands. The money has gone right into the stock price, um, not to mention and some of these companies it was that they were in investing rounds. So, some of that money has gone straight into uh the bank account of those companies, and maybe they're still private, maybe they're still public. My point is the projects aren't getting built because the applications that were projected, the ability to truly have a human conversation, or at least a conversation that was novel enough, that felt human enough, that had real human discernment and maybe even real human moods, right? Maybe even maybe even had a different intonation in their voice on a day, you know, on its voice, on a day whenever um it had had to deal with three tough customers or whatever that might be. You know, those kind of of projected outcomes simply haven't started to add up. That's not to say they won't. I'm not sitting here saying that there's not a future where AI is integrated into, you know, whatever it might be 60, 70, 80, 90 percent of each kind of um division, you know, of a company. You know, there's some kind of an AI backbone that is working there um to make everything better, make everything more optimized, um produce better content, blah, blah, blah. You know, all these things. But for today, you know, the lesson to take out of AI continues to be use it where it's useful, use it for what it's good at, don't overinvest. And and really the backbone of your company still needs to be you. Now, it needs to be you in a way that can be um converted to data. It needs it needs to be you in video, it needs to be you in text, it needs to be you in a guest blog post, uh, you know, on the Shed Geek newsletter, right? Like it needs to be you. There needs to be lots of it because for these models to continue to get better, they're going to continue to need to learn and learn and learn and be able to smush these words, these uh this language that we use into better and better and more accurate little packets that we can then use uh you know to talk to customers or whatever it might be.

Privacy, Data, And AI Limits

Cord 52:52

I was actually saying to Shannon the other day, you know, for um for these language models to be what they say they are, what they could be. And they could be, you know, if it had every piece of data that existed in every conversation, not the conversations we're having here that goes out, you know, over the internet, you know, uh into your Apple Podcast or your Spotify or you know, in video on YouTube, but but every single conversation, if they could all be picked up on these things, if these microphones, and we know that this happens, right? There's there is a system that you know is not openly talked about, you know, where these apps are gathering voice-to-text information and then are specifically targeting consumer keywords, right? You know, um, there's so many examples. I mean, I could say uh um, oh gosh, I don't know, Pampers, Pampers, um, diapers, diapers for babies, right? Like I can just say it aloud, and all the apps that are running in the background here are doing some level of voice to text that is then being captured either directly by the um by the platform, the actual app owner, or is being captured in the throughput. I mean, for me, I right this moment am not um am not attached to the Wi-Fi here at um here at the shed geek offices. So my phone is is going through the the wireless network. Of course, all the the data that I'm all the my voice data is being transferred um you know through a wireless network. And so, you know, there's even a potential that that there's some throughput that is being captured. You know, it's all kind of backroom type of stuff. You know, it's a lot of different deals um where at a high level, you know, consumer reporting agencies have data deals with the apps themselves. There's probably some black hat stuff going on that then gets fed you know into specialized consumer databases um and allowed then and is then purchased on, you know, uh I guess if some of that happens on the black market, you know, on outside of what is kind of legal or expected, at least with the protections in the US and the UK and Europe and a lot of the Western world has those protections in place. But my point is, I guess what I'm getting around to is I'm sorry, that was a bit of a divergence onto how we wind up seeing uh diapers, you know, now that are going to come across my uh newsfeed or whatever when I open up Facebook. But what I'm saying is is for these LLMs to truly have the answers, have the strategies, um, have a higher level of discernment without having the true AGI component. Theoretically, it's possible we would just need to willingly feed every private, non-internet-based, non-video call, non-anything, every private conversation we have, whether it's in our office, in our car, or in our house, we would just need to willingly give it to them, right? And just let them train off of every conversation, and it would start to pick up on um what the real strategy is of pick a company, right? It would probably be disturbing uh because we would start to get answers, you know, that in a lot of ways, you know, uh cut crossways with with what we kind of think of as conventional wisdom. Um, whenever, you know, I mean, uh it lays out why XYZ company is aiming to uh support uh a dictator in another country, right? Because they advantage them, they provide enough stability through their dictatorship that cobalt or lithium can be mined. I mean, some things that would probably be a little unsettling for the the uh big data set to be able to pull down, but I'm really just saying all of this to reinforce the fact that these programs are limited. They are limited to what you put into them, they are limited to how you train them, they are limited to not having actual discernment, which is so so important in customer service. You have to be discerning, you have to be um you have to be able to hold fifty different ideas in your mind, you know, and decide what the best path for that customer is. Um and and that can't be faked, not yet. You know, maybe one day it will be, but these LLMs just simply do not have the capacity um to do that work.

Practical Takeaways And Invitation

Cord 58:26

So, I will leave you on this thought. Um be real, be genuine, use people to interact with people when it comes to your cut company and your sales and your customer service, um, you know, integrate it where it works, integrate it into the back end, integrate it into those response campaigns. You know, if you're just if you're trying to get a um you know, a customer to be responsive, if you want to have some of that AI interaction in those first couple text messages, um, because it just makes sense to have to have a limited, you know, three-text response or whatever that might be, just to try and get your customer going, um, rather than relying on your salespeople to be able to text back all of your 50 leads for that day within you know 30 seconds, which is basically the expectation that we all now have if we're having an active conversation through text messages. We think that 20 to 30 seconds is basically the amount of time that it should take someone to text back. So like allow those, allow those limited use cases where it works, but you know, don't get out over your skis on this. And I know that that um I probably don't have to preach that to this industry. You know, this industry does want to wait and see. Um, but I just I guess you know, I wanted to um for the whole shed geek brand, for the whole network, we can we can certainly be critical sometimes, and we can certainly um push people to adopt the new things, to get the new solutions, uh, to go grab up the results that are sitting there waiting. If we would only just make the decision, if we would only just take the step. And so, in this um in this way, I just wanted to say uh I wanted to eat a little crow for the Shed Geek brand, not that we've been pushing AI or necessarily been out here just, you know, uh uh encouraging everyone to do as much AI as possible. Um, but I do want to eat a little crow for the narrative that we have sometimes, which is, you know, we all just need to adopt this new stuff. We need to move to the new era. We need to uh get on the bandwagon. Um, I I think in this case, um, you know, like I said, a round of applause, um, um, an acknowledgement to the uh the slow movers, to the Luddites, uh, to the the um those who are slow to adopt, you know. Um kudos to them because I think in I think that there is much wisdom in in not getting too far ahead of yourself on this particular topic. So, I hope you enjoyed listening to me for, oh gosh, I don't even know. It must have been 45 minutes or so, and I'm very self-conscious about um all that rambling. So, I hope that you enjoyed it. I hope that it furthers the discussion on what is valuable uh in these AI tools and AI programs. Um I hope that it, you know, I hope that it even generates some pushback. I hope that somebody out there says, hey, I've got an AI agent talking to every customer, and we're closing 30% of all responsive customers, and our AI agent is doing it all. You know, um, like I I want to hear that. Um and I want to hear how you did it because my my uh inkling would be that you had really great data to start with. I mean, maybe I'm wrong. Maybe maybe you trained it on generic data and you found the right data sets, but but I want to hear that too. I want to hear that I'm wrong if I'm wrong. Um, but in any case, I hope I hope that uh you've enjoyed listening to me for a little while here, and uh and I look very much forward to continuing this conversation. Catch me next week on another edition of the Shed Geek podcast. And until then, uh give me a ring, shoot me an email, uh, reach out to me. I'd love to hear from you. Thank you all.

OUTRO 1:03:08

Thanks again, Shed Pro for being the Shed Geek's studio sponsor for 2025. If you need any more information about ShedPro or about Shed Geek, just reach out. You can reach us by email at info@shedgeek.com. Or just go to our website, www.shedgeek.com, and submit a form with your information, and we'll be in contact right away. Thank you again for listening, as always, to today's episode of the Shed Geek Podcast. Thank you and have a blessed day.