So the benefit of modular design, modular construction, or one of them, is that it’s a known quantity and you have these units that you know every nut and bolt and detail about. So you can cost things really accurately and you know exactly how big things are, and you can dimension things very accurately. And so you have all that detail that then you can apply in a computational model. we’re doing at Modulus and what I’m part of in the tech team is basically procedural generation of buildings. And so the idea is that an architect or a developer will have a site in mind, and they’ll be able to use our software to rot out the site, and it will generate modular homes on that site. Wow.

Hello everyone and thanks for tuning into another episode of the Bricks and Bytes podcast, your go-to for off-limits construction and property technology. On today’s show we have Sam Grexson, lead computational designer at Modulus. Modulus created the first globally scalable model for the design and delivery of homes. In this episode we discuss effective design solutions, computational design, why digitized design costing programming and procurement is our future and many many more.

If you’re enjoying our podcast, please check us out on Spotify or Apple or wherever you listen to your podcast. And if you enjoyed, please leave us a review. This helps us to get more amazing guests to give you guys the best and most informative content on technology in the build world. And shout out to our sponsor, Beta. If you want to connect with some of the biggest players in the construction tech world, including tier one building contractors, some of the biggest construction tech companies, investors and advisors, check them out by visiting www.d-beta.com and this is www.d-beta.com. You are listening to Bricks and Bytes podcast, where we take you on a journey in construction, technology and business. All right, let’s get this episode started.

Tell us about your journey from a structural engineer to your current role at Modulus. Yeah. In terms of like, taking those kinds of technology side, I guess, like at university, I actually started doing an aeronautical engineering degree and there was like a little bit of coding in that, so that kind of started there. And then I decided to switch to structural engineering more because I didn’t, I wanted to be creating. buildings that kind of were more useful to society and maybe was worried about the career trajectory of maybe like designing planes that bomb and kill people. So it was like a slightly ethical thing. But anyway, kind of digress. So yeah, I worked, went into structure engineering and there’s not a lot of kind of tech in the education there. So I saw a bit of a difference there, which was interesting, but I played around with Grasshopper even like during… university because my architecture, my structural engineering degree had a bit of architecture in it and kept on playing with that and trying to force it into all the work that I was doing, even if it was like very basic mundane stuff, trying to weave it in. So I’ve always had an interest there and ended up doing like a lot of interoperability. So just like passing data between software packages, because as mentioned is, we use like lots of different structural. lots of different packages like Revit and analysis and you always like struggling with passing data between the two. So I did a lot of that and then developed optimization algorithms and stuff alongside my structural engineering. And then a year ago I took a bit of a kind of career shift and joined Modulus. And what kind of excited me about Modulus was that like beforehand I did some like amazing projects, really interesting. amazing architecture like Foster’s and Partners designs and stuff like that. And that was where I saw kind of computational design being used mostly, but a lot of it because of the like the crazy geometry that you get on these like projects. And then Modulus was this first time that I saw like an opportunity to apply computational design to the everyday and going back to the homes that people live in. And that kind of goes back to why I went down the kind of structure engineering route, as opposed to the aeronautical was like being able to design homes and stuff that like beneficial to society. Yeah.

Rather than the luxury things that only affect the rich. So yeah, so it went full circle there in a way. Yeah. So that excites me about the changing computational design is this in a way trickling down to the more mundane stuff. And whereas before it was just like the fancy projects. And that’s pretty cool seeing that change. Sam, go on, man. I’m very happy that I can hear you very well. And you can hear me as well. Sam, you said that you were quite heavily involved in interoperability between software and usually engineers or professionals within the design industry. they struggle with this, with these models transferring from one to another. I think, I feel like this is a major problem actually. And the reason, and people also don’t want to take responsibility for their models and there’s a caveat that the models is only used for X, Y, Z. It cannot be used for the rest of the things, for other professionals to work on the models sometimes. And why is it so complicated in general? Why can it not be much simpler? And yeah, what’s, what are your thoughts on this? Yeah, every time you move between one piece of software to another, you lose data. You lose something, you can’t keep everything because it just… Yeah, there’s always a compromise and that kind of makes things difficult. But also I think a fundamental problem is that every bit of analysis you do or every piece of software requires something different. So one task might be to just generate a nice render. And that doesn’t require like your geometry to be perfect, to be perfectly aligned in your vertices to node out and stuff like that. It’s just got to look and then when it comes to like an analysis that an MEP engineer would do, everything’s got to be completely like watertight because you don’t want energy kind of escaping it. You don’t want holes in your model. But it can be 2D. It doesn’t need to be 3D. And then when it comes to Revit, everything’s got to be 3D. So kind of software requires different things and it’s like translating your geometry into something else and closing those holes or making it 2D and 3D and kind of meshing it for analysis and stuff that makes it quite tricky.

So like broadly speaking, do we say that we don’t have enough computational power to deal with like having one stop shop for all of the designers? I’m very broadly speaking about it. Yeah, the problem with a one-stop shop approach is that I just, I don’t think it’s possible. Like I don’t think it’s possible for, I don’t know if Autodesk eventually trying to take over the world and trying to do everything, but it’s really, you need like a giant to, if that’s possible, then you need like a giant to do that, to try and buy up as many companies as you can and like, like shove them in as plugins to one piece of software. I think that’s the only way that you could. That’s definitely something that’s becoming more popular in construction. You just mentioned Autodesk. They’re trying to get as much as they can bring in. And also Procore as well. They have a marketplace where everything is just plugged. Other people develop software and it’s all plugged into this central place because a lot of software is very fragmented and you have to have 50 logins just to run one project or whatever it is. Yeah. But the other approach is, I guess, that is quite common is that you have your one… kind of source of truth that you try and keep in one place, and then you feed it out to the appropriate software. So you try and build up as much information in that central kind of source of truth model that you can, and then you use plugins to fire it out. So people doing that in Rhino and Grasshopper is that kind of source of truth because you can, there’s lots of plugins that feed into that. You can send it to Revit, you can send it to do daylighting analysis and stuff. So that’s the other approach rather than just having one piece of software. Yeah, it makes sense. And when you have a separate company, like they’re more specially focused on solving one small thing. So exactly. Yeah. Rather than doing that very well. Yeah, exactly. Yeah. Okay, I’d love to ask you more about how like someone may get into computational design that’s like going for a traditional kind of career like a structural engineer, but we can touch on that later. Before we get there, what is your definition of computational design? Because it’s probably some people may only be introduced to this term, but it’s obviously been around for perhaps a lot longer than we envisage is just more recent with like technological evolution, it’s becoming a lot more advanced. Yeah, I don’t know. I actually struggled to define it.

I guess it’s just like designing with computers in the broader sense. And maybe it changes as well as technology changes. Like the yeah, like the term AI kind of changes because what’s clever today, and yeah, in terms of technology and seems very clever changes like 10 years later, it won’t be AI anymore, because it’s not clever anymore. I’m sure. And I think computational design is similar, but I just, I like to just define it in the broader sense of designing with computers, I guess. Keep it simple. On this technology change, like when we spoke first in December, there was no concept of chat GPT, right? At that time, it just emerged a few months ago. Yeah. So things changed quite quickly. Do you see any machine learning applications to computational design and construction engineering? Yeah, it’s funny because my opinion on that has changed. Like even in the last six months or something, I was quite pessimistic about neural networks, deep learning, machine learning for architecture. And I think my main kind of hesitancy was twofold. One is that it obviously requires like a lot of data and I don’t feel like we have that data, like we don’t in the construction industry, we don’t share data like wildly. So you’d be what I was saying. Yeah, that’s what I was saying. Everyone has their own. Yeah. Yeah. And then the other thing is that well, my original thoughts anyway, were that like, we need something very specific in the construction industry, we don’t just want a pretty picture at the end. If you want a pretty picture, you can get that you can get AI to or machine learning to generate that if a few pixel values are a bit off, it’s fine. It still looks gritty, but if we’re, what we want is something that stands up or does it. And I was, that kind of concerns me the accuracy, exactly. But I think there are use cases and using machine learning. And one way of getting around the data problem is I see people using deep learning for surrogate models.

And that’s an approximate model to your, to what you’re actually trying to do. So for example, computational fluid dynamics is quite expensive to do. Like it takes, takes like at least a few minutes to run like one analysis. So it’s quite heavy from a computational point of view, but if you can generate those, that data yourself, but you just need a lot of time that you could imagine like leaving a computer to, to run these models for a couple of weeks, then suddenly you have a lot of data that you can. you can put it put into a model which generates an approximation, a surrogate model. But the benefit of that surrogate model, even though it’s just an approximation, is that it’s super quick and you can get results in milliseconds. And then you can start playing about with that model. So that’s a really interesting approach that you generate the data slowly and then you feed it into a model and then you can get snappy quick results and then you can apply it to your design. Say if we move this here, what happens? And then it becomes interactive. You know what I mean? Rather than waiting 10 minutes to see what happens when you move a column. So that’s a really interesting use case. That is super interesting. When might that be applied Sam to perhaps, let’s say the real world. Like it might it be someone who’s designing like a house or a multi-story building or any ideas? Yeah. I mean, anything, anything that. that requires like analysis that takes a bit of time. And I guess like it’s maybe, yeah. I mean, you could do it with day lighting studies, although that’s a little bit cheaper than CFD studies, but like radiation analysis and any buildings that has a building and you put in your context and where the sun is and you can get your kind of radiation studies. So as long as the data like translates to what you’re working on and it’s not like crazy world.

And actually, going back to what I was saying before, the more mundane or typical your building is, the more that it will match the data. Because if it’s just boxes and rectangles, then that’s going to match your data much more if that’s what you’ve trained it on compared to maybe if you stick in Zaha hadid building, then maybe it doesn’t quite translate. It’s slightly more difficult. So again, computational design being used on the more mundane is quite interesting. Yeah, that leads nicely on to my next question. Given that we have used computational design for a long while, what are more some of the recent developments, say, within the last, I don’t know, five years that perhaps people with businesses now are not necessarily using because they started 10 years ago just using CAD and basic system? Is there any recent advancements people should be aware of or any incoming? Perhaps you might know? Yeah, I don’t know. Obviously we’re speaking about it. Deep learning. I can only speak to my experience of what I’ve been doing. So obviously like parametric. You’re way more advanced than most people out there. Don’t know about that. You scroll through LinkedIn and you get like tons of stuff. Lots of people like doing this kind of stuff now. But yeah, so parametric design, like that’s been around like 10 years plus now with the likes of Grasshopper and stuff. So that’s where it started out for me as well. And then beyond that, like once you have a parametric model, you can optimize it. You can feed it into a like genetic algorithm is a kind of typical algorithm that you use, which automatically changes your parameters and optimizes for whatever fitness function you want. And that’s been ticking along nicely for 10 years, I think. And everyone’s caught up to that more or less now. All the kind of big firms at least have a couple of guys who, you don’t know how to do that.

But I guess what’s interested me is actually like, I’ve taken a lot of inspiration from the games industry. They’re doing a lot of like research on what they call procedural content generation and looking at the kind of algorithms that they use and exploring the different algorithms. So it’s not just genetic algorithms and going a bit further. Like sounds are advanced. Yeah, it doesn’t have to be. It’s just, you can have. Yeah, taking inspiration from what other guys are doing. Like the algorithm that was the AlphaGo that kind of be the champion. Yeah, that’s kind of a play. Like they had some deep learning in there, which was like really fancy and cleverly engineered. But the kind of core of the algorithm was something called a Monte Carlo tree search, which is actually quite a simple algorithm. And that was what got them there and looking at like this, a sequence of moves. And that kind of, I find quite interesting, like looking at architecture or engineering as a sequence of actions that you take and framing it in that sense, you can use some of these algorithms. And so I’ve been exploring like a lot of different algorithms recently that’s been quite interesting and fun for me. Yeah, nice. Okay. And so on to like more modulus, should we say related stuff, what role does computational design have in developing? innovative modular solutions. Do you have any examples? Yeah, so the benefit of modular design, modular construction, or one of them, is that it’s a known quantity and you have these units that you know every nut and bolt and detail about, so you can cost things really accurately and you know exactly how big things are and you can dimension things very accurately. And so you have all that detail that then you can apply in a kind of computational model.

And so what we’re doing at Modulus and what I’m part of in the tech team is basically procedural generation of buildings. And so the idea is that an architectural developer will have a site in mind and they’ll be able to use our software to plot out the site and it will generate modular homes on that side. Wow. A modular apartment. Down to the nut and bolt or? Yeah. Yeah, essentially. Already? Yeah. Yeah. Wow. So we have, um, we have our own modular units that we’ve designed to it down to the nut and bolt and so our software is web-based, but at the end you can download the result as a Revit file. And you can download the plans and PDFs of all the kind of metrics. And we even cost it as well. So that shows the power. Because we have all that knowledge embedded in the models, we can do all that and we can design it down to a T. But that’s the fun bit, is trying to put an architect’s brain into a computer or make clever decisions that is not just completely random. And that’s the challenge of where do you put your buildings and how do you solve this problem? And the idea is not to replace an architect. It’s not to like we don’t claim to, to be able to design better than an architect that will spend a month on this, but it’s to speed up the process and allow them to iterate much faster and say, what if I do this or give me some ideas? Oh, that seems interesting. What if we explore that a little bit more and move this here and tweak that? So it’s about iterating faster. basically. And from an architect’s point of view and from a developer’s point of view, it’s about getting some quick results very quickly.

And they might not be tuned 100% developed, but you get some very quick feedback, 90% there, and then they can get those numbers back and understand if a site is viable or not or whatever they might be looking for. This is something that’s like a next thing to me that needs to happen just to improve the project how it happens is there’s an idea, architects have an idea during the project, or there’s some change from the investor’s side. And then there’s five, six people gathering in the room and talking about these changes and then agreeing on the next time scales for the next changes to be implemented for the other people to see what’s the outcome of these changes. And that’s basically a one month gone, 15,000 pounds for the design teams as well. It feels like it should be done much quicker and without so many… professionals getting involved on the, such a deep detail level. Yeah. And I think don’t talk yourself out of it. I think it can build trust as well, potentially, because you have essentially a live model that you can play about with. And so rather than just in conversations with an architect or an engineer saying, why didn’t you do this? And just having to trust them because they say it didn’t feel right. We haven’t tried it, but like my. professional judgment says that’s not a good thing, you can say, okay, let’s do it. Look, you can see that it has this impact and you can kind of play about with that model. So conversations with clients is potentially easier because you have that that you can show them the results for. It’s just not the core project offering is it of of modulus? Is this something you’re personally working on? Or is this what like modulus is going towards right now? No, this is yeah. massive part of what we offer.

But as I alluded to, we do have our own kind of product or design for modular units. But kind of what makes us unique is that we don’t kind of outs the fabrication. So we don’t have our own factories, which as a startup, having factories is a massive investment required to do that. And that makes you more vulnerable. And it makes it harder to scale as well, like trying to scale it. across the country or even globally makes it more difficult if you have these kind of set factories. So we have design, but that’s outsourced to partners. Yeah. And the business model and actually what we’re going to do is to finish on that. Although we also have our own design, we are also looking to partner with other fabricators of modular units and embed their modular units into our software. So the idea is that they provide us with a Revit model that we upload to our software, and then we can start playing about with their units. So that’s another source of revenue for us and another kind of way that we expand our software for different use cases. So the business model and business case of Modulus is to provide software and solution for modular fabrication. Would you say that? Yeah, yeah. That’s the main bit that I’m part of. Yeah. Okay. So it feels like everything is going towards consolidation.

So there is a software, great software, let’s say Modulus, which can provide great designs, which are scalable and used in the various types of developments. And then these can be sent to the fabricators all across the country or in various places in the globe and can be In construction, we are going towards consolidation of the process towards one kind of entity of software provider which solves all of it or is it not the case? So what we’re focusing on is the initial stages, the kind of feasibility concept. So we would expect there to be further development and that’s one of the reasons why we provide export options, for example, to Revit. So we’re not designing foundations, for example, and that kind of thing.

So there’s still stuff to be worked out. And I think, yeah, I think we’re far from having a one-stop shop anywhere. And I don’t think we, yeah, we’re not going to be providing that anytime soon because it’s a massive challenge. In a few years. Yeah. Give us a few years guys. It makes sense because that is like the idea behind modular, right? It’s more consolidated than another process. Yeah. And that makes our job easier to some extent. Yeah. Nice. And what does a day in the life of someone, a computational designer, a modulus look like? So I’m like pretty much a hundred percent coding at the moment, working from home in front of my laptop. Yeah. Just coding in TypeScript, JavaScript. Having now and again, having meetings with the other guys and the various teams that we have, in-house experts in their fields, MEP engineers and architects to make sure that what we’re doing on the computations side kind of matches expectations and is in line with rules via engineering rules and what the architects expect and all that kind of stuff. And how many people work under the software side, technical side? So we’re about 15 out of 15. Wow. Oh, so that’s 20% of the business, right? More or less. Yeah. 30%, 30% of the business. That’s crazy. So thinking about it, like construction company, 20%, 30% of the staff working on the coat, which for example, 10 years ago, you probably have zero. 15 years ago. Even now, even now, the majority of the things, no one’s doing any technical stuff in businesses. So other guys who are working on software, are they also like engineers and programmers or they are more programmers? Yeah, it’s quite, it’s a bit mixed, but I would say, because we’re also split into front end and back end. We’ve got like… the database guys and the guys who are doing the UI and the UX and everything on the front end, making it work nicely with users and stuff.

And there’s me in the middle with a few other guys looking at the algorithms and how we translate architectural intent into a building. So I think that’s the fun bit. So a lot of those guys come from… a developer background, like they’re used to doing their backend database stuff or their front end design or their front end coding and stuff. But then we have a couple of guys more like me who have a more of an architectural background or an engineering background. So a bit of a mix. Yeah, that’s an incredible talent to be able to do like structural engineer background, then go into coding. Yeah, it’s just it makes sense. And I think more people will be required to have that skill set as well as technology advances. Can you share any breakthroughs or innovations that maybe modulus or even you personally have had developed in the previous years? In what sense breakthroughs? Technologically with design or the modular side of things? In construction broadly I would say, right, as well. Even construction, yeah. Yeah, generally speaking, yeah, could be. Like the question I would say, if I could add on this bit, like what is generating that? What is generating that? the possible change? Is it like better computers or is it better software that coders are creating along with engineers? So interoperability between the technical knowledge of the design industry, engineers, architects and the coding staff who can put the things together. Like what is the thing that is improving the innovation also? I think it’s mainly down to a change in attitude in the construction industry more willing to adopt technology. Obviously, there’s a lot of breakthroughs in, like we’ve already discussed in the last few years, in technology and deep learning and stuff. And I don’t think they’ve trickled through yet, really, in a meaningful way.

But they provide kind of inspiration when people maybe realize that there is something to technology and it can do interesting things. But I think, yeah, well… One of the challenges is maybe some companies don’t kind of realize that they need a diverse set of skills, maybe. I think they want a chartered engineer with so many years’ experience who has worked with these materials and that’s what they fill their offices with. And maybe there needs to be… bit more and you’re seeing it a bit more of a mind shift of saying this guy doesn’t take these boxes but has these skills here and need to get more of a broad set of skills in the office I think and with I think that’s what we’re seeing. Yeah, I would definitely agree with you on that. You touch on awareness and obviously with things like chat GPT, like people that is like such pop culture in terms of the tech world, that people now look at that anyone can use it. It’s so easy to use. And you wonder, people are not using that, then what disadvantage are they going to be at in terms of running their business? And other people that can’t use a tool like chat, GBT or whatever, will they fall behind as a business? I don’t know. Maybe they will.

Maybe they won’t. Similarly, if you’re not hiring people that have got like a technical, technologically technical skillset and just like a chartered skillset, then you may be in the same boat. I think so. I like to compare a lot of, as a structure engineer, I like to compare this with FEA. For example, finite element analysis, just, you know, what every structural engineering firm uses to do their like analysis and make sure a column is strong enough and they build their models in, right. That’s just so commonplace. Everyone does that now. Bar like a few very traditional like companies, but like… Martin, do you use it? Yeah, we use software and software is based on this. Yeah. Okay, cool. But yeah, so it’s really commonplace, but at one point it was just for those fancy projects. Like it was just when you got like a Sydney Opera House come along, maybe like, oh, let’s get the computers out. And I think it’s the same. Can we blow the dust off this? Wipe down the screen and hopefully it switches on. If it had a screen back then, I don’t know. Yeah, so I feel like it’s just a matter of time and… what feels fancy now is not necessarily going to be fancy. Yeah, absolutely. What’s your thoughts on like generative design and AI driven tools in terms of designing and even bonus point, can it be applied to modular construction? So when you say AI, what do you mean? Or what do you mean? Generative. You mentioned earlier that you had someone that was using like neural networks to create a comment with the exact term you use. Yeah. Diffusion models. Diffusion models, yeah. Yeah. So you should check him out. So the company is called Plan Finder. Okay.

So he uses diffusion models, which is like the mid, the same stuff that mid journey is same technology mid journeys based on. But instead of generating pretty pictures, you generate. apartment floor plans and so it requires different debt to be fed into it. And yeah, I think it’s quite interesting. But again, like it’s that question of data and what data you have. Like I’ve seen people kind of toy with the idea of using like the data on OpenStreetMaps or Google Maps. to train a model. So this we have, although not like detailed data, we have street networks and building outlines. And can you train a model to produce building shapes at high level and then extrapolate or go into more detail afterwards? So a kind of more urban design perspective. Yeah. Sam, I just wanted to gauge your thoughts on talent in construction yourself, a structural engineer, transition to computational designer. How do you see construction being attractive for young people to stay within the profession of engineering or architecture? You mentioned earlier game industry. I think there’s not enough gaming in the design, I would say, for young people also to get attracted to it and be willing to work in the construction. What do you think? Yeah, I do think it’s challenging. You can disagree, of course. better paid jobs out there with a similar set of skills potentially going into like more traditional like coding or something is some well paid jobs. But I think like partly the kind of love of the like the industry or like what you what I really like is like being able to see and touch like something that you’ve worked on. creative. Yeah, exactly. Rather than it maybe just being something on a screen. So it’s like I said something similar actually.

Yeah, they had, yeah, we had one recently and they said that was the exact kind of reason for being in construction was that they loved being involved at the start. Yeah. And then seeing the emporer as opposed to creating something that then just like disappears from there. Yeah. So I think if it was all about money, then I probably wouldn’t be in this career path because but it’s I’m not saying it’s is badly paid by any means, but the skill sets are maybe transferable to better paid jobs. But yeah, I like what I do as well. S1A3L 11 It’s absolutely not all about the money either. What about attracting people? Because we obviously have a big thing on this podcast of, I say it a lot, but attracting the next generation to construction. People don’t want to work in a stinking office that’s cold and dusty. They want to go and work in all the glass panels and No, so any thoughts on how we can attract people? Yeah, I don’t know. I’m biased. I really like the kind of, I like the computational design bit. If it wasn’t for that, like that, that really excites me. And by like tinkering with things and that’s yeah, writing a bit of code and then seeing it work on your screen as well is really exciting for me. And I think it’s difficult because like universities have a bit of a challenge. I don’t think. There’s not many courses unless you do like a separate masters or something that will feed coding into an undergraduate degree. And I think like part of that is because they’ve got so many things they need to teach, they need to get those like basics because you don’t want to send out like engineers who don’t know how to design a beam. or something, but no, how to code. Yeah, exactly. So it’s like a tricky thing for them to deal with.

And they’ve got loads of boxes they need to tick. But at the same time, it’s difficult to nurture those skills if you don’t do that at an early stage. How old is Modulus? That’s a good question, actually. I’m not entirely sure. I think it’s probably around five years or something. but it’s like really picking up speed in the last couple of years. Yeah, so just based on this, five years or so, let’s be safe and say eight years ago if there was no modules and 30% of the workforce here are coders and they are working towards improving the construction design and then for fabrication. That means that there is a need for this skill set and there will be need for this skill set in the future. to have people who are within the construction, but also are able to code and understand changes within the code, what implications they have on the software that they design in. And then so how does someone go about that Sam? How does someone who’s interested in computational design and they’re coming into construction, should we say? Perfect example is you. How would, what advice would you give them to succeed? Is it like late nights learning how to code on top of your day job? How does it look? Yeah, I don’t like that approach though. I don’t, I like my free time. I like getting out and like going for a bike and whatever else. So I don’t think that’s a healthy work life balance, but then again, so my approach and I’m not saying that like, I necessarily recommend this, but like, I was just trying to wedge it into every single project that I could, you know, and like at the beginning, like you’re learning and you’re learning something new. So. Although it has the benefits of being faster, like at the beginning, it’s not going to be faster to do something in a conversational design sort of way, because you’re learning and it’s going to be slower, but it will pay off in the long term. But so that was me.

I just like plugged it into everything and forced my way to do it on projects rather than doing it after school, so to speak. Yeah. How might someone do that? Is it like a matter of? Educating yourself slightly then applying it to a project you’re on and then just keep it going and making change Yeah, and like grasshoppers a great kind of gateway drug to coding because you can tinker with things and Put things together very quickly and then something doesn’t quite work the way you want it and you want to inject a tiny bit of code maybe and you Google it and you find someone else’s bit of code and you do a bit of copying and pasting and changing a few numbers and that’s kind of how it develops. That makes sense. And now you can ask ChatGPT. It’s an industry for geeky people. Exactly. Yeah, you just ask ChatGPT and it does it all for you anyway. We’re advanced. Okay, so should we move on to something off topic? Martin, you kick us off. Yeah, I’m always interested in what people read. What is the book that you could recommend for our listeners to… to read. Is there anything that you’re reading at this moment? In terms of like knowledge for construction industry kind of computational design, I think I do a lot of YouTube watching. I watch, especially since like the pandemic, like loads of universities have just chucked a load of material online because they’ve had to record it and then they’ve just gone, oh, let’s chuck it online. So there’s loads of great stuff that you can watch on YouTube. And that’s great because you can listen, watch while you’re… going for a walk or making a coffee or something. That’s what I do in the morning. It’s like a stick of video. Keep education. Yeah, continuous learning. And in your career as well, it’s being a coder, which is something I’ve noticed about people that do code for a profession, is they are lifelong learners. You don’t really have a choice. Like a lot of other professions is you just do your job and go home or do your job and you know it and you learn from experience. But you actually have to learn the stuff.

Yeah, it changes. That’s the thing, isn’t it? Because physics doesn’t change. So as an engineer, like of course you can continually learning through experience, but you don’t have to learn necessarily any major new concepts because the physics is the same. But with tech, like the tech changes. Sam, if I may ask, the chat GPT and all of these machine learning things, does it mean that there will be a collapse of renumeration for coders in the future? because of the fact that ChatGPT can code for you and you can just ask them to ask it to code something and it will do it for you. I think about the future, I don’t know how it works now because I have no idea, but feels coders are obviously well paid because that’s like a high tech skill. And if this can think can do it for us, for them, then does it collapse the salaries of coders at some point or not? Maybe for the basic stuff. I don’t know, it could probably build you a simple website pretty quickly. I’ve seen like Twitter and stuff. I built a website in five minutes. Yeah. Someone takes a picture of a piece of paper, don’t they? With scribbled website and it just gives them like the whole HTML. Yeah. Yeah. So yeah, for that, for sure. But I think anything that’s like properly creative and requires industry knowledge or kind of proper understanding of the challenges, I think more challenging for something like chat GPT to tackle. Absolutely. And you still have to have, you still have to know how things go together. Yeah, you want to modify it. And you want to, someone says, oh, this isn’t quite working for me. Can you change it? If you haven’t written it, it’s very difficult to do. Yeah. Okay, Sam, where can people find out more about you, Modulus and perhaps any other endeavors? check out our website, hondulus.com. Oh, for me, I don’t know, probably LinkedIn, Sam Gregson, I don’t know what my thing is.

I sometimes post stuff. That’s where I’ve, it’s usually your name. That’s where I found you. Oh, okay, yeah, yeah. It’s a good source. Yeah, yeah, so I don’t really tweet or anything. I am on Twitter, but mainly just following other people. But so, LinkedIn, I’ll post if I’m posting something, yeah. Great, okay thanks. And hopefully I’ll be like doing some conferences. Now I do conferences now and again. I quite like to do that, share knowledge, love kind of sharing things. Any ones in particular? So I’ve not got any lined up, but last year I did built. Have you heard of built? No. So what’s it? I can’t actually remember what it stands for, but it’s… Let me have a Google. Digital Built Environment Institute is the institute that put it on. So that was the first time I went there, I really enjoyed that. So lots of interesting stuff about BIM and C and like games engines and yeah. Is it in the UK or is it somewhere else? It’s everywhere. So there’s like a built Europe. which changes location within Europe.

And then there’s one in the States and then there’s one in Australia as well. But I really enjoyed that because there was lots of people there like willing to share their knowledge. And sometimes you go to a conference and they’re just there to show off the projects they’ve worked on. And that’s not really why you go to a conference. Or at least not- That’s the old mindset in construction. Right, yeah. There’s a new generation coming for a change in that. Cool. All right. We best wrap this up. Sam, thanks for coming on. We appreciate it. Thank you. Yeah. Thank you, Sam. Thank you, Sam. Enjoyed that. Thanks so much for tuning in to this episode of the Bricks and Bytes Podcast. If you are enjoying the show, please feel free to rate, subscribe, and leave a review wherever you listen to your podcasts. We really appreciate it, and we’ll catch you in the next episode.