Ep 22: Are Construction Companies Viewing A.I. Wrong?

Join our mailing list & gain more
construction business insights from our experts.

Episode Transcript:

This transcription is provided verbatim from the excerpts of the podcast video “Are Construction Companies Viewing A.I. All Wrong?” uploaded on the YouTube channel “Construction Hot Takes Podcast.”

So about a week ago I was watching football on Sunday and a really interesting commercial came on it happens to be for IBM and it was about AI and since that’s all we’re talking about kind of in the world these days I I paid attention and the the Watson commercial no so it’s it’s it’s new i think it’s a newer commercial but the gist of the commercial is it’s two people walking down a hallway in an office and the woman says to the man “Hey um I love AI.” You know and the man says “Yeah um it’s supposed to streamline our lives right and it’s supposed to make our jobs easier right?” And the woman goes “Well that’s that’s what it’s supposed to do.” And then the guy goes “Well but doesn’t isn’t AI just something else that we have to manage and doesn’t that mean that it’s more work for us and so isn’t that the opposite of streamlining?” And the woman goes “Yep.” And that’s the irony of it and I thought it was very interesting that in the world that we live in which is kind of all all AI focused people need to think about AI the right way not just as like this magic pill that’s going to fix everything for you so I think it’s really interesting for us to talk about in this episode some of the pitfalls of AI and how to overcome those pitfalls so that any business owner of a construction company can know what they need to be doing and how they need to be thinking about AI for it to be the tool that it is interesting yeah okay all right so so so I’d love to kind of jump into that with thinking about you know what is AI right it’s a tool like we use a tool for everything else it’s a it’s a tool like we would use a tool like software or a tool like we would use a screwdriver right it’s something to make something else happen yeah i mean it’s also in my mind it’s also like an assistant so instead of perhaps delegating something like write me a letter to uh like a coordinator or an administrative assistant or a uh somebody lower down on on your project team you have the AI do it because it’s something that it’s really good at it’s a tool yeah but it’s also I treat it sometimes like another employee that I can give tasks to so that’s a great segue so so that’s exactly kind of where I was thinking about this if it’s an employee if it’s an assistant what do you have to do with an employee or assistant you have to manage them ai you have to train them you have to manage them you have to oversee them you have to check their work yes ai for me is a tool that has to be managed and so I think thinking about this in the right way I’ll use your example something you want it to write something for you you don’t just send that out no you have to read it you have to check it you have to maybe change the model you have to you have to polish it and so make sure it’s accurate it’s got to be accurate it’s got to be right for what you want so tone all those things tone style punctuation all those things so the question is is AI a tool to be managed or is it a tool that you think is just supposed to work hands down day one to make your job easier uh well it’s definitely not at the latter stage yet it is certainly still in in its infancy as far as what it can do on its own um you know I find that it it’s it’s a very quick learner but it’s not omnisient or anything like that yet it doesn’t know exactly what you want it’s still it’s it’s a very advanced child and children take management and teenagers take management and some adults take management development and you know the more you teach it the better it gets but it’s it’s not independent yet and and and so in that same vein of AI needs management I think what’s important for people to think about is how do you think about AI as something to be managed right how do you think about AI as something that you can train each hover over well so in the last 12 months when I’ve been talking to contractors and I think you’ve probably had the same experience most of them aren’t using it or don’t understand how to use it they we have we have contractors that have not seen it before i’ve literally shown it to people for the first time in just in the last three or four weeks yeah construction companies construction company executives owners and they’re like “Oh wow i didn’t know it could do that.” And I’m just showing them one little thing like analyzing a spreadsheet and giving me some some KPIs or or giving me spotting trends helping me do some quick data analysis they’re like “I had no idea it could do that i do all that stuff manually.” So so that you know again a perfect segue into the next thing that I was thinking about this is that in order for AI to be valuable to a construction business owner or the construction business in itself I think you have to know what you need it to do and how to make it innovate for your business and those are some of the things that I think are really or can be really interesting for us to talk about which is sort of how do you make it work for what you need it to work for if you need it to write for you if you need it to build something for you if you need it to do estimating takeoffs for you those are things that you have to know in advance that you need it for and then you actually go in and do the work behind it to actually make those things happen mhm and I think that’s one of the things that can overcome one of these pitfalls is learning what it can and can’t do learning what you need to know how to train it and make it better and then letting it do what you now trained it to do yeah um I I think some of it starts with learning what it’s capable of like it can do so much so well and then also learning what it can’t do very well or where it has struggles um you know we’ve we’ve both experienced um chat GPT making stuff up like like where did you get that from i’ve had it analyze spreadsheets and give me incorrect data output and I’m like like you forgot this or what about this and it’s like oh you’re right i forgot I left that out let me redo that so double-checking its work but understanding what it’s capable of doing so so you because if you want to ask it to do something you have to kind of know what you want the output to look like so I I like to find uh samples of things I’ve done before and show it like this is what I’m this is what I created for some on this other project now I’m going to give you new data and I want you to create something that’s similar but with this data so can you can you understand what I did here and then replicate it with this new data set this this new uh meeting summary or this transcript if you wanted to write something like a meeting summary hey I wrote a meeting summary i’ve been writing meeting summaries for years here’s samples of meeting summaries and now here’s a transcript of a meeting I had please write a summary and put it into the same format that’s kind of how you train it and that’s maybe a starting point yeah here’s another example of that i was thinking this too financials are really important to most to most people who run businesses yes and construction businesses are no different financials are really important so AI can analyze financial data but if you just drop let’s say a P&L into your chat GPT or whatever you use it’s going to say “Oh I can’t really read this i don’t understand.” Or it’s going to give you inaccurate data but if you teach it what kind of format of data it can read then you have now taught it what it’s going to do for you and you now know on your side well then every time I load that data in I need to load it in that way in that format then it is going to work for you and that’s a way to innovate for your business let it teach it what you want it to do for you not the other way around and I think that’s another thing to think of right it’s a tool like I said before it’s a tool to use it’s a tool like any other tool that does not work by itself right i would agree yeah you know it is it is meant to I’m using it to like enhance productivity or um analyze big sets of data like massive spreadsheets that would take me a couple of hours to sift through and and do the math and find the trends it can spot those in seconds uh if I’ve taught it what to look for so giving it again giving it sample data like here’s here’s I was doing this the other day like here’s three years worth of P&L data by month i want you to spot trends and revenue fluctuations give me the highs and the lows you know spot those those uh yeah trends are repeating patterns yeah uh hey looks like every fall your revenue drops off and your costs are flatlined or your costs don’t go down in in relation to your profit shrinks in those months having it do things like that are very basic but but that’s some of the basic kind of financial analysis we’re talking about here yeah and it’s not going to do it on its own right so that’s it doesn’t know what you want one of the examples that I that I like to use is you know cutting your grass you know you can cut your grass in a couple of ways but the easiest thing to use is is a lawn mower but the lawn mower unless you’re you know very very very lucky and you’ve heard of something I’ve never heard of you have to push the lawn mower right you have to push so you walk behind it you do a little bit of the work and then you the tool does the work for you that’s kind of how I feel about AI we we have to still do some work there and I think one of the pitfalls that we can overcome by thinking that you know AI is some magic pill is to understand that it’s a tool for us to use but we’re important we’re important in the AI infrastructure we as people we’re integral we’re integral exactly it has no purpose without us yet yet it’s not self-aware yet it’s not Sky now it’s not okay perfect so as we as I was thinking through the sort of thinking that AI can be a pitfall i was wanting to make sure that we keep a positive spin on things because I personally believe that it is a gamecher for the entire world including the construction sector it it’s a game changer and I think it’s important to maybe talk through some of the ways that it can be positively used keeping in mind that you have to do the training do the work push the lawn mower a little bit right so I was thinking what what are some ways that we can help people think about things that AI can do for them with the right training and tools you kind of brought up financial analysis one of my examples was going to be estimating takeoffs you know that’s one of the things that we hear from our clients a lot of times something that it can do that is I don’t want to say noncreative but those are just facts right looking at something and saying this x you know x is x and y is y now so what are some other kind of things that we can think through that AI can be a gamecher for the sector or things that you’ve heard from some of your clients yeah I mean the financial analysis is a is a big one because um I’m working with a client right now and and they have a monthly work in process a whip meeting and they go through sometimes hundreds of active projects over the course of several days and they’re and they’re uploading all their projections and everything like that well if you can dump that into a big spreadsheet then the AI can do all the analysis and point out or find the the projects that need the most attention yeah and kind of help you prioritize where you should be spending your time these are the projects that have the most risk or they are showing early signs of project fade or cost overruns or schedule extensions and having the AI kind of highlight that stuff for you helps you identify where to focus your time i think risk analysis is another huge I just there’s so much risk in what our clients do and what we do and risk is just a part of life but looking at something like say a contract right a legal document you reading a legal document looking ahead to a project that’s you know failing um trying to find a way to let the system identify risk for you but still keeping in mind that you have to you have to do your part right you have to load in the right data you have to look at it you have to check in on it you have to you know verify it so I I I think there’s a huge upside that we probably haven’t even seen yet um for what it can do for construction companies there’s there’s a lot out there and there’s a lot of companies that are developing their own special AI agents or software platforms to solve specific problems or pain points in the construction industry you know we’ve talked to you know we know softwares that can help do the purchasing the invoice processing as well as the field ticket match to make sure it’s a three-way match and then make sure that your invoice matches your purchase order and your delivery ticket for price quantity and everything was you know actually delivered no back orders things like that that’s a very specific one there’s ones that can do on-screen takeoff automatically now they can do the counts and the linear takeoffs and all that kind of stuff and dump that into your estimating software so you can run your your pricing math on them y um there’s soft there’s things out there now that are using drones to go scan job sites to measure progress and safety safety violations absolutely unsafe conditions like it’s it’s it’s pervasive already and we’re just starting it’s just starting i’ve got a client right now we were just with them and we were talking about daily logs and we were talking about taking you know photos you know in your daily log pretty standard they’ve implemented drone daily drones they send a drone up three days a week so yeah it’s not daily three days a week they send a drone up take full sight photos bring them back and then I said “Well then what do you do with those?” And they said “We just drop them in the daily logs.” And I said “Well are you looking at them?” And they’re like “Well not really.” And I said “Well would you ever would you ever consider using some kind of you know AI you know system to actually look at those photos and actually help you identify what the site sees daily changes?” And they said “That’s a great idea.” So I think that people are sort of partly there with what they’re using it for but that’s but that’s where the human interaction becomes really important you have to think ahead this goes back to what I was saying earlier like they need to understand what it’s capable of then they can start going well if it can do that then this is what I would like it to do with that skill set give me this type of an output think or reverse engineering I would love to have this can the AI do that I would love to have a yeah uh a financial snapshot emailed to me every Monday of all my open projects by project manager can the AI do that well if the data is there yeah we can you know the AI is able to do that work and then the automation of sending out an email is very easy but yeah understanding both what you want it what you would love to have kind of you know magic wand what what I’d love to have is X can the AI do that yes or no or not yet or not yet or can it do part of that and then does you need a human to kind of take it across the finish line all of that and and that’s changing almost daily for sure and and I think where we kind of started the episode where was talking about how to maximize the value of what AI is I think it’s interesting to think about it this way But AI is like you know I think you just said it it’s changing daily if not hourly i mean it is vast and I think maximizing the value of what you need it to do is the part that’s the pitfall that I think people can overcome they can be thinking ahead for what they would like it to do it might not be that way today might not even be that way next week but in six months it might be and if you’re kind of lagging behind AI I think that that’s probably not a great place to be you know I think I’ll say this i think other than maybe a select few of cutting edge AI developers I think we’re going to be playing catchup with AI for years because it’s going so fast that we can’t even we don’t even know what it can do now much less what it’s going to be able to do in three months and I I think we’re going to be it’s going to be leading us down the path of how powerful it’s getting it’s not us that’s trying to push it it’s getting smarter faster than we realize and and I think that’s a great way to think about it that that could happen and it probably will happen but one of the things that I think is valuable for people to think of is how can you become the master of AI at least at least in your business in your business and not the other way around right think think ahead plan ahead use the tool like you would any other tool you know we talked a little we talk a lot about software here right and you mentioned something earlier about software i I I think if you think I would think about software as an example here Procore really obvious you know a lot of people use it excel and Procore are both tools right sure you can do a lot of what you can do in Procore in an Excel document it’s it takes longer it’s more manual sure prone for error but then Procore comes along and says “All right we’re going to put all that into a system.” So those are two tools that do relatively the same thing one that’s how I like to think about AI what is the next version of the tool that’s going to help you run your construction business going to be think ahead plan ahead make sure you’re kind of on the cutting edge of that so that when it’s ready for you you’re ready is kind of how I think about it yeah i’m I’m what instantly came to mind was how long until it can build a real CPM schedule not like like you give it the plans give it the takeoffs and then it knows how long it takes to build things because it’s got enough historical data from real project builds that it can build a realistic uh critical path schedule for a project for like a large project not a small one and I think that’s I don’t think we’re there yet but I think that’s probably within a year i I think there’s so many areas where we’re going to be in a year or less another one that just came to mind when you said that think about the ability to think ahead and have your AI alert you to a future potential safety issue if your AI can tell you that someone is about to get hurt like high probability high probability of stepping in the hole they’re in a ladder without being tied off something like that and it took a photo of that and it kind of alerts you i mean there’s a million ways that it can go and I think that that to me is kind of where I wanted to kind of this to land is now you’re reminding me of that Tom Cruz movie where the crime hadn’t actually happened but it was very likely about to happen and they started Minority Report yeah and they arrest people in advance right we want this to be a good thing we want it to be stop someone from getting hurt but so they don’t ever get hurt in the first place but but I think for construction there’s so many ways for us to think about how to make AI be the tool that it is and will be as opposed to being a slave to it and that’s kind of where I think the pitfalls are coming from and that’s kind of where that commercial was kind of leading was don’t make sure that you are not make sure that your AI is not dragging you down is is sort of the overarching kind of thought i don’t I don’t I never thought about it as becoming a slave to your AI i think about people who adopt it and leverage it and people who don’t adopt it and get left behind i think both of those are problems and I think they’re both really important to think about and as we continue to talk about AI which we clearly will i I just don’t think that there’s we’ll probably never run out of conversations that we can have about the best way to maximize AI at some point we’re just going to have AI sit here and be interviewed and we’ll be able to interview i could inter we could interview our our AI right now one of the ones that we use and have it tell us what see how it tells us that might be fun we could turn on the uh the chat feature with chat GPT and have have a conversation with it that would be interesting maybe that’s the next episode we’ll do it um well I I think I think it’s been I think this is exactly kind of what I wanted to hear and talk about was you know is AI a pitfall how to maximize it as a tool how do we think about it for our you know our construction clients how do our clients think about it um what sort of your overarching thoughts on how to not let AI become a pitfall and how to maximize it for our construction clients well one would be start using it and learning how to use it yeah uh because you don’t want to get left behind and then and then you’re out of business because at some point you’re not going to be competitive without using it because everybody who’s using everyone else is going to be they’re their quotes are going to be more accurate they’re going to be able to turn things around faster they’re going to spot things in advance they’re going to have better risk profiles they’re going to have lower accident rates all the stuff we just talked about yeah so they’re going to be highly competitive and if you’re not part of that part of that group you’re not going to be as competitive and you’re not going to win as much work if any so one would be don’t don’t miss out don’t be scared of it just don’t miss out don’t be scared of it uh two would be I think going back to what you said as it being a tool it has a time and a place and it has a use case there are other times and things that you want to do or need to do in your company where it is not the solution it’s not suited for everything so learning where to apply it and not trying to force it to do something that it’s not ready to do or not meant to do or that you haven’t trained it to do or that you haven’t spent the time doing or that you don’t fully understand and you’re hoping that the AI will make up your shortfall i I don’t ask the AI to do anything that I can’t do myself because I need to be able to double check its work so don’t lean on it like I don’t have to learn how to do that because the AI will do it because the AI can make mistakes and if you don’t like financial analysis and stuff if it doesn’t if you’re not smart enough to check the AI’s work find somebody who is so that’s a way you could really that’s that’s a that’s a a problem or could be a problem it could be a problem people leveraging it or relying on it to give it real trustable output when it’s not meant to do that it they there’s a disclaimer on chat GPT it’s like it can make mistakes please like double check all the work i would sum that up as to say don’t take it for granted yeah is one of the ways that I would kind of put that right it makes mistakes all the time when I use it yeah and and uh the other thing that comes to mind is like it as we said it’s evolving rapidly yeah like what it can and can’t do today is not going to be the same as what it can and can’t do a week from now or a month from now even and and definitely six months from now it’s going to look a lot more powerful and be able to do a lot more so stay up with it as best you can i was going to say stay tuned is what I was going to say standing system it’s not like Microsoft Word where every couple of years they roll out a new little button this thing is changing and evolving and getting more and more powerful you know as we sit here so as expected this has been a really interesting conversation i knew it would be and I I would say thanks to all of our listeners who you know tuned in to listen to this to this episode but and you know if any of of those if any of you out there would like to talk to us more about our AI experience we’d love to talk to you about it there’s an endless array of conversations we can have for sure and I think that we’re gonna continue to talk about AI for the rest of our our lives for the rest of our lives yes yes all right thanks Greg that was fun uh check us out on Apple Spotify and YouTube