PODCAST TRANSCRIPT:

Beon DeNood
This is Inspector Toolbelt Talk.

Ian Robertson
Hey, David. On this episode, because the two of us are together, I really think we should keep the giggling down, as someone did complain about the giggling on a previous episode.

David Nyman
Really?

Ian Robertson
No, not really.

David Nyman
I thought that was the best part of it.

Ian Robertson
Somebody actually did complain about the inspectacles episode, that we laughed too long about it.

David Nyman
What can I say? We’re too funny.

Ian Robertson
No, probably not. Only to each other. But it’s good to have you back on the podcast again today because we have something that is, it’s important to our industry, whether we like to talk about it or not, and it goes beyond, you know, AI—what are we using it for? This and that. Everybody has an AI episode floating around every two minutes. This one is entitled “The Home Inspection Singularity Is Here.” So we hear that expression singularity. In physics, what’s a singularity, David?

David Nyman
It’s a black hole.

Ian Robertson
It’s a black hole.

David Nyman
Even though I’m not sure that’s the only definition. I actually haven’t looked that up. It’s the first thing that comes to mind for me, anyway.

Ian Robertson
It’s not the only definition. It’s like saying, scotch is a whiskey, you know. A black hole is a singularity. But whenever there’s a singularity in physics, basically you can predict most of the universe mathematically. You can say this is going to interact there. That’s going to interact here, and this is what it’s going to do. As soon as it reaches a singularity, it’s where everything converges, and then when it comes out or not comes out, there’s no way to mathematically predict any of it. It’s why when you end up in a black hole in a movie, it bends space and time, and all this weird stuff happens because it’s unpredictable. So, we in the home inspection industry have had some predictable things that have happened. We predicted AI was going to become predominant in the industry. Home inspectors would adopt it. People would start accepting it differently, and all that stuff. We’ve kind of reached a point where all the predictions have kind of come true for the most part, with a few exceptions of advancement, and now we’re getting to the point where mathematically, it’s going to be really hard to predict the next stages of things. Would you agree with that assessment or no?

David Nyman
Yeah, for sure. I mean, on a daily basis, I talk to inspectors that, they run their comments through one of the AI chatbots. They take business advice from their AI, so it’s definitely, it’s all over the place. But you do have some holdouts too. You know, you have the paper and pen inspectors out there still, as well as inspectors that say, I’m never gonna AI touch my data.

Ian Robertson
Yeah, there’s always gonna be a holdout. If anybody wants to look up what a luddite is, there’s always gonna be luddites in any advancement.

David Nyman
Look it up in a dictionary.

Ian Robertson
Yeah, look it up in the dictionary. But no, it’s just people who are opposed to advancing technology. But it’s gone quite a bit beyond that. Just every day, we’re inundated with AI in ways that I never thought possible. There was a home inspector that emailed me, and this happens all the time. They write up an email or they connect their email right to AI or Gemini or ChatGPT or whoever, and all of a sudden I get this message, and I’m like, this is very obviously AI. So I knew the inspector. So I’m like, I’m going to have my AI responder respond. The two AIs argued with each other on something, so he emailed me said, this is wrong with my website. He didn’t know what it was talking about, and it actually wasn’t true. It just couldn’t access it, so the AI just made something up out of thin air. So I let it go on for a couple hours, and I started looking at him like, what in the world are these two AIs talking about? So finally, I told my AI to just, you know, put in a message to have his AI, we know this is an AI conversation. Have it check what you said, and here’s the actual information. And his AI came back and said, oh yeah, I was completely wrong about that. Scratch that. Thank you for all the good work that you’re doing for us. And then he sent me a message afterwards. He goes, that was weird. He was using Grok or something, and I was using whatever, and they were just going back and forth. But that’s the era that we live in.

David Nyman
Yeah, I did hear that one of the fears they have, you know, all these doomsday scenarios about AI is, when they start talking their own language, that it’s more efficient for them to communicate with. Instead of using, you know, English or whatever they’re using currently, so that they can come up with a language that’s more efficient for them to talk to each other, and then humans can’t any longer understand what they’re saying to each other.

Ian Robertson
That’s already happened, though.

David Nyman
I don’t know. I don’t think, like that’s what they’re saying. Like once they get to that stage, like that’s when human control is out the window completely. So they can still see the thoughts behind. I guess.

Ian Robertson
Oh, okay. I guess. Maybe I misunderstood because there have been two different AIs that they put in the room, so to speak, together, and they come up with their own language to be able to talk to each other. But like you said, I guess the difference is being able to see that, like you can see the reasoning and all that stuff behind it. But how that relates to home inspections is very interesting, because now it’s not just, will home inspectors use it, but instead, what do human beings, our clientele, expect? One of the things that I’ve predicted for a long time is that people lower the floor. In other words, the basics of what they expect out of a service, the standards have gone way down. Here’s an example. They did a study, and I don’t know why it’s such an odd number of people, but 1,301 people.

David Nyman
That last person really made the difference.

Ian Robertson
Yeah, were we not going to trust it at 1,300? They’re like 1,301. They’re going to trust this study. But they had these people review all of these emails. Some were written by AI and some were written by people. When they did it the first time, they identified the AI-written emails and people didn’t like that, but then when they showed them all of the emails that were written by AI and people and didn’t identify which ones were written by AI, they all said that they didn’t care which ones were written by AI. So basically, it says “AI recipients showed essentially no skepticism. Their impressions were virtually identical to those receiving confirmed human-written messages.” So they use that as a way to say, people just kind of expect things to be AI slop now. So they have no feelings on it. Two years ago, if I dared to send a client an email written by AI, they would have been like, did you did you just write me an email with AI? How unprofessional! Now I’ve written emails, and they’re like, maybe you could run that through AI to be more complex for me. You know, it’s like just in a couple of years, people have completely changed their standards. There was a couple of other studies. Gallup found that 79% of Americans had seen obviously AI-looking advertising and said it was acceptable for companies to use AI to create the final text, images, or videos in advertisements. Again, you ask people that same question from three years ago, and everybody hated it the first time they saw AI slop. Now they’re like, yeah, sure. 80% of people accept it as yeah, this is normal.

David Nyman
Yeah, and also on top of that, it’s at the point now where basically it’s such high-quality video and things like that that people are identifying a lot of things as AI that aren’t AI. So across the board, everybody accepts anything that’s exceptional that you see. You’re like, oh, that’s got to be AI. So yeah, it’s really made it a lot more difficult to say you don’t like it because people think everything is AI.

Ian Robertson
Yeah. And the young people, Gen Z, or Gen Alpha, whoever the young people are nowadays, they actually have an expression, when you’re lying, they say, that’s AI. So they use AI as for anything made up. If you’re like, oh man, I ran a four-minute mile today. They’re like, that’s AI. So people expect it, but it’s lowered the floor. It’s also raised the ceiling in certain ways, but it’s also kind of made the ceiling a little bit fuzzy. But for our industry, for home inspections, it’s done something unique that I said would happen, but I didn’t think it would quite happen like this. And what’s that expression? Correlation is not causation. So some of it is a bit of assumption, but a couple of things. The price of a home inspection has nationally been trending downward again. It trended downward whenever there was anything big that made home inspections easier, like licensing. Now people are like, licensing doesn’t make it easier. Yeah, kind of does. We’ve had that discussion before, but either way, some guy fresh out of school comes out and he’s going to have the cheapest prices. You know, instead of the guy who’s been working in the family business for 30 years and then becomes a home inspector. Big difference in the two scenarios. So whenever a licensed state starts licensing, the cost of a home inspection goes down for a period of time. So whenever it lowers the floor or creates a barrier of entry that is lesser than it was before, i.e. I had to work 30 years to have a skill level to be a home inspector and a whole bunch of credentials. Now I can just get a license and have the same credentials as a guy with 30 years. AI has done that, though. So there has been a trend downward in pricing. Matter of fact, the home inspector just last week emailed me, and he’s just like, Ian, I can’t believe it. And he showed me all the trends. I’m like, I believe it. It hasn’t gone down like $100 an inspection, but in most markets it’s like minus 25, minus $40. We’re still making now what we did in the 90s, but AI is going to continue pushing that, and for the reasons that we talked about. First of all, people’s standards of what they expect in a report are different. How many AI-written comments are in most reports that you read, David?

David Nyman
I have a hard time setting a number to it, but it is interesting that I’ve seen home inspectors that are complaining about formatting issues, and I don’t know exactly how it works, but I know AI is putting some formatting in the comments that you get out of them. I’m assuming it’s to be able to identify it for itself, so that formatting doesn’t show, you know, when you copy and paste the text, but it can show later in your report. So, just that fact alone has told me that there’s a lot of AI comments popping up in people’s reports.

Ian Robertson
Yeah. So, and that’s an interesting factoid for the side, what was it Claude, or one of them is starting to put imprints in their text and everything, so that AI can recognize its own AI content, and other AIs can as well. So that’s going to actually become more prolific. But when people read reports now, it used to be, again, just two years ago, did you use AI to write my report? Now you might even have some clients or agents saying there was a couple of grammar errors in there. You should use AI to help you write your report. They expect weirdly long comments instead of being concise. People expect the weird flower, flowery, flowery. Am I saying that right? If you say a word enough, it doesn’t sound like a word anymore. The weird flowery language that AI puts in things. I ask AI a simple question, a math question, like, oh, you know, what’s this divided by that, real quick, because I was doing something quickly. I was working in my garage. It was nine paragraphs, and I’m just like this. This is silly, but that’s what AI does. It makes things big and flowery.

David Nyman
Unnecessarily complicated.

Ian Robertson
Yeah, and now a good portion of clients are running their report through AI when they’re done. So like in ITB, you can do a PDF version and you just pop it into AI, and it reads it, and it reads all the pictures and things like that. It really, it creates a weird dynamic because that used to be half our skill as home inspectors. We would one, recognize a defect, which takes a lot of skill in and of itself. You need to know a lot of stuff to be able to recognize defects, and then two, be able to, you know, narrate that, put it into words. It’s kind of removed that barrier.

David Nyman
And then also third, explaining it to the client.

Ian Robertson
Okay, well, that goes back to our episode. Remember the episode on don’t let AI write checks that you can’t cash. That is happening a lot. Okay, so I’m not going to use any names, but there was this inspector that one of the inspectors I work with, he got asked to be an expert witness because they were suing this home inspector. His report was full of so much AI slop, but for somebody that didn’t know, it all looked magnificent. Like he spent days on this report. It was formatted weird, hard to read. He had other states SOP mixed in with his state’s SOP, which was all weird. But interestingly, he fooled the client for lack of a better word, and I’m not using names. I know this inspector personally to an extent, and he struggles. Like he works hard, but he struggles. He had magnificently huge errors in his report. Interestingly, this guy couldn’t really sue him because it’s just like all you’re going to get out of him is his fee. You know, paid 400 bucks. That’s all you’re going to get. You’re going to spend 10 grand to get 400 bucks. So that’s another point. But he wrote checks with AI that he couldn’t cash. He couldn’t say why he wrote that in the report. He didn’t understand what he put in the report. He didn’t write most of it. And I’m like, that guy needs to listen to that podcast about writing checks that you can’t cash. So that’s a very good point that you brought up. But that was an extreme case, and I wish I could go into more detail, but it’s not anything I’d put on air. But that’s why we’re at the singularity. Whenever anything converges on a singularity, the math starts to break down, and it’s like, where do we go from here? So I can see the math breaking down in our industry right now. It’s like, okay, this guy lowered the floor, doing a $250 inspection. I see guys out there doing $199 inspections because you know what? They can do five reports in a day, spend an hour on site, and then sit there on AI and then just say, hey, analyze these photos and write some comments for me, and you’re going to get a report that will fool people. It’s not going to be a pretty report. It’s not going to be consistent, but that guy just lowered the floor. Do you see people raising the ceiling with it, though?

David Nyman
I could definitely see that happening. I think the one thing that you probably always had in the market is, you know, some people view the home inspection report as a formality. They’re like, okay, I need to get this out of the way before I buy a house or sell a house. But you also have those that genuinely care, and all these, like you say, slop reports, they will probably end up with reviews that are very mixed. So someone that writes a clear and concise report, finds all the defects, describes them in a way that’s easy to understand and really gets the point across quickly. I could definitely see those rising up above the slop.

Ian Robertson
You know, I agree with you there. I do think that it’s just disappearing middle ground. We’re going to have low end and high end, and very little middle ground. Whereas most of our industry for the past 30 years has been a lot of middle ground, and a little bit of the low and a little bit of the high. You know, it’s kind of like website building and things like that, and SEO and all that with AI, people are like, oh, I can go online and get this really cheap, and I can do it myself. There was one guy that built his own website, and he came back to us like he worked with us, and he says, oh, I can do this with AI, and it’ll be amazing. It’ll do automated SEO. And I’m like, okay, go for it. Got some guy from some country to build him a website, automated everything, and he wasn’t a jerk. But he came back and kind of rubbed it in our face a little bit. It’s like, ha! Like five months later, he came back and we built him a new website, and he’s back on a plan with us. Nice guy. But what ended up happening was it lowered the floor of the industry. Yeah, can you build a nice-looking website and add a bunch of quote unquote SEO stuff? Yeah, but that doesn’t mean it was a power tool in the hand of a skilled builder. You give a bunch of guys that don’t have skill a bunch of power tools, they can put together a house really quick, but it’s gonna not be a great house. They didn’t think of things like, not that anybody really wants to know any of this stuff, but pre-rendering. They didn’t think about proper schema markups, redundancies on the back end. They didn’t think about how their structured data was laid out, regenerating sitemaps and XML sitemaps. And once you put all that stuff together, that’s what makes a website good, you know. Not just having it. But it’s the same thing with a home inspection. People are eventually going to figure out, okay, if I want a cheap home inspection just to get through the formality, like you said, there’s going to be no end of cheap guys out there using AI.

David Nyman
Yeah, it’s interesting. Yeah, when it comes to the industry as a whole, I feel like very few of current home inspectors actually want an app that does all the work for them, because they realize that’s the death of the industry. But there’s still apps being developed that do all the work for them. So it almost seems like you have a stronghold of current inspectors that they would hate to have that. So it makes it a little bit harder for these companies to make the full AI, you know, analyze pictures for you. Basically you walk around as a human camera and that’s all you do. It’s a small market share right now, but it’s definitely something that could get out of hand when more and more people start going for the low end.

Ian Robertson
And there’s the point. The high end is going to say basically, do you want AI slop or do you want a real human doing this for you? Because I mean, the technology is there to do exactly what you said. Matter of fact, in some industries, not the home inspection industry, they have it where people just walk through with a camera, and then AI just extrapolates everything that it needs out of there and produces a report. But the fallacy is all of us are worried about that, and then we make ourselves feel better saying things like, yeah, but you can’t get AI to crawl in a crawl space. You can’t get AI to go in the mud and check this out. You can’t give AI, you know, the experience of touching the heating unit on the side to see if the heat shield is broken or whatever. But there’s a fallacy in that of when the floor is being lowered. With the floor being lowered, are people going to care about that. So think about appraisers. I bring them up because they’re where our industry may end up. You hire an appraiser. Most of us never look at the appraisal report. As long as it’s within the range that they’ll accept the mortgage, we just move along with it. You know, there’s very few times it is a problem. It’s a formality, and nobody goes around looking. I want the best appraiser. You know, if the appraiser uses AI, nobody cares. Just do your job, and as long as it’s what I want, then there we go. We don’t want it to be that way with home inspections, and it won’t because our job is a little more direct. So if they get the value of your $400,000 house off by you know $10,000, big whoop. Wait a couple years for appreciation is what it is. If the home inspector tells you that you have a gas furnace and you have electric baseboard, now that’s a different issue. AI is still a long ways off from being able to completely take some college kid walking around with a camera, but it’s not so far off that we shouldn’t at least think about it. And there’s where the singularity comes in. Interestingly, I keep seeing diverging reports. Do you see the same thing that I see? I’ll read one report saying the accuracy of AI has gotten better by up to 80%, and then I will see an article saying that it has gone down by like 40 or 50% because hallucinations are happening more, and I’m like, they both can’t be right.

David Nyman
Maybe an AI wrote those articles.

Ian Robertson
Yeah, exactly.

David Nyman
I’ve seen, yeah, I’ve seen both, and from my personal experience, I feel like it’s definitely the more accurate thing, I feel like it’s gotten worse, like it got better for a while, and then all of a sudden they had a big dip. Part of it might also be that they’re trying to monetize it more than they were, and that means that if you have a free model versus a paid model, you’re getting wildly differing results. I almost suspect that the free models are just consciously being made worse, just because they’re dumping so much money into the AI bubble that they have to start making money out of it. To me, I’m still not 100% that AI is ever going to hit the spot where they claim it’s gonna hit. It almost feels like it’s a little bit overblown. Especially when it comes to destroying the world or ruining life as we know it. I remember year 2k for all you young kids out there.

Ian Robertson
Oh, Y2k. Yeah.

David Nyman
Yep. Everybody’s like, it’s gonna ruin the world. Everything’s gonna break. And then, you know, year 2000 came and passed, nothing happened. And I feel like sometimes they kind of don’t know how things work exactly, and they just start hypothesizing.

Ian Robertson
Hypothesizing.

David Nyman
Why do I use these words? But they start making up all these, you know, scenarios because they don’t know what’s going to happen, and kind of feels a little bit like that with AI right now to me.

Ian Robertson
There’s always, to your point there, there’s always going to be a bit of excitement and over-excitement, like Elon Musk is like, don’t save for retirement. If my predictions are right, the quality of life is going to go through the roof. And I’m like, come on, Elon, it’s still a large language model. It does have a long ways to go before it starts replacing people permanently. But in the hands of a skilled person, it really does add a lot. Like a home inspector who’s really good can, you know, scan some things, run it through AI, and you know, put it in a report and recognize things that are wrong and things that aren’t. I was talking to my accountant who’s been on the show before, Tom. And he was funny. He’s like, even with AI, AI can completely do your taxes without us involved, and he goes, our business has grown. Because number one, people don’t trust AI, and they’re like, what if it made a mistake? And number two, they still want to be able to talk to a person, and not just, like we think, oh, well, obviously older people. No, even the people that weren’t alive for Y2k, they’re young people. They’re just like, all right. Yesterday, it told me that blueberries were cheese-flavored, so let me at least ask somebody who knows about my taxes.

David Nyman
Wait, they are?

Ian Robertson
Blueberry cheese. Those of us who are 25 and above are going to have an advantage because we will have job security well after retirement because we will be the only people that know how things work without AI. So when an AI model goes down, here’s a good example. In New York State, they used to have this programming language called COBOL. Terrible programming language, awful, ridiculous. But there was only a few people that knew it. New York State is still using it in certain things. So there’s a guy I know, they actually called him back in because it broke. They called him back in and paid him stupid amount of money because he’s one of only three people left in the world that know how this COBOL works. And there he is, and he goes, I have a buddy that just works for the state, maintaining it until they transition over. They’ve transitioned over most everything. But that’s going to be us. Like AI is going to be like, oh man, we’re hallucinating now, and we don’t know what to do. It’s like, okay, let’s go find somebody that actually knew how this worked before AI.

David Nyman
Yeah.

Ian Robertson
And then there’s quantum computing, which you mentioned the other day, and then I went down the rabbit hole. Should I say rabbit hole or black hole? I went down the black hole of quantum computing. They already have 1% quantum computing systems already. They’ll do 99% traditional computing and then 1% quantum computing. And they say by 2030, they should have at least one model that does fully quantum computing. If that happens, I mean, it’s limitless what they can do.

David Nyman
Yeah, it is crazy. You know, when you see development over the past 20 years, it is definitely getting exponential.

Ian Robertson
You gotta stick to the small words.

David Nyman
I know.

Ian Robertson
Me and you will both stick to the small words.

David Nyman
Me, David. I got rock.

Ian Robertson
AI real smart.

David Nyman
Yeah.

Ian Robertson
But there’s advancements coming. That’s why I say though, if quantum computing becomes a thing, it’ll be another five to 10 years before it becomes ubiquitous after they have it.

David Nyman
You’ve been practicing that one.

Ian Robertson
I’ve been practicing. I have to practice my ubiquitous. It’s going to change everything. I mean, you’ll have a supercomputer in your pocket at any given time. We should do a whole podcast on quantum computing because I completely do not understand it. I have tried so hard, but the clearest explanation that I’ve heard is basically ones and zeros is how we compute now. If you’ve ever read the book Three-Body Problem, they use the Chinese army and they do ones and zeros, and they can do advanced computing if you have enough ones and zeros. So that’s like looking at one side of the coin or the other. Quantum computing is like taking two coins and flipping them in the air, and those two coins are flipping at a state where you can see multiple sides at once. So that’s what quantum computing is, and then also those two coins can interact with each other. One hits the table and vibrates the table so that the other one lands slightly different, or they could hit each other or not. So quantum computing basically takes a qubit, they call it, and they spin it. I’m sounding like I know what I’m talking about, but I totally do not. This is so above my head.

David Nyman
You just reminded me, you know, there’s 110 types of people: those who understand binary and those who don’t.

Ian Robertson
I apparently do not.

David Nyman
Oh, that was an awesome joke. There’s some people out there that are going to crack up like crazy with that one.

Ian Robertson
They’re laughing hysterically right now. But quantum computing is-it’s not just a new higher language. It’s taking away old computing. The ones and zeros don’t exist anymore. When you do one through 5 billion, it can do a calculation that would take years for regular computing to do, and it will do it in seconds. So you know how you sit on ChatGPT and you type in something and you wait a couple of seconds. Those couple of seconds will turn into milliseconds, and it’ll do things like your entire report. You say, clean up my report. Boom, half a second, it’s done, and it’s going to do it with incredible accuracy. But we’re 10-15 years away at best. Like they’re saying, at best, five years, four to five years for quantum computing in general. Another five to 10 at best for something that the average consumer can use.

David Nyman
Yeah, I mean it’s like when the first computers came out and they were big as a house and much slower than current calculators. But yeah, once it starts going downhill, it just goes faster and faster.

Ian Robertson
But until then, I think, again, this is the singularity. It’s the home inspection singularity. I can’t predict it. I think that AI is becoming too expensive too quickly. So if you’ve looked at like OpenAI and all of them, their losses are becoming fairly heavy, which is one reason why some AIs are now doing marketing. So, if it continues to go up, the cost of computing with all these data centers and things like that, it’s gonna be a problem. They’re going to have to start charging more for it. So, for instance, we use some pretty advanced AI for what we do here, and it’s not cheap. It is not cheap at all. I can’t imagine the average home inspector spending several grand a month to have AI just do normal things that we’re used to it doing for free for us now. But we’ll see how it goes. I do think it’s going to become more and more expensive. Like even the AI integration we have with ITB, I’m amazed sometimes. I’m just like, wow, you know, computing now is quite a bit more expensive than it was just six months ago, and every six months it seems to go up, and they’re still not making money, those companies.

David Nyman
No, that’s true, and they’re gonna have to build more data centers. Most people don’t know that I’m completely AI-generated too.

Ian Robertson
What’s that?

David Nyman
I’m AI. 100%.

Ian Robertson
You’re AI generated. Well, Kim, one of our support people, is one of our best ones, and she’s always getting accused of being AI. So I’m like, Kim, make a mistake or something. Be like, oops, that was misspelled or something. I don’t know. Stop being so good at your job. They think you’re AI. But it is amazing how many people get irritated. They’re like, are you AI? I want a real human. And I find myself yelling into a telephone. I call it a telephone. Wow, I’m old. I find myself yelling into a phone often enough. “Human agent!”

David Nyman
Yelling always helps.

Ian Robertson
It does. Well, I had to do it recently because I kept saying it calmly. Human agent, and they’re like, “We don’t know what homogenate means.” I’m just like.

David Nyman
Speak slower.

Ian Robertson
And so I’m like yelling it, and so the person who gets the call just hears me going, “Human agent!”

David Nyman
Yeah, I mean, we do have to do an episode on the one thing that AI will never be able to do right. Identifying smells on a home inspection.

Ian Robertson
I don’t know. That is true. It’ll take it a while, but I’ve walked into some houses where I’ve not been able to identify it. I’m like, what is that smell? That is nothing I’ve smelled. My brain doesn’t register that. That’s actually a good research point. You should look that up. There was a, I don’t remember enough of it to mention it here, but I was reading about how computers were trying to identify smells, and they were only having mild success with it. But what would you predict is coming? We’re in the singularity, so we can’t predict what’s happening. And all of my predictions have come true. Some of them not the way that I thought, but I like being right. But from this point on, I’m just like, man, it’s anybody’s guess. So, what do you think, David? What’s coming?

David Nyman
Yeah, I do think you will see that skill gap, I guess you can call it, between the cheap inspections and the people that actually want an inspection that they can read through, understand, and make use of. So, it’s gonna get wider, and that will mean for a lot of home inspectors, either they have to lower their prices to match what’s around them or raise their prices and be premium inspectors.

Ian Robertson
And raising your prices is the way to go, because if we lower our prices to keep up with the low guy, that’s a one-way journey. Like that’s hard to come back from. That’s why I’ve been saying that whatever home inspection company we are, don’t be a luddite. Don’t hate AI. Don’t think it can’t help you. Grab it, embrace it, and use it to become that way high-elevated home inspector, is charging $1,000 an inspection, because you know what? Prices are going down. Our prices have gone up for my inspection companies because we’re using AI as a way to enhance what we do instead of replace what we do. But we can’t stop the guys from replacing what we do with AI either. Just know that it’s a lower-quality report. Until AI becomes completely self-aware, it’s not going to be a complete replacement for a good home inspector.

David Nyman
No, for sure. And I want a good inspection. I know that.

Ian Robertson
Yeah, you want a good inspection, but as the floor lowers, there’s going to be a large cohort of people that are going to settle for whatever AI slop they’ve become accustomed to in society. If your attorney spits out AI slop, your agent spits out AI slop, Zillow spits out AI slop, you get AI slop within all your research. Why would they expect something different from their inspector? So it’s lowered the floor, but it gives us a chance to raise the ceiling. So the singularity is here, but it’s definitely not the end. It’s just the beginning. There’s still some things that we can do to make sure that we’re set apart.

David Nyman
Very true.

Ian Robertson
David, thanks for being on. Appreciate it.

David Nyman
Always a pleasure.

Ian Robertson
See everybody next time on Inspector Toolbelt Talk.

Outro: On behalf of myself, Ian, and the entire ITB team, thank you for listening to this episode of Inspector Toolbelt Talk. We also love hearing your feedback, so please drop us a line at [email protected].

If you’re enjoying the conversation, don’t forget to hit the subscribe button. Our podcast is available on all major podcast platforms. For more information on our services and our brand-new inspection app, please visit our website at Inspectortoolbelt.com.

*The views and opinions expressed in this podcast, and the guests on it, do not necessarily reflect the views and opinions of Inspector Toolbelt and its associates.

PODCAST SUMMARY/BLOG

The home inspection industry is hitting a turning point where artificial intelligence stops being a novelty and starts changing what buyers, agents, and inspectors expect as “normal.” When AI tools can draft comments, rewrite narratives, summarize defects, and even read PDFs with photos, the old differentiator of writing a clear home inspection report gets blurred. That shift feels like a singularity: we can predict the early impacts, but the next steps are harder to model because the technology, pricing, and customer behavior are all moving at once. For inspectors, the big question is no longer “Will I use AI?” but “How will clients judge my work when AI language is everywhere?”

One of the most revealing changes is how quickly standards drop for everyday communication. Studies discussed in the conversation suggest people often react negatively when AI authorship is labeled, but become indifferent when they simply receive the message without disclosure. That same psychology is showing up in inspection workflows: buyers and agents increasingly accept AI-generated text, and some even prefer it because it sounds polished, long, and confident. The danger is that “confident” is not the same as “correct,” and AI hallucinations can create believable nonsense. If consumers get accustomed to AI slop from advertising, real estate portals, and emails, they may stop demanding precision until something goes wrong.

That expectation shift is tied to real market pressure: home inspection pricing can trend downward when barriers to entry drop and when report production gets faster. AI can help a low-price operator crank out more reports per day by generating narratives from photos and templates, even if the result is inconsistent or hard to read. The risk shows up later in liability and reputation. If an inspector relies on AI to produce statements they cannot explain or defend, they are effectively “writing checks” they cannot cash when challenged by an agent, a client, or an expert witness in a dispute. Skill still matters, but the consequences of using AI carelessly get sharper as more consumers treat the report as a formality.

At the same time, AI can raise the ceiling for premium inspectors who use it like a power tool instead of a replacement brain. The market may split into low-end, fast, cheap inspections and high-end, high-trust inspections with clearer writing, better defect recognition, and stronger client communication. Future tech like quantum computing could compress timelines by making AI faster and potentially more capable, but economics may push the other way as compute costs rise and AI vendors monetize harder. The practical takeaway is simple: embrace AI to enhance accuracy, consistency, and client understanding, while protecting your credibility with human judgment, tight language, and the ability to justify every recommendation you put in a home inspection report.