AI in Construction for Business Leaders to Fix Broken Processes & Avoid Deskilling - Jonathan Marsh

Speaker A: Welcome to the construction technology news of the week with Jonathan Marsh and Travis Voss. From groundbreaking innovations to industry shaking updates, this is your construction technology news of the week. Join us with our guest of the week as we dig into the latest news shaping the future of construction. That's right, guys. We're going to be dropping in these news episodes when we feel like it's worth it. And uh, hey, it's 2026. We're here. We're going to have some fun. I've got John and I've got Travis with me. We're going to go through and talk to you about a little bit of everything. Mr. Travis Voss. How are you doing this morning?

Speaker B: I'm doing really well, really well. Enjoying uh, another fall spring. So uh, we'll get some cold and snow again and then uh, hopefully spring will come around a couple of weeks.

Speaker A: We're supposed to be back in winter here, which we never got there. So I'm hoping to be a uh, brighter, bushier tailed person because we get snow and I can get out, but we'll see.

Speaker B: You know where I live, Northern Illinois, not exactly known for mountains and skiing, but my daughter had a ski trip yesterday in 65 degree weather.

Speaker A: Well, at least it was in 60 degree weather in Colorado, which we were having. So let's hope, uh, hey, breed those new skiers. We need them because, you know, I love me some skiing. Jonathan Mars. How you doing man?

Speaker C: I'm doing good too. I'm in the middle of my Indian summer. I'm loving it. It's a little bit sloppy here because everything's still melting off. But boy, that was a long like, since Thanksgiving it has been below freezing here. So like the ground's really, really frozen deep. I found out that I probably have some uh, water issue with my driveway because it now has a giant crack down the center of it. The joys of living in places where the ground decides to lift up and drop down every year. You know what I mean? It's always interesting.

Speaker A: Yeah, it is. It is man. All right guys, you know the drill. We're back on YouTube. Check us out at YouTube.com forward/the content crew. And you know, I'm noticing not a ton of you are heading out there and signing up for that emailer. And you know, you might want to because there's some new things coming out on it.

Speaker C: Maybe some things you can't even find on the ship.

Speaker A: And hey, follow those social medias, follow these guys. They're a great follow, great place to keep up on information. Follow The Comtech crew. We appreciate it. Are you tired of all the talk about solving the industry's problems and you just want to roll up your sleeves with the industry's best and brightest? Well, the CPC's returning to Denver, Colorado from April 8th to 10th for the 5th annual AEC integration Summit. That's right. Participate in one of the four teams topics from frictionless change orders to AI for issue resolution. Identify root causes of these shared pains and discover new approaches to solving them. Learn more at constructionprogress.org AECI Summit 26 the problems you learn to solve could be your own. Join us there. All right, so News 2026 kind of, uh, I wanted us to grab a few. These guys grabbed a few to talk about sort of trends into the future. Of course we have to start with AI, but I always like this. When someone brings me an article, I like to give them a little shout out. So this is to our boy, Rob Sloyer. I caught it on his LinkedIn feed, really dug into it and can't say I fully agree with the whole thing or the way it's written. We were having a lot of, uh, interesting discussion off camera. So it'll be fun to see how much of it I can pull out on here. But this is from Foundation Capital dot com. It's AI's trillion dollar opportunity and the title is called Than Context Graphs. And I think this is interesting because the conversation has been around, you know, is AI going to kill the systems of records? But you know, we know that that rise of enterprise software, whether it was your salesforces, your workdays, your saps, you know, all of those things were these systems of records for your data. And then it was, you know, trying to analyze and get analytics and grab things out of there. But that, you know, as we go into this AI world with agents and other things, they can't really just connect system to system and actually do much with that other than maybe move some data, prevent some duplicate entry, you know, maybe fetch you some things if you needed it. And the idea here, that he's stating, and I want to be very careful here on he's stating that what we need is also to connect those agents to, to what they're calling the decision traces. So think about this. If you're selling something and you're tracking your client and salesforce and you know, you're on a conversation with a marine or slack with your boss and you're then giving them a discount because of, uh, something else, they're going to do it's a discount above say, uh, a trigger that you could usually give them. That's happening in other systems and that decision's being made, but it's only coming down in Salesforce as you gave them a 20% discount instead of a 10. And you don't know why. Well, this is saying, can we capture all of that in a way to persist it and use it in the future? And in this respect, kind of in saying they're going to create an enduring layer that will help automate those decisions for you in the future. Now I disagree with that end state, but I do think it's worth the conversation around it and around to me, what it, what it sees. So Rob, what I saw from this when you posted it was think of the complexity of every business and how many businesses out there really even understand that complexity. Where those decisions are made, how they're made, why they're made, where they're triggering, where they're going on. And to think that, you know, you can just automate that and be done, uh, that's too far of a jump for me. The fact that you can connect some of the tissue together and maybe use the agents to understand that might be interesting or to map your process might be interesting. So, uh, I mean that's sort of the preface of the article. The article is saying, you know, that it's this, the context graph is connecting to that slack, connecting to that email conversation, connecting to those trigger points and automating it. Right, John? Like there was a little.

Speaker C: Normalizing it.

Speaker A: Yeah, normalizing it, automating it. Which, and I will say this, and I completely, I've said this for a while and it's, it, it goes to sort of why I work. Where I work is I think anytime you're thinking about taking humans out of the decision making loop, it's a bad idea. I think it is. But I think providing. And you know, John, we got this out of mep. You know, one of the things we heard over and over on our session was time to be in that decision making chair. The opportunity, not that we, that we don't have enough of them, um, the people and the opportunity time and what it could drive, productivity and standardization perspective.

Speaker C: So yeah, and really what it is is this question of scaling. Okay. So when we look at our foundational pieces of software that are software of record, they've kind of scaled as far as they can with the data that we give them. Okay. So we've given them enough data to kind of get where they're at. And you Know when we're at MEP Innovation, we, we're talking to all these companies that want to scale and what do we tell them? You got to write down your processes in detail and make up maps around your processes in detail. Why? Because we want to inform your future decision makers, the people that will take over your post, what to do when they're in these situations, or at least what you did when you were in those situations. And you also want to tell the rest of the organization, hey, this is how we want to do things. And I think that's what this is about. You know, like we're in the beginning of our data journey, most of our companies, and we're catching the data that's easy to catch, that has stuff set up for it, you know, in Salesforce or wherever. And now we're saying, okay, we can industrialize our use of all the files with that and the use of the data we have. But there's a real drive. AI is a drive to industrialized knowledge work. So industrialized decision making and knowledge working. And so if you're going to industrialize knowledge working, scale knowledge working, you start with these diagrams. Like he was stepping around it very softly because I think that this is a, has a lot of problems to it from a, ah, human perspective and even from a business perspective. You know, we talk about kind and wicked environments. The environment's too wicked, uh, for normal software to work. But an agent they think could do it right. An agent can deal with all these extra things that come in off to the side as long as we give it some examples. So we'll see how far we can get to the examples. Because you know, construction's awesome in that some of the stuff that happens on our job site only ever happens once. You only ever gave that guy the 10% discount once. And if you start to base systems on that, those systems go haywire. So there's always a, uh, upper level, I think. You know, Jeff, you talked about this in this article. Gives you some really good sort of patterns to look for within your organization. Here are your blind spots. And they define these really well because they're like, okay, here's where we cease to be able to make good decisions with AI or automation or anything else. And here's the blind spots, here's how to look at it. Here's where there's that tribal knowledge left within the knowledge workers of your company. And you got a lot. Because Travis, you know, I'm like going to sheet metal shops and I'm like, so, uh, so what's the tribal knowledge inside the sheet metal shop that makes this place work? You know, because it's not just these processes I see on the wall. So what are your thoughts on it?

Speaker B: Yeah, I was, I've been jotting some notes because there's a lot of words that I picked out and I want to go back to the uh, context graph. And I think that's, it's kind of cleverly named and we've talked about this a lot with, with AI and including agents is a lot of times AI is, is missing that context, right? It has the knowledge, you put it at your different systems, you put it at an LLM. It has the data, it has the knowledge. Even if you start to uh, to include Slack conversations or email conversations that maybe point to why a decision was made. The context graph still, again, I guess this is probably my views, like there's still context up here in my head. There's context that, that exists in your tribal knowledge, exists in your, in your thought leaders or your thought workers. That isn't gonna easily come out. And it's not, I mean some of it may be self, um, preservation, but it's, it's not always that either. There's just things that happen that we all do day to day and if you had to go back and explain why you did it, you're like, I don't know. I, it's just the choice I made, right? It's just the decision I made. And companies run like that in good, good, bad or indifferent. I think that we all, we all love processes and agents love rules. Right? And that was another point that I was going to kind of hit here is to your point of, of things that only happen once. Um, now it's important a lot of these agents are hopefully smart enough to bubble up the exceptions for humans to address. But AI is still software. Software still likes rules. And there are parts of our world that, yeah, the rules are kind of there, but they're a little more fuzzy. Uh, so I do think there's benefit here. Anything that forces you to look inside your business and look at your processes, look at what's going on, try to capture some more of that tribal knowledge. As um, again, we'll touch on an article later that talks about workforce. Like as that workforce becomes either shrinks or we just get new people, you have to bring those people up to speed somehow. So anything that takes you through that process I think is, is beneficial. And Jeff, I'll circle it back to you now where you're talking about having A human in the decision part. I think that's still going to be crucial. I do think. Um, we were talking before, um, we got on the call about how sometimes it's just really hard to automate what you think is a simple task. And sometimes automation can go really wrong and it can really screw something up because it's just happening all the time in the background. So as long as these systems still have a human in the mix, I do see that there's some augmenting effects that could happen.

Speaker A: Yeah, I love that you brought it back there, too, because we talk about this over and over. Process, process, process. Right. Mapping it and understanding it. But also, you know, that's the, that most companies don't have that done and this is where they're going to fall over and this isn't going to fix it at all. But I think what Rob's point too was, is like, it's kind of nice to see where all that's coming from and know the scope of where things are happening and to gather that over time. See, for me, it's more about gathering and analyzing still and pushing your businesses forward. But this is where, like, this is also assuming that you know how to do the business, John. You know this. You go everywhere. Most people don't. Right. They, they, they knew what they were doing when they started, and they've grown to a point or gotten to a point where they're outside of their own expertise. So you can't solve a problem you don't know you have. And if you throw software at it, it's going to kill you.

Speaker B: You know, it's m. A little bit of a microcosm. But your point earlier about why did you give a 20% discount over a 10% discount kind of dovetails in with my point. I don't know. At that moment in time, I just decided that they deserved a 20% discount. How does software suss that out?

Speaker A: Yeah, and maybe, maybe it would be nice to track that 20% discount because I gave it to them because I thought, um, they're in a pinch right now and I could, I could gouge them. Cause the data will tell me I can gouge them because they need my product. But I'm thinking I'm, uh, going to be on the other side of this in a year or two. I'm going to give them the 20 and not gouge them when they're in a tough spot. And then I've created a relationship. Right. Because it's a relationship business. The computer's not going to tell you that now, what would be nice to know is, like, how many times am I right about that? Et cetera. But also, I don't really care in some respects as long as I'm coming out ahead, because then I'm doing the right thing, too. And by the way, uh, rules. The computer doesn't understand empathy and compassion. It does not. And it will apply a rule that could really be awful, and we don't want that. So I love this from the context graph. Like, I want to steal that part, but I'll spin it too. And, uh, I've asked him if it's okay, and he said it is. So we're going to start this debate. Thanks to Trent Line and back here for Is it the new SaaS, is it software and a service, or is it software as a service? But either way, I think the value of software here is your business. It's not the software itself anymore. It's whether or not the people providing it to you and coaching you can help you with your actual business and your processes and get to where you want to go. They can't tell you where you want to go. You know, John, you can come in and I might say, hey, Look, I, uh. $250 million, and this is where I want to stay. I just want to, you know, make more money, uh, more profit. I don't want to grow above. Or, uh, you can't tell me that when you come in. Or I could be like, hey, I'm 250. I want to be a billion in two years. Let's go.

Speaker C: Absolutely, man. I always teach software as part of your workforce. I teach it as like people on your workforce. This is the same thing. This is the workforce that you're hiring from someone else. Just realize that they work for somebody else. Like, this is a workforce that's just subbing. You're subbing out whatever you're doing here to whatever is providing it. And I think there's a lot of room in this to, to really sort of grow your business along with the software. So I think, like you said, you have to do some of these things to scale it all. Just be cautious.

Speaker A: Well, and I want to bring this back. Right. It's great people. And so I'm going to dovetail the Travis's article. Travis, you had somebody recognize, and I want to hit that before we get into your deep article.

Speaker B: Yeah, yeah. I just want to touch on this. Um, we all know and love Angie Simon. Her, uh, and Rick Hermanson created the Heavy Metal Summer Experience. Um, and they were recognized at this year's 2026 MEP Innovation Conference. Angie particularly as the, with the Industry Advocacy Award comes with um, a donation to, from the event and our three associations to continue this uh, this path that she's on. Um, and we talk about workforce all the time in this is Angie and her team are up there. They're, they're trying to solve it right. They're going to. I've already lost track. They're already over going to be a hundred camps this year. Like I think she said 1200 kids if I believe right, if I remember right in her acceptance or thank you speech. And even right out of that um, just being at that conference and being able to present there, she, she told me she had about four other contractors reach out. So she's a worthy recipient. It's a great organization teaching these kids, getting these kids at the, at the high school level, getting them into a contractor shop, getting them into a training center. Show them the different trades and the opportunities that they have. And some of her, you'll have to watch the video if you haven't seen it from the conference. Um, some of the stories that she tells of lives that she's helped. It's very touching.

Speaker A: It's people, right? And HMSE for people is the Heavy Metal Summer Experience. It is a program that is created by Angie and Rick Hermanson. Like you said it grew from like one or two at Hermanson and I think at ah, ah General. Not general sheet metal but her. Was she a general.

Speaker B: No, Allied.

Speaker A: Those two Allied. And it was, it grew from those two and now it's you know, like you said it's going to be over a hundred camps, over twelve hundred kids and it, it brings them into shops and it gives them hands on experience with the tools themselves and it funds it and they all get a set of boots and it, it's showing the alternative career. Well, should we stop calling it that?

Speaker B: I think we should.

Speaker A: A career path in the trades with your hands. So uh, again I'm going to try and change my language there. So yeah John, you, you, you know you got to hear Angie. I'm, we're just a huge fan. She was on the show so guys go listen, you can search her up. She was on our show so it's

Speaker C: awesome and she has some great stories of the families that are impacted and I think that's where it really comes home is those people that actually go to the camp that it changes their lives because it's not just being successful at learning something with your hands. She was up there and she said one of her goals was not just to get people into this. She's like, no, I want to get shop classes back into high schools. She said, this is my goal. My goal is to get shop classes back into high schools. And I think it was on the, it's the empowerment, the empowerment you get knowing that you can create things with your hands. And you see it in the pictures that she shows of the kids and the letters she reads from the parents. This is something that changes people's lives. You know, we have a big focus on producing a lot of paper in high school. You produce a lot of paper, man. But, but you used to. I mean, I remember shop class. I got to build robots, I got to program robots, I got to do carpentry, I got to do, I learned a lot of hands on skills that sort of ended up shaping my life more than, more than the paperwork did. So I, I, I really hope that she, I hope that someday she is able to, to sit back and say I was able to start this many shop classes. I don't know if she's realized that she's probably what, like five, eight thousand different kids she's going to touch between the beginning of this program and now. And those lives are definitely affected. So she's an amazing person. This is an amazing program and every kid should be in it.

Speaker A: Please go check it out. Please follow Angie. And to your point, how great is it when someone wants to basically eliminate themselves by making this go back to all the schools? Oh yeah, It's a lofty goal and it's uh, going to take a long run to get there. But we hope she gets there. So Travis, what is your, uh, other article?

Speaker B: Yeah, so my second article, we're talking about giving shout outs. I got this one actually from your good buddy the Contact Trio. Oh, gee. Josh Bohm shared this with me because he knew that uh, I was out at the HR Expo this year. And so the article says what AHR Expo revealed about the future of building automation and controls and the business imperative behind it. Now this goes into what we saw a lot at the show. There's a, there's a lot more systems integrations, there's a lot more of these, it OT platforms. There's a lot more of these, um, you know, high efficiency lighting, all these systems that we've been talking about and we've been installing for these over these years that, that are meant to kind of tie different portions of the, of the building together to help you run it more efficiently and whatnot. And the follow on conversation was, was what really was cool with me. M When Josh and I talked is we all know one of the biggest limiters in the kind of the, the AI explosion is, is power. Right. A lot of energy. Right. So we're seeing with a lot of this imperative. Yes, it's a strategic imperative for the owner operator of the building. But um, it's also an imperative for the industry as a whole because as they take these buildings and work to make them more energy efficient, they're shoveling that. They want to shovel that energy back into the grid. Right. So that the, these data centers and AI platforms can use it within construction. That provides a lot of opportunity. Right. I think that Josh, I don't remember where he cited this from, but he said there's, there's like 129 million like upgrade projects that are available as we, as these buildings continue to get modernized and pushed towards this. And it's not maybe the, the green or the energy efficiency type push that we were getting before, but now it really is squeezing energy out of these buildings and pushing it back into grid so we can power more of these data sets. What do you think, Jeff?

Speaker A: Yeah, I, I, I mean this is, this has been near and dear to Josh's push. And it's, you know, from his days, back in our days we used to hear it now where he's added electric. He's really pushing this idea on the future of electrical contractors. And I think he's right. I, I have to say I've, uh, you know, I've spent quite a bit of time researching and, and working with him. And it's this systems integrator idea, right. It kind of goes to my article on the context graph. Like it is a context graph of your building and those electrical contractors or service providers that can understand the systems and learn how they integrate and learn to create those efficiencies. So what this means to everybody listening is like a lot of the systems inside of a building now are electrified and they're data centric and, but most of them don't connect into one another. A lot of the old building controls, those systems don't work with other systems. And now it's kind of like this. It's like your ERP doesn't work with your new field technology or your new scanning technology. And it's the same thing in a building, but a building, you know, the construction's 20%, a building is 80% of its life cycle. So a systems integrator can not only make money today, but make money down the road and drive efficiencies back into the power grid. You know, look, we went through what we went through with COVID right? Like, we can't ignore that that happened and that, uh, there may be a time and space where our environment changes again. You know? Well, how are you going to do that? A systems integrator is going to be one of the first people you have to go to to understand what that change is. Do we go, you know, maybe it's not. Let's not doom and gloom it. Maybe we all of a sudden realize that open offices were really a bad idea. We got to put a whole lot more walls up and give people space. Great. How do we do that? Systems integrator is going to have to know. So, just trying to make light of it. But John, I'm wondering what. What are you thinking? Because I love this super.

Speaker C: I'm kind of, uh. I'm sitting in a weird spot in the middle of it M. Especially on the electrical side. So we are starting to see now on the electrical side, monitoring of equipment becoming requirements like it used to be. Recommendations NFPA 70B, which is about your electrical maintenance plan, used to be a recommendation. Have this because it's going to make it safer for electricians to work in your building, and you're going to have less downtime and all this other stuff. And then about three years ago, they're like, nope, it's a requirement. And these systems are getting more and more complex. They're getting more and more dangerous. And so they're like, you have to have an, you, uh, know, electrical maintenance plan to go with this. We're going to make it a requirement. And IoT is a big part of that. We can look at that switch gear that's going to go out and when it hits a certain temperature, as a certain IR infrared scan, when you done, we can say, oh, wait, wait, we're going to have a problem not today, but in four days. That idea of predictive maintenance is actually baked into those new requirements. So legit on the electrical side for maintenance. This is the time to get involved. Okay? This is because Iot, just like they're saying it's going to be hooked up to everything. You know, we had it in mechanicals for a while. Now we're going to start getting some context and feedback from other systems. I think this really is just a very positive thing from a safety standard, from a future standard standard, from a standard of use of energy. But the one thing I'd say watch out for as building owner is okay, so if it's now becoming a requirement on the safety side, when does it become a requirement on the energy side? When do we say okay listen, you're building, there's an energy code and it is poorly enforced at best. But now we have IOT to be in there and sort of police it a little bit. And I don't want to say that as a bad thing because you know it's more expensive than energy downtime. Uh, if your systems go down you get to lose a lot of money really, really, really fast. So like this is also a hedge against that, you know. So I think this is a really good article. I think it really covered Iot and what's happening. Well, I just can't help thinking, and again Josh Bowen's heavily involved in that one too, is on the electrical side. IoT is now going to become the requirement. You're going to have to have a building that is able to give you some information so that you can make good decisions.

Speaker A: Yeah, I like that. NFPA is the National Fire Protection association, right?

Speaker C: Yes.

Speaker A: So this is, this is also a safety requirement in the end as well. It's not just see. And uh, that's what I like is it converged with a business purpose for that. Like hey, we can't have downtime, we can't have things not working, people not being productive, et cetera. But ultimately it's kind of fire protection. We don't want that gear. The wattages we're at are insane, insane and they could create pretty, pretty big problems. So let's uh, but, but hey, can

Speaker C: I add one thing real quick, Jeff, if you guys are interested in more on that Josh Bone, just put out an electry report on it through, through Electric. I was what, I was the researcher on it. It was a lot of fun. It's a big project but it's out there and you can go and sign uh, into electry and get a copy of that.

Speaker A: I didn't even think about it but John, I almost grabbed that article as mine and I was like, I don't know if that would look too self serving but it was awesome. It was a great article and, and if you don't mind, drop it into the show notes today and I'll make sure it gets off for everybody listening because they should be watching that because nfpa, this is also a um, I call it a canary in the coal mine moment around a lot of what you're talking about. So it'd be good to have that. All right, man, well, bring it home. You get to, you get to close us out with what you've got.

Speaker C: Okay, so this, this is like, you know, in. And I've been the sort of the doom and gloom on AI anyways, but the reason I am is because there's this tension with AI you know, AI this the article is workers gain hours with AI but risk losing skills, you know, according to Microsoft's new Future of Work report. So Microsoft's looking at this because they're using a lot of this and there is such a thing as deskilling a person. You can deskill a person really quickly. You know, that whole thing, if you don't lose it, you use uh, it, you'll lose it. That's not a lie. That's a real thing that happens. And I think this article does a good job of talking about where it's being used, how it's being used, thinking about how you're using it. But really, this is, you know, if, if anything is my biggest concern right now is a lot of the stuff that's hit in this article is really whether it's going to amplify inequality across the roles within your company, whether it's going to, you know, sort of augment your skills or whether it's going to skill you. Because I think used correctly, it can augment your skills in a positive way. Used incorrectly, and I know I'm m going to catch a lot about this and maybe it gets cut off. I feel like it's steroids, man. I feel like it's steroids. You want to look intelligent really fast, just use some AI you can look really smart to other people and you can appear really strong. But at the end of the day when you start use it, stop using AI you may not be able to be strong for yourself. Like you use gear too long.

Speaker A: You can't like I, I, that's what this is. I, I love it. The AI is the steroids for the brain. Look, it has a lot of consequences.

Speaker C: I love man. And remember, you can still use it to get strong though. Like when people I was watching some dude in the gym and I watched him and he was doing curls and the guy next to me goes, yeah, he's not doing those right. But he's on gear, so it was still get big. And I'm like, oh, so. But he won't know how to actually do a curl right? And he's like, yeah, if you did, you don't have to learn the form. If you have a shortcut. So I think in all of this, yeah, learn the form.

Speaker A: Learn the form.

Speaker C: Learn the form first. Then you can use the shortcut.

Speaker A: And by the way, you can use the tools to learn the form and learn how to do it and get better at it and build the muscle. And by the way, I am seeing this in kids right now. I am completely seeing it in using the tools to create. And I love it. They call it work slot work. It is not reviewed. It is it. It starts off good, it ends good. That's all they looked at. They perused it. But in the middle there's. And by the way, uh, in life right now, I am seeing this. All three of us get hit up constantly to either be on this show or be on the dorks or do whatever. And it is appalling what I see from these things with a, uh, with a plus sign here and a, uh, plus sign there. But you're not using that anywhere else. And it's like, this is slop, guys. You can't. And I'm watching students who can't regurgitate what they just slopped out. That's not right. Now if they can go back and forth with it and like you said, do the muscle right and learn. And it's. You know, by the way, John, you've done a lot of this in yours and taught me it recently is like how you can control the outputs that you're getting to be really, really specific and don't make things up and don't just placate me and challenge me and, and sight things so you can learn that muscle way, way faster. And by the way, like you said, without it, if you're doing the curl correctly, you're actually going to grow faster. So doing it right goes faster.

Speaker B: I'll just tag onto it. I see this too. Um, Hugh Seaton, who's done a, ah, webinar series for Smackdown. AI, uh, we talk about this all the time, how it's unskilling, it makes you dumber if you don't use it properly. I'll admit, you know, I think everybody kind of went through the same usage curve. Like when it first started, I basically used it like Google. Right. I'm just searching, I'm asking. And then maybe write this email for me. Do this for me, do this for me. And now since I've gotten more comfortable with it, it has grown like, uh, the models have improved, they've gotten more data, I've turned it into more. Using it more as a coach. Right. I already Know what it is I want to say and I want to do. Maybe, maybe I'm writing a report or I'm writing an article or working on a manual or something like that. So I'll draft something and be like, okay, what am I missing? What points could I make better?

Speaker A: What?

Speaker B: Um, you know, search the Internet, anything, any new data. So. And then in. In stuff in my personal life, I. I use it for kind of the same thing, like in my. In my weightlifting, in my nutrition plans and stuff like that. This is what I'm planning to do. Help me refine it, to use it more of a coach. Something that I can, like a research assessment or something, bounce things off of instead of. Instead of having it do all the work for me. Right? Because then you do just. You just get that slope.

Speaker A: We can't afford that slop in our world, right? You can, uh, I'm sorry, with you, marketing. I love you guys, and I do a lot of that. You can afford a little slop. Nobody gets killed from it. But we cannot afford slop in the field. We cannot afford slop in construction and design and engineering. And that's where, you know, a lot of it says in there about generative AI. And, you know, we've seen so many runs at generative AI playing a role in designing buildings and designing infrastructure, and we can't have slop there. We need knowledge we could really use for great engineers to put out more work because we're not gaining more. If you look at, by the way, a couple episodes back, Quantum Rise, that's where Hugh Seaton is right now, so you can go check him out. And we just had Bryce Wisen on that. You're going to see, and that's the accounting piece, and he says it there, too. If you are doing your accounting muscle correctly, then you can then use the tools to amplify that and get stronger and be better and make better decisions and know where your cash flow is, et cetera. So we're saying it could be done, and we. We do need to capture a lot of that, right? We're losing it before we can replace it. So it would be nice to codify it in some way. The AI can do some of that for us, but we gotta know how to. John, I love it. I'm gonna. I'm gonna start using that one. You don't go into the, uh, to the gym and just start picking stuff up and using it. You need to know what you're doing, why you're doing it. Like you said, Travis, you're like throw in your workout routine and it's like, hey, Travis, you're skipping leg day, man. Stop skipping leg day.

Speaker B: You know, like, gotta, uh, work on those calves, man.

Speaker A: Gotta work on those calves, buddy. So, uh, yeah, but ultimately that's what that article is really leaning on.

Speaker C: I love it. I would say one thing that, that me and Jared Christman and Jeff Elwell are doing a panel on necessary effort. Because to sort of counteract this, what we started to say is, okay, well let's start really finding out what are the things that require necessary effort to really grow. Like what are the things you should not cheat on? And I actually think that could be a real positive because we haven't had enough, I would say tech like beating on us yet that we really had to step back and go, okay, what do we actually need to know and what do we not need to know? You know, like, like we have had calculators. Do you still need to know how to do basic math? How much math do you need to know how to do if you have a calculator? You know, we spent some time on that back in the day and now I feel like we're hitting the point where we're going to start spending some time on that. And that I like because that also means now I know where to put my effort in order to grow. Well, because if you know what's necessary and how, what's important to grow now, you can put more effort in that area and dismiss the stuff that's unnecessary to grow. You know, I, maybe I don't need to do as much homework, but I do need to do more like back harder problems or something. I don't know. But like that idea I think is there and I think that's one of the positives that's coming out of AI is we're looking at what's important and what matters when productivity is at stake.

Speaker A: I like that. Uh, I am going to put you all on the spot real quick. Looking forward into 2026. Do you have a prediction? Do you have a thought? Do you have anything to give the listeners? Do it quick and we'll wrap it up.

Speaker B: AI is going to still be huge. I mean it's just not, it's not going away, you know, just to tag on. It is kind of funny. Like when we do these news episodes or I think about something to write on for some articles I have, it's hard to find anything that's not AI anymore. It's going to continue to be noisy, I think.

Speaker C: John, any thoughts I think that AI is making a lot of things very, very simple to achieve at an 80% margin. So like, I think that this is going to be the year of 80 percenters. Like we stuff is 80% done. That's never actually going to get to 100% done because the reason it stopped at 80% was, was real. Um, I also think this will be the year of micro coding, because I know that with some of the vibe, coding has hit a point where like, I have a little like, assistant that I made off to the side of my desk literally in 15 minutes over lunch. That's really helpful for what I'm on. But I'm going to throw it away. I, uh, literally emailed somebody today. I'm like, I'm going to build throwaway tech for a long lot this year. You like programs where I'm just going to throw it away because, like, it gives you a chance to experiment, it gives you a chance to figure out what it should be. And hopefully that means you move it forward to somebody who's going to make it sustainable. But I think this could be years, a lot of throwaway tech. We're going to throw away tech, microcoding and throw away tech because you can do it so easily. And I would even encourage people to experiment with it. Not on your work computers, on a computer that is not important that you can zap after you're done. But like, I would encourage people to get on and experiment because I think a lot of that, that's, that's where AI's really moved ahead. I mean the, the personal assistants are there, but there'll be more there next year than they are there this year. But, but the coding's really, really taking a big bite.

Speaker A: Yeah, I'm going to steal from, uh, our boy Trent Linen back and just say it for him because I do think this is the year as service, as a software. And that, that is actually going to be the differentiator. The groups, the technologies, the companies that can figure out how their service expertise provides more for their customers will get them more using it. Because you guys are right, AI can build almost any feature at almost any time. And with good tech and less people, even you can incorporate that into your stack very easily at an enterprise grade level. And um, like John said that microcoding is not enterprise level, but he's throwing it away, which I think is fantastic because that's teaching him to move towards enterprise level. But if that is your job and you actually know the service and how people perform their work and how to amplify what they're doing and incorporate it into your system and get adoption and help them grow. I think that's going to be your only differentiator, uh, not only in 2026, but into the future with that. So that's my double down. But I got to give total credit. And by the way, I've sat down with a couple people since Trent since said that and I've heard it quite a few times. I texted him and and made sure it was clear that we got it. So. And by the way to these two dorks and Trent himself, the dorks. If you guys aren't checking out our other show, the Construction Dorks, you need to look it up on your favorite podcast app and listen in. That's where we have a lot of fun and we will be bringing you more episodes over there. If anything. We said to hear connected with you, connect with us. We really appreciate the time. Thank you for tuning in today to geek out for the Construction Technology News of the Week with Jonathan Marsh and Travis Voss. To read all our news stories, learn more about our guests and to listen to this show, visit thecontechcrew.com this is the Contact Crew signing out. Until next time, enjoy the ride and geek out.