The Real Reason AI Hasn’t Taken Over Construction

talk about Autodesk, right, [music] where you were previously. Yeah. >> You helped them shift from that from desktop and application to [music] cloud-based SAS, which is the way most the the industry has been going. >> What do you think the biggest barriers that keeps [music] these firms AEC firms tied to legacy tools? >> I mean, there's two things that really pace the level of adoption of [music] technology in the this A world. One is simply the risk everybody takes on. So, I mean, like, you know, you're building things [music] that need to stand up. you're signing your names on documents that you know sort of guarantee the safety [music] and sort of the performance of these buildings. So people think about risk differently [music] in this world. What do most successful AEC firms have in common with the ones that are still struggling to modernize? On paper, surprisingly little separates them since both groups are usually buying the same tools and reading the same headlines about AI. [music] The difference typically shows up somewhere else entirely. whether the organization has rebuilt habits, workflows, and mindsets that were built for a desktop era because no platform, however sophisticated, can outpace a team that is not ready to change the way it works. Joining us today is Amar Hanspal, CEO of Motif, where he's building AI first cloudnative design software for the AEC industry. In this episode, we'll speak about why AI hasn't yet transformed AEC the way many expected, what value of AI and design turns out to be once you look past image generation, and whether AI is handing creative agency back to architects rather than taking it away. If you are a firm leader trying to figure out where your organization stands or wondering why your last technology roll out did not land the way you had hoped, this conversation is going to give you a better lens for thinking about what readiness looks like. But before we jump in, I want to tell you about AECPM Connect, a series of in-person events we've created at EMI for AEC project managers and [music] the leaders who develop them. If you want to stay on top of PM trends, elevate your team, and deliver stronger project results, these full day events are built for you. I'm coming to Kansas City this October. Learn more and register at apmconnect.com. With that, let's jump into today's episode. Okay, it's now time for our conversation of boot with Omar Hansball. Omar, thank you so much for taking the time to join the show today and welcome. >> My pleasure. Thanks for having me. >> Absolutely. So, uh, we are going to dive right in because I think this is going to be a very interesting conversation for the audience. So, please be sure to follow along and and pay close attention. So Amomar, where are you today in your career and then how did that path lead you to your current role at your company Motif? >> All right, sounds good. Uh, I'm the CEO of Motif System. So this is kind of like this is maybe my fourth role over 30 plus years of uh working in the software industry. I've always worked in building software for physical industries. Uh my path here was I started at Autodesk uh which is a company that builds software for AC and manufacturing companies way back in 1987. uh I was a mechanical engineer by training but I switched to the whole universe of building software for physical worlds back at that time and you know I grew up basically on the customer side of the the operations where I was doing technical support first and then product management and so you know over the years I've just sort of learned to uh hear and understand people's technical problems like what are they trying to do what are they trying to build what kind of support do they need in software and try and translate that into meaningful things that people can can use. Um and uh yeah and I've really developed a deep appreciation for the built world and this you know these are my two loves. I love the build world and I love the building software and this gives me a chance to do both. And you know hopefully now at this this is probably the last I you can never say never but I think this is my last job in some ways uh that I and I want to really make a difference with uh this particular project for the for the world. >> Absolutely. So I'm already talked about Autodesk right where you were previously. >> Yeah. So you so as as I understand it, you helped them shift from that from desktop and application to cloud-based SAS, which is the way most of the the industry has been going. >> So and and funny enough, I was actually just having a conversation today about a firm. They still have all of their their files on a server that's located in their office, right? So they're on prem and not in the cloud. So I I I I laugh because as you know, different people, different companies are at different stages, right? So my question to you is what do you think the biggest barriers that keeps these firms AEC firms tied to legacy tools? >> Yeah, my sense of it is that you know the the industry has I mean there's two things that really pace the level of adoption of technology in the this a world. One is simply the risk everybody takes on. So I mean like you know you're building things that need to stand up. you're signing your names on documents that you know sort of guarantee the safety and sort of the the uh the performance of these buildings. So people think about risk differently in this world than they do in other industries. So I think people h need a much higher level of confidence when they're adopting new technology and things. But that there's one element of like this isn't like my my my profession software development people change tools all the time because the level of risk is kind of different and the level of risk tolerance and when you're writing software is like all right I'll try this if this little language didn't work I'll switch to that when you're building things that need to stand up and keep people safe I think you just think of it differently the second thing I think is that you have project cycles that can take anywhere from 2 years to 5 years you know you're building working on a building or you're working on a piece of infrastructure and so your adoption cycles you generally you won't change tooling in the middle of a project. So your adoption cycles naturally become longer and so I think those two things sort of change the core dynamics of adoption. I do think there is overall you know you the industry switches when there's a younger generation moving into the workforce. So it kind of needs that injection of fresh blood as well from time to time to sort of push it along. It in of itself it doesn't sort of have a natural tendency to look for frontier technologies. It kind of tends to need a push from time to time. So I think that's why as you're pointing out your the customer you talk to like it doesn't surprise me that people still sticking with on-pre stuff not because yeah because they're just trying to be very cautious in their thinking you know they're not pressed for uh you know trying to take risks and things like that. Yeah. >> Transform your technical [music] experts into impactful leaders with EMI's engineering leadership accelerator tailored for AEC professionals. This program [music] enhances communication, delegation, and decision-making skills. Flexible, interactive, and designed to fit your team's schedule. Elevate your team's leadership capabilities today. Visit engineeringmanagementinstitute.org [music] and click on corporate training. >> Thing that I notic is that you'll you'll tend to hear the loudest voices that talk about how the industry is behind are typically the ones that don't have lensure and aren't responsible for public safety, which is kind of ironic but what I will tell you know anyone who asks is that it's not it's not that there should be FOMO or your business is going to evaporate overnight because you don't adopt. >> Yeah. >> What I say is the firms that do get a head start that advantage is going to compound. So the earlier you get started the more of an advantage you'll have. >> Exactly. And you know, I mean, Nick, to that point, there are moments in the technology cycle where it is really a bit of a step change. And if you don't do something to stay up, you're going to it's really going to hurt you. And I mean, the last time that happened was when the industry adopted BIM, right? If you were still or even CAD if you think way back to the 1980s like if you were still drawing by hand today you would be so far behind if you were you know like there are moments where you say hey these are step functions I really do think that AI right now there is a moment in time where if you don't start adopting it at a reasonable pace you will be very far behind you know in 3 or 4 years like this is there are some technologies that are not like the others that you have to lean and do that's that's all I would say. So >> absolutely and and and you know to be clear to the audience right it's like it's not going to happen tomorrow maybe not a month maybe not three months maybe not even a year but at some point that compounding Yeah. goes on like an exponential path and then suddenly you're you're >> left. There's this famous saying that says change happens slowly than suddenly, right? It's like it's like these moments where you don't feel it and then all of a sudden it's there and if you haven't kind of had your toe in the water, you know, you just kind of are, you know, washed over by the title wave, you have never sensed that coming. So I think that that is one of the just the way you have to prepare yourself for that. So you can't just be on the sidelines. You have to actively, you know, participate in this and decide what makes sense and what doesn't. Yeah. So >> beautiful analogy and and it can be hard for us as humans because we're so used to living in this linear world. >> Yes. >> And then this exponential technology comes along and uh yes a colleague of mine for it just imagine yourself sitting on a graph, right? you look over your left shoulder, it's like, okay, you know, everything's in linear world or uh >> you know, if you to call back to Nine Tab, right, Mediocran, >> and you whip your head to the right and you're like, what the heck happened? Now you're an extremist and that exponential curve and it and it happens and it's like uh what the heck >> and it is happening. It just hasn't really reached us as an industry to that point. And that's actually a great segue into our next question, Amar. So >> you hear that you'll hear from various parties, research studies, etc. saying that AI really has not transformed this industry yet. So >> why is that? And then how does that relate to what you guys are doing as a company? >> Yeah. Well, I mean, I think goes back to the previous question you asked me is that because this is a place or an industry where people build things that have to stand, you can't be 99% accurate. Like it's this thing of like this bar of like the the work you do has to be perfect or almost overperfect. Like I mean you you know as far as I know when people build buildings the safety buffers and all they put in they try to be like 120% right not just 100% right. So I mean AI is a probabilistic technology at the end of the day when you look under the hood and is still based on underlying you know technology all these transforms all these things that they do that they are probabilistic. So I think until it gets good enough on on on solving certain problems and delivering physically accurate things, its use is limited. So I mean to be fair in my conversations with customers, people are using AI for things that don't require 100% accuracy. So like visualization is one thing in AC. I see that I see uh people definitely using it for text for checking you know documents [clears throat] and proposals and things like that where you know being 99% right is good and to the point you just made where change happens you know ex exponentially this is a technology that is moving at a rate that is nonlinear and you know human brains can't comprehend that but I do think the this the speed at which AI is improving. It's going to do more and more solve more and more problems for people in the built environment very rapidly. So I think you can't be oh it doesn't work let's ignore it. You have to kind of engage and see what it can do. So companies like us motif we're building software that wraps around AI to solve those problems for customers. So we started with visualization, we're migrating to code checking and you know document checking and then we're migrating to actually doing generative design for simple things like you could think about interior furniture. You don't have to be 100% accurate like you could be lay chairs out or tables out in an interior space you know being 99% accurate. So we as a software company are applying AI to design to review to visualization for our customers and you know helping people transform the way or shorten the cycle improve the decision-m process that they go through uh during the during the front end of the building design process >> and Amar I think that that brings up the the topic of discussion of creativity or agency both are are related. So we don't >> don't hear about it as much from the engineering space because I think the perception of engineers is that it's all qualitative which is absolutely not true. It's it's probably more qualitative for us than it is for architects. But the worry is >> AI stripping away that creative agency, innovation, creativity from architects. >> Yeah. >> Which the way I've heard it is like okay like you're going to get put into this box. So you're going to be essentially become like serve it to the machine. But is that the way you guys see it or we missing something here? >> Quite the opposite. Nick, I'll give you the answer in two parts. One, I feel like when I talk to people in the industry and we really they feel like even today, way before AI, their time is not spent doing creative stuff. It's doing all the other stuff around it. like the the the daily work of an architect isn't like they're sitting in the corner coming up with these great ideas. They're like swamped with a hundred things that they have to do to take the idea and express it or communicate it. Right? That's where they're like they're in in meetings. They're in like hey can you create this document set for this person to look at it? They're like sitting getting feedback from someone. They're checking for this. They're doing that. Like that is where AI can be helpful is to take all the work that you're spending your time in and doing that for you. And it's like I always think the AI is like your helpful intern like would you like to say to somebody let me do the creative work. Can you do this document set for me? Can you go get the feedback from Bill over there in the corner? Can you go find me that detail that we worked on structural detail or the vapor barrier we built? you know for the last project like that's the time savings that AI can do. So I think that's the core thing is can can this thing become the thing that frees you to do the creative work. So that's one thing I would tell you and if it that just sounds like words I would just tell you that in reality let's take one industry that has transformed itself using AI and again that's my industry software development. If you look at the way we write code today and you walk talk to any of the people working at motif today many of the engineers now are not writing by hand every single line of code they're asking AI to do that so where are they spending their time they're actually spending their time in creative judgment so I I look at what our engineers or our team is doing now is they're actually spending time on well the customer wants this capability how should we design it like they're spending their time actually thinking about the thing they want to as opposed to can I get every comma, every brace, every memory allocation, did I write this thing and does it have a security bug? Does it actually what like all those things the AI is taking care of and these guys are thinking about okay I need something that kind of like let's say it helps people create structural columns you know like that's where the burden has shifted so I actually am much more optimistic and bullish on the fact that AI will liberate people to do creative tasks then stifle that >> beautifully said Amar and I I'm in agreement but I uh I I think the the the the piece I'm really interested in is that you still need competentmemes behind the technology. >> And how do is is there a way to help or can you can you get what would typically be a 10-year architect in seven years because you're able to cut out so much of the monotony out of the day. >> Yeah. >> Especially when it comes to things that are in a non-digital environment. So the real world, right, physical world, like you can't shortcut any of that learning. Like I can't just I can't just get 10 site visits in a day like I can you know >> well unless we're talking about simulations right but my point is you still need competent operators and those take time to develop 100% 100% I mean I think that that is the you know it's like and can you scale that knowledge that's the other thing that I hear from people is that they have all these people that have developed expertise I'm sure you in your daily life in your job you've got some deep domain expertise that people re rely on you for and then you know you could h help answer 10 questions a day like how do we help you scale that to 20 questions a day can you help the person like that's the other promise of AI is can we take Nick and everything in his head and kind of help that be available to more people in the company that you work in like scaling that kind of knowledge and judgment is the other I think the long-term promise of AI that's one thing that I think over time being able to say this is the way we do these things and help more people access that is another you know promising avenue but it you know that that there's more work for us to do to be able to do that >> and you know that's uh it's a great point and I'll give a shout out to another technology company uh in your neck of the woods in the Bay Area so knowledge architecture who is episode 104 on the show because that's exactly what they do. So, knowledge management, knowledge retrieval, scaling people's knowledge. So, to the audience, if you're interested in that point that Amar just made, go check out episode 104 with Chris. It was a it was a great one. That being said, around the same path, Amomar, but uh I've just heard for the first time today the term vibe modeling, right? So, I'm sure a uh predecessor to vibe coding, but tell me more about that cuz I think I can grasp the concept, but I want to hear what you have have to say about it. >> Well, I I heard customers throw this term around. I was at a sort of an evening session with these uh amazing architecture firms that were looking at using AI for their operations and they were starting to use that term. It it was a way of them sort of saying hey can we generate ideas using AI that are sort of like examples so can we create like eight options for customers to for their clients to look at for you know something that they were working on. So I I think the idea here is vibe coding was this term that came up where it was like hey man I'm in the flow of things uh can AI generate code for me and I can look at doing lots of you know simple simple little things and I think in the world of architecture in the built environment this is just not a fully fleshed out thought but it's an idea here where back to sort of the creative exploration we were just talking about is that can AI come up and help you think through Hey, there are eight ways to solve this problem. Can you do go do that for me? As opposed to me having to do each of these eight things. Again, we kind of, you know, AI can help me do six of them and then I can look at that and say, let's doubleclick on that particular idea. So, I think it isn't a widely used term. We've been at botif using agentic modeling a little bit more to just kind of describe the idea of like you can have agents that go do a task for you and come back and propose something and you know whether that is a piece of geometry or you know go research a product on the web and say you know find me a window that fits these dimensions and these conditions like that's the idea of like using agents to do the work delegated to agents to do the work for you but it's early early days and we we still it's not as well thought through or well practiced as an idea of what in software you know software development I think by the way the one other thing I would say in software development you don't hear the term vibe coding anymore because actually vibe coding has gone and been replaced by people actually just coding so people are not you know what was like the experimental thing like software has shifted to just using AI to generate code so I've actually seen that term disappear over time So I would say it'd be interesting to see if you know people customers stick to the term by modeling over time or not. So >> well and so far what what this whole revolution has done for the lay person, right? Speaking as a lay person. Um when it comes to anything software related is it's it's really opened my eyes to how much you when you use a computer right it's a user interface but at its core it's just binary. It's all zeros and ones that if you think about >> think about vibe coding, it's just English to whatever form of code and eventually some layers of abstraction to >> Exactly. >> So just take that concept to modeling. >> Yeah. >> Okay. So you're telling the computer what you want to do. >> It's converting it >> to something in Revit or other other BIM. >> Yeah. >> But it's all just zeros and ones. >> Exactly. I mean, I think that's a really good point, Nick. I mean, the history of compute has been a series of abstractions. I mean if you go back to the beginning there was like uh assembler the the rawest form is like you know microchips have like their language that they run the processes on and then we had assembly language that drove that and then from assembly we got to the first level of you know things like C or C++ that you compiled into assembly language and now we're going one level above that where you're like giving instructions that generate the code that generate the assembler that generates It's the instruction set for these things. So I do I think there's a direct equivalent to you know the design practice where you're going from you know handdrawn things to BIM that kind of you started with a 3D model and generated the lines and now you can go one level above where you can give instructions that generate the geometry that generate the drawing you know etc etc kind of thing. It's and I think it's it's interesting when you talk about like so I'll get I'll go to code as an example and code is a great place in my mind to start um because it when you write when you compile a piece of code it either spits back some errors or it doesn't. >> Yes. >> Now just because it doesn't spit back errors doesn't mean it's working as you the developer intended to. But there's a there's a binary state of hey it either works or doesn't work in this context but with what we do as design professionals it's not quite so binary. >> Yes. >> Yes. I mean I think there's a judgment there's taste there's other ways to say it works right in your in your profession. I mean even in software development like it works and then you get feedback from customers that say hey I wish it did this I wish it did that or in this condition it doesn't work. So I I do think that you know in the built environment there's a lot more stuff that happens that you you know I I I think you get feedback on aesthetics, you get feedback on sustainability, you get feedback on all these other criteria that determine whether it works or not, you know. So >> yeah, and and then I guess maybe the equivalent is like the the quantitative portions of the code where it's either it's binary and that it either meets the code or it doesn't and there's no there's no judgment involved which is not the >> other hard part about the physical industry is that feedback loop can take years like >> it's not >> you know not immediate. So it's it's cool because there's so many parallels that you can draw. >> Yes. >> Um >> it's just a it's just a different industry. Is that a different industry, different pace? I think that's an excellent point. It does take longer to do things in the physical industry, >> physical world. So, so you know, talk to us about buy versus build. So, this right, so this market is getting flooded with AI native platforms no matter what segment or niche you're in. So, how do leaders think about that? Should we buy? >> Should we build? Yeah, I mean I I think the first place to start is like you know I if how unique is the problem that you're trying to build for like I mean I I do think that you know people like building things in uh you know some fun to like creating things yourself there is a natural tendency to try and build things but if it's really unique to your firm it's worth doing that if it's a problem that is going to be pervasive across the industry. It is more likely that someone will build a solution that kind of learns from a bunch of people and drives that and generally moves faster because a single team let's say us is getting feedback from 20 customers as opposed to being like you know and so it's forced to move quicker than like it's moving at the pace of one company. So I I do think that I there are many things that are more suited for a not to be just built but to be bought and and customized if you will. The other thing is that when you start it's always easy to start something it's really hard to finish it. So like the version one of any piece of software has is is very exciting and you can get it done and maybe version five and maybe version two you know when you write software you have to sustain it for years for it to get better that's the hard part and that's the decision you have to take so let's even take AI right so you write something and you write it on uh opus 4.5 well then you know 4.6 six shows up and then Mythos shows up and then Fable shows up and then GPT 5.6 six shows up like you're constantly like you have a team now we do that does nothing but what we call eval like every time something shows up you figure out what do you have to change what do you have to keep the same what is the new thing you have to do so you have to be of enough scale where you can afford to sustain that so my my net net is I think it makes sense to build some things but I don't think it makes sense to that to be the default I do think that it for many missionritical things that you have to take years to build it's more prudent to buy that's the way I would describe it you know as a software development company we don't build everything ourselves we I even anthropic like this is the irony that people keep pointing out is that even anthropic uses Salesforce they use rippling they use all these companies other companies to manage their HR to manage their CRM to manage their you know they they themselves although they have this incredible capacity. They've just decided that therefore these certain capabilities. They're just going to trust people who only focus on that, you know, kind of thing. But I'm sure Anthropic has written tools for many things that are, you know, just are custom tools for their company. So, >> well, it's it's a great point, right? We we all only have 24 hours in a day. >> Yeah. >> You know, a certain portion of that is sleeping and not working. So you're you're you're resource constrained in time. So like sounds to me like anthropic is just making a an intentional choice to focus on their core business instead of getting distracted stuff that's not >> and who knows, right? Like uh I think >> I think Open AI called them, you know, side quests at one point that they just killed a bunch off to go focus. >> Exactly. So, and furthering the irony train, the the core human principles of of just picking what you focus on and setting down a path don't change just because we're developing perhaps the most uh disruptive technology we'll see in our lifetimes. >> Maybe >> that's true. >> Yeah, I agree. >> Excellent. You know, this has been an excellent conversation. I've, you know, really enjoyed the time here, but um if the if the audience wants to reach out to you, learn more, ask you a couple of follow-up questions, what's the best way for them to connect with you? >> I mean, I'm um they can connect to me on LinkedIn, on DM me on Twitter. My email is amar.hunspotif.io. Contact me any of those ways and be more than happy to engage in a dialogue. >> Excellent. Omar Gab, thank you so much for taking the time to join us today. >> My pleasure, Nick. Thanks so much, sir. Until next time. >> Cheers. Please remember you can find the shows for this episode at aechpodcast.com. There you'll find a summary of the key points discussed in today's episode, as well as links to any of the resources, websites, or books mentioned during this episode. Until next time, I wish you the best in all your engineering and technology endeavors.