Subscribe
Share
Share
Embed
FUTURE OF XYZ is a bi-weekly interview series that explores big questions about where we are as a world and where we’re going. Through candid conversations with international experts, visionary leaders and courageous changemakers- we provoke new thinking about what's coming down the pipeline on matters related to art & design, science & innovation, culture & creativity.
Future of XYZ is presented by iF Design, a respected member of the international design community and host of the prestigious iF DESIGN AWARD since 1953. The show is also a proud member of the SURROUND Podcast Network. For more information, visit ifdesign.com/XYZ.
00:00:04:00 - 00:00:27:08
Speaker 1
Hello and welcome to this week's episode of Future of XYZ. I'm a little overwhelmed by our guest today. Martha Tsigkari is the senior partner and head of the applied R&D group at the incredible Foster and Partners, the architectural firm based in London, all around the world, led by Lord Norman Foster.
00:00:27:09 - 00:00:31:00
Speaker 1
Martha, thanks so much for joining us on Future of XYZ today.
00:00:31:02 - 00:00:40:17
Speaker 2
Thank you for having me. I'm really excited to to be here and speak with you. I don't know why you're overwhelmed. I mean, it's going to be super easy and straightforward and fun.
00:00:40:19 - 00:01:00:24
Speaker 1
Okay, if you say so. I don't usually get so overwhelmed, but, you know, we met last year or earlier this year, I guess in April at the iF Design Trend conference 2025. You had contributed to the trend report, on this topic. But that talk was, mind blowing, which is how we ended up in this conversation.
00:01:01:00 - 00:01:34:24
Speaker 1
So that's maybe why I'm a little overwhelmed. Your background spans architecture, engineering, and computer science. You've been with Foster and Partners for nearly two decades, working across all these different projects. You're an associate professor at Bartlett at University College London, where I think you also got your masters science and architectural computation. Correct. And you've obviously juried and and taught and lectured all around the world, including at places like MIT, University of Pennsylvania, Chalmers, and all over in terms of computational design.
00:01:35:01 - 00:01:56:03
Speaker 1
And I know Greece is home, although London is where you live. So you bring that into kind of everything that you do. Martha, I always like to start us in our topics with, you know, kind of grounding in the roots of it, which is how do we define applied R&D in the context both of today's discussion and your expertise?
00:01:56:05 - 00:02:22:06
Speaker 2
That is actually a very, very interesting question. And that is one of the reasons I want to do this with you, because I thought the questions that you're posing are quite topical and they're scratching a little bit beyond the surface because, people keep talking about innovation, R&D, these are all wonderful things, right? But in many situations, particularly in the AC industry, you see a lot of these things becoming, pretty much blue sky research.
00:02:22:06 - 00:02:46:18
Speaker 2
And by blue sky, I mean the sort of things that you're looking for the future, may have nothing to do with your workload now, but they're like good to have. And so you can collaborate and do a variety of things that shows that you're thinking about the future. The entire point of applied R&D and the way that we've defined it and we build the team, was that we want research and innovation of the future that can be applied to the way we work today.
00:02:46:20 - 00:03:12:18
Speaker 2
So what technologies exist out there that can completely transform our profession, and how can we tap into that way before other people would, and not just technologies that exist, out there within our industry, but technologies that exist outside our industry. How can innovations in healthcare be part of like what we do? How can innovations in film or computer science can be part of what we do?
00:03:12:18 - 00:03:29:05
Speaker 2
And, change the the way that we do things. And so this is how I would define it, future looking innovation that can be used right now to change the way that we do things for the better.
00:03:29:07 - 00:03:55:15
Speaker 1
I love when you break something really complex down to something relatively simple as you just did. I mean, the truth, however, is that you're a specialist in a wide range of areas, including computational design, performance driven design, optimization, machine learning, interface and interaction, design to production, fast feedback, integration, all of these different things. How does this broad range of skill sets and tools get applied to architecture and design specifically?
00:03:55:15 - 00:03:57:14
Speaker 1
00:03:57:16 - 00:04:19:21
Speaker 2
Well, I can give you example for all of them. And so I can take you through a little history tour of like how we, we developed. Initially our bread and butter 20 years ago was computational design and starting with parametric and generative design. So how can we use new tools in terms of reassessing how quickly we can do things?
00:04:19:22 - 00:04:44:04
Speaker 2
A lot of these things were kind of obvious in the aviation industry back then, but they were not that obvious in the architectural industry. And so we started kind of creating our own algorithms or using new software to be able to make different permutations that we can analyze very quickly. And that's how we started. But as you go along, you realize that there are a lot of different things that you can be doing to optimize a process.
00:04:44:05 - 00:05:06:13
Speaker 2
You mentioned performance driven design. Performance driven design for us is this idea of being able, in real time, to analyze what you're doing to break these very long cycles of “I designed something, I need to export it, send it to like five experts who need to run their own analysis in different software, clean up the model,” by the time they come back with their reports and findings, two three weeks have gone along.
00:05:06:19 - 00:05:28:12
Speaker 2
At that time, we have done 50 different options, 150 different options. So how do you inform them in real time? How can you help designers create intuition in real time? And we thought, right, let's start building simulations that are related to sustainability criteria, structural criteria, visibility criteria that can help us very quickly to understand how well we can develop things.
00:05:28:14 - 00:05:55:06
Speaker 2
And so we started developing our own applications. And then we tapped into technologies like GPU computing. So doing things in the graphics card that at the time was developed for the game industry in order to have real time interaction in games. And we thought, well, if you take that from the games and you put it into our own simulations, you can start having applications that in real time can tell you how well something performs without having to export or even having to run an analysis that takes one day, two days.
00:05:55:06 - 00:06:24:02
Speaker 2
But it happens instantaneously, right? And from there, you start thinking, now that I've done that, and I know that I can design and analyze, design and analyze in real time, can I go the other way around, kind of go top down and say to the computer, this is the DNA of what I want to be doing. And based on that DNA and these objectives that I have, these different analytical engines, can you offer me the optimized options that can be provided through something like a genetic algorithm?
00:06:24:02 - 00:06:58:14
Speaker 2
But that's something that, you know, computer science was investigating all the time and comes from evolution, really how we develop. And so you start including optimization and okay, now you have to not do just one option but you have to do 200,000 options to deliver some optimized result. How can you do that? Well you know what? Rather than doing it one by one, can we use distributed computing that is used for visualization in the industry to start like, being able to not run a 100,000 options linearly, but to run pieces of them to all the hardware that we have in the office.
00:06:58:16 - 00:07:22:12
Speaker 2
And we do what used to take two weeks in two hours. And so this is a little bit of like an idea of how you take certain things and you apply them in architecture. And all these are true for things around immersive environments and augmented reality. How we took these ideas that have to do with how you immerse yourself in the environment, not just as a visualization tool, but as a design collaboration tool.
00:07:22:14 - 00:07:53:03
Speaker 2
How do you use the same ideas through AI, through business cases and how you can drive innovation, automation through that? How do you take digital twins from something like, an airplane, which is where they used to do it, in aviation they have the digital twin since the 1960s. And you're thinking, well, this idea of operationally being able to monitor a building and not only use the findings to optimize the operations of said building, but also bring it back in the design process and optimize how you do things to do them better.
00:07:53:05 - 00:08:18:13
Speaker 2
All these starts kind of coming together, and this is how R&D that exists out there in different spaces can be directly incorporated in what we do just by having a vision. And it's not a one person vision, right? This is a team vision. This is a company vision of yes, we want to be able to transform the way we work to the better with every possible means.
00:08:18:15 - 00:08:51:06
Speaker 1
You've just answered why it's an overwhelming conversation topic, so thanks for that. No, I think, you know, it was interesting, Martha, because there are two things I want to pull on here, and I'm not sure which order. When you presented at the trend conference, your topic was Disruptive Technologies in Architecture, exactly what you're talking about. And what you showcased were a lot of examples of these models in real time, whether that was light resourcing, whether that was sustainability, whether that was engineering for, you know, and structural stuff for, for buildings that needed to be retrofit.
00:08:51:06 - 00:09:09:24
Speaker 1
I mean, there was there were quite a number of applications that were pretty fascinating to me. And I think, if I'm not mistaken, the statistic you gave is that you could get to with a preliminary design, and then using the applied R&D teams tools, you guys could get to an 80% accuracy. So then the design could build on top of it.
00:09:10:01 - 00:09:20:02
Speaker 1
Was that accurate? Is that an accurate memory that I have? And maybe you could give an example of a of a project that you, you know, of of, of, of these kind of lessons that you imparted in that talk.
00:09:20:04 - 00:09:50:06
Speaker 2
This is curious in relation to analytical results. So we, what we strive for at Foster and Partners is performance driven design and evidence based design. So the the buildings that you see that we're creating, they're not done the way they're done because aesthetically it’s pleasing to us. They're done the way they're done because we want to provide the best possible user experience for that building in its context and its environment with the type of uses that it's going to have.
00:09:50:06 - 00:10:12:17
Speaker 2
Right. And to do this, it's not enough to just design something. You need to analyze and drive it based on what performances you're trying to hit. How do you want to make that space be unique? How do you want the experience to be unique? Now, in the past and still now in many offices, in order to achieve that goal through this design,
00:10:12:17 - 00:10:33:09
Speaker 2
it's a multidisciplinary process. It's not just an architect going there by themselves. We're not in a nine run board, right? We are in a in a reality where tens, if not hundreds of people are being part of delivering a fantastic project. And so how can you facilitate all these people to get insights on the design really, really quickly?
00:10:33:11 - 00:11:17:07
Speaker 2
And that's where some of this performative like that's where that comment went in terms of like how we can provide performative criteria and outputs very quickly. You can do that through, as I said before, GPU computing or distributed computing, actual results directly, really fast. You can do it through AI. So we have built a what we call sorry, not what we call, but they are called like surrogate models. And what surrogate models are they are also machine learning AI, let's say a machine learning models that can replace long analytical processes where you're not getting the actual analytical process, but you're getting a prediction on that analytical process, having trained your system to that
00:11:17:07 - 00:11:39:02
Speaker 2
analytical process. And in that sense, that's where my comment was like, at that point you can get 80, 90% accuracy. So the prediction is very hardly going to be 100% of what the analysis would be. But in early design stages, does it matter if you're special connectivity is 80, 90% accurate? It will point you to the right direction. If something is really wrong, you will see it.
00:11:39:04 - 00:11:53:20
Speaker 2
You will build the intuition to it and you will move on to a better solution. And that's what we're trying to achieve. We're a very small cog in a very big machine, which is the design engine at Foster and Partners and we just do our part. But again, that's a small part, right?
00:11:53:24 - 00:12:17:14
Speaker 1
It's a it's a pretty big part. I want to talk about the group, and the profiles and kind of like how, how exactly how it interfaces with the hundreds, if not thousands of architects. But I want to talk about something that you mentioned first, to understand it. You've mentioned a few times in your work, is known for incorporating deep neural networks and genetic algorithms in the design process.
00:12:17:16 - 00:12:22:09
Speaker 1
What what are these what does what does what does that even mean?
00:12:22:11 - 00:12:42:20
Speaker 2
So, genetic algorithms are algorithms that we're using for optimization processes. And as I explained before the work, a little bit like evolution, the way that the genetic algorithm works is as as explained before, you can have a bottom up process where you start building a design, you analyze it, you see how well it performs, and you change your tact to keep optimizing things.
00:12:42:20 - 00:13:02:24
Speaker 2
Or you change it and you see that it performs even worse. Then you create an intuition oh I need to go the other way around. A genetic algorithm says, give me the DNA, give me the the rules by which you want to design this thing and create the DNA of how this options need to be created. And then give me the objectives.
00:13:02:24 - 00:13:21:02
Speaker 2
Tell me it needs to excel in daylight. It needs to have this structural stability, needs to have that many sunlight hours in the yards. It need to keep that particular area. It needs to have optimal connectivity with the parks nearby. Okay. So these are the objectives. And then you allow the computer to do the heavy lifting for you.
00:13:21:02 - 00:13:44:14
Speaker 2
So what the GA will do is it will automatically take the design DNA that you created through a parametric model or a configurator. It will create a few options. In this case, usually every generation is a 100 options. It will analyze each one of the single options against all this criteria. Using in this case is our distributed computing capabilities and our fast, real time analysis.
00:13:44:16 - 00:14:12:18
Speaker 2
And it will sort all this options from best performing to worst performing in relation to all this criteria. And then it will get through an algorithm. The best performing, from the pool of the best performing ones, but with possibilities to get from the least performing ones, some individuals, let's say, to breed them together, their characteristics and create a new individual and replace all the individuals in the population with new ones that are bred through this process.
00:14:12:20 - 00:14:36:04
Speaker 2
And so you do this thousands of times, creating hundreds of thousands of options. And little by little, the solution converges to a multi-dimensional pseudo optimal space, each dimension being one of the objectives that you said that allows you then to visualize. Okay, what are the solutions that are trying to do better? I can solve this objectives even though this objectives may be conflicting.
00:14:36:09 - 00:15:03:22
Speaker 2
If you want, let's say to shade things, but also have daylight in the room. This is a conflicting objective. And and so it places all this position in this multi-dimensional parental front where you can start interrogating what are these characteristics and how you can move from something that already works to the design aspiration that you have. So this is this process described is a genetic algorithm is the starting point of our design process
00:15:03:24 - 00:15:54:00
Speaker 2
in some cases. Then there, the deep neural networks are machine learning networks. And they kind of resurfaced in 2018, 2019. And it's really a way of being able to create algorithms. In that case, you're not optimizing, but you're training systems to do the right thing based on a particular output. So we use deep neural networks in 2019, in one of the investigations that we've done where we were trying to identify, for example, if you have a passively actuated material like a laminate and it's heated and deforms, how can you control that particular deformation?
00:15:54:04 - 00:16:19:12
Speaker 2
To give you an example, imagine that this material scale up in 20 years time, ten years time, and you can have organic materials on your facade. Then when heated, can transform to a shading device, right? In that case, that transformation to a shading device is not straightforward because depending on the heat, it really depends on the layering of the laminate that will provide the particular transformation.
00:16:19:14 - 00:16:42:12
Speaker 2
And in order to understand what that is, you need to run thousands of different laminates to analyze them in a nonlinear, analytical way, and then start kind of predicting what, understanding what the layering should be. So a deep neural network can reverse engineer the process. It can say, okay, let's run some of this, a few thousands of them.
00:16:42:12 - 00:17:03:03
Speaker 2
Let's understand the different laminates, what’s the formation they produce. And then let's reverse engineer that and take the output that we want, the formation that we want for the neural network to tell us what the predicted laminate clearing should be. So these are two different ways of dealing with with this these things, this two different examples.
00:17:03:03 - 00:17:34:08
Speaker 1
It's totally insane. I think one of the things that's really interesting, Foster and Partners, obviously, Lord Norman Foster was given the second ever iF Design Lifetime Achievement award at our award ceremony the evening before you spoke at the conference. And largely, I mean, the year before, it had been Dieter Rams. I mean, the fact that, you know, Foster and Partners is doing the level of innovative work that it does, and that's where you guys are, whether it's materiality, whether it's light bearing, whether it's, you know, connectivity to the local community.
00:17:34:08 - 00:18:06:24
Speaker 1
I mean, these are all things that are really important, as you said, the parameters. But some of these are like the Sheikh Zayed Museum in Abu Dhabi or the iconic Lusail Stadium that was, you know, the 2022 FIFA World Cup venue. But, also the Mexico City airport, I think is, from what I've seen, forget super yachts and all the other things that Foster and Partners touches and the Reichstag in, you know, in Berlin and all these kinds of incredible, beautiful design feats that also have unbelievable engineering aspects to them and materiality considerations.
00:18:07:04 - 00:18:30:04
Speaker 1
But the Mexico City airport, to me in your presentation, Martha, was one of those where what you just described, you know, in terms of how what is really possible came to life. How did your team, you know, first of all, when when was your team, the applied R&D team at Foster and Partners formulated what's kind of the profile?
00:18:30:04 - 00:18:40:21
Speaker 1
And then I guess I'd use the Mexico City airport as an example of like, what is it that you guys then did that made that realizable, executable?
00:18:40:23 - 00:19:07:04
Speaker 2
Oh, well, I think it's, almost 15 years ago that we have now been created the Mexico City, the what could have been the new Mexico City airport is a sad story for me because it's the one that got away. There are very few, I mean, you learn to deal with disappointment in, in a profession like architecture because you do a lot of things and some of them never end up being created.
00:19:07:06 - 00:19:42:03
Speaker 2
With Mexico, it was actually a big build. And then through different political decisions and what have you, it stopped, which was really sad. But, this is still one of the best examples that I can give of how technology has expanded the art of the possible in what we can do in architecture. And, we were having discussions within a lot of Foster about that because this idea that he was having about the reinvention of the airport and how the team wanted to approach it, were not ideas that were new.
00:19:42:03 - 00:20:07:13
Speaker 2
I mean, we did first with Stansted, but then if you start thinking how technology made some things possible, this idea that you want a space frame that doesn't distinguish between wall, column, or ceiling, that works very hard environmentally, is totally optimized but has to be super, super lightweight because we're building a 1.6km space frame that sits in what used to be a swamp.
00:20:07:20 - 00:20:11:23
Speaker 2
So the conditions of the soil are not the best. You have a volcano nearby.
00:20:12:23 - 00:20:53:13
Speaker 2
For extra fun, and you have a highly seismic zone. So you have to to have a super, super lightweight space frame that does all these things. And we managed to do that. I think the the weight of the space frame altogether was less than 5% for the weight of the entire building or something like that. And that was only possible because the entire space frame, we worked very, very closely with the design team and the engineering team to make sure that it was created programmatically from an optimal topological configuration of each one of the hundreds of thousands of nodes of the space frame that was then pushed up in the right height and
00:20:53:13 - 00:21:04:18
Speaker 2
dynamically relaxed to create the optimal shape and compression, which is exactly what Gaudi did for Sagrada Familia, right? With, with his hanging change. But we did it in a digital format.
00:21:04:18 - 00:21:06:15
Speaker 1
In Barcelona you're referring to.
00:21:06:17 - 00:21:29:19
Speaker 2
Yes, exactly. So we like going through that and then going through developing a level of standardization and detail for financial, economical, sustainable reasons. The producers at the end of the day, a B model that is so rich that it becomes, a sort of painting in the Royal Academy here in London for people to see in terms of like output.
00:21:29:19 - 00:21:46:13
Speaker 2
I think that this is a really remarkable feat and we wouldn't be able to put all this ideas to action had we not had in our hands, the technological advancements that we work through in order to be able to deliver something like that?
00:21:46:15 - 00:22:14:19
Speaker 1
I can't believe how the time quickly goes. So I have I just have two other questions before we get to our last one. I think one of them is, you know, it's fair to say and based on this conversation, it shows as well everything that you do is diving deep into kind of the revolutionary impact of, let's call it technology or computational design. As we look, you know, kind of in the near term, what most excites you that's happening?
00:22:14:20 - 00:22:21:10
Speaker 1
And, and, and is there anything kind of I don't know that like, what's the impact of that going to be?
00:22:21:12 - 00:22:53:22
Speaker 2
I think if we’re talking about technologies that are completely going to transform the way we work, AI should be at the top of of that list. I think, the changes that machine learning and AI will bring in the industry are extremely transformational, and they're going to reinvent a lot of the things that we do, not not necessarily the designs that we do, but how we automate pieces of the design in order to be able to focus much more on the creative process.
00:22:53:24 - 00:23:18:24
Speaker 2
Myself and my colleague Sherif Tarabishy, who's leading all the AI initiatives within the team, have been working since 2018, ‘19 on ideas like that where pixel picks, that's an algorithm, was all the rage. And like, people didn't really realize that we're going to have this revolution with diffusion models like, Major Nell Dali or with, large language models like, chat GPT and the likes.
00:23:19:01 - 00:23:46:16
Speaker 2
And we were trying since then to find ways by which, which can we can incorporate AI within different aspects of our process, from design aspects to financial aspects even, in a way that is secure and reassures that our IP remains our own, but in the same time automates parts of the process and makes them much faster while still retaining and augmented our design capabilities.
00:23:46:18 - 00:24:15:01
Speaker 2
And so I think this is going to be revolutionary in ways that people don't even imagine, because they're not only just scratching the surface of what is possible. And I think the the power of automation, that generative AI is going to bring in our processes and what we will be able to do with far less is going to be something that completely makes us rethink how we do architecture.
00:24:15:03 - 00:24:17:14
Speaker 2
00:24:17:16 - 00:24:37:22
Speaker 1
I love that, and I think that's a very forward looking vision and it seems doesn't matter what the topic I'm discussing on Future of XYZ these days is the hope and promise of AI is is very, very profound in the design profession, which is interesting. The thing that always concerns me a little bit, and I wonder how you guys are dealing with it.
00:24:37:22 - 00:25:00:19
Speaker 1
Before we get to our final question, and sorry for the for the bone in this, but of course the sustainability aspect of of our use of computational design is, is quite negative at the moment. We haven't figured out how to solve for the, environmental impacts. Do you guys at Foster and Partners have any kind of, or you as a leader in this space globally,
00:25:00:21 - 00:25:08:03
Speaker 1
have any specific thoughts on and how we mitigate the environmental externalities? Let's call them.
00:25:08:05 - 00:25:31:04
Speaker 2
I think it's it's very interesting because we I get this question quite a lot, especially when we talk about AI. And it's very important to say that a lot of the negativity around like all this models are being trained and all the energy that is being used, etc. we as an architectural firm or any other architectural firm who's using these models, are not having that footprint, right.
00:25:31:06 - 00:25:49:09
Speaker 2
These models have already been trained. How big companies are training these models, and I'm not going to name any names, but we all know who they are. How they do it in a more sustainable way is a discussion that's been had in conference for the past five years at least. There are different ways that have to do with the hardware, have to do with a location.
00:25:49:11 - 00:26:10:05
Speaker 2
I personally have a vision that we can do a lot of that training in orbit, potentially to mitigate a lot of the issues that we're currently having. Right. There are a lot of ways that this discussion can be shifted, but I don't think you can pinpoint by you, I mean, the, the collective you, you can pinpoint the sustainable footprint on that in every given company.
00:26:10:05 - 00:26:38:02
Speaker 2
Because, yes, it takes a lot to train one of these huge foundational models that take a year to train. It actually, like takes very, very little to use some of this thing for the things that we're doing in the in the practice and the office. Long term, I think how we approach sustainability, it has been very, very much engraved in the DNA of the company, and we would not do something that would grossly jeopardize that.
00:26:38:02 - 00:26:57:11
Speaker 2
But we also need to be conscious of the fact that albeit you might be the end users or something, the way that that something is created is not in your hands always. And all you can do is try to drive the conversation how things can change. But you don't always have the last word in that.
00:26:57:13 - 00:27:18:20
Speaker 1
That's fair. I, I'm I'm not going to push on some of that. And I totally appreciate, that there's responsibility that's distributed, and that we need to use the tools we have to do other things. The last question always, Martha, is always the same, which is, in conclusion, what is your greatest hope for the future of applied R&D in 25 years?
00:27:18:20 - 00:27:22:21
Speaker 1
So we're looking at 2050.
00:27:22:23 - 00:27:42:07
Speaker 2
For, well what is my greatest hope? Well, above all this, that's a hard one. I think my greatest hope that's that we're still going to be the dreamers of the impossible, right? And that you're still going to be pushing what, can be done with new things that exist that are out there that we don't even know them right now?
00:27:42:09 - 00:28:08:17
Speaker 2
I think that there's always the hope that you always keep using whatever you're better at as a means of exploration and of invention and I do hope that applied R&D as a whole allows people constantly to push themselves outside their comfort zone and rethink everything that we have been taking for granted, because that's the only way that progress is being made.
00:28:08:17 - 00:28:25:23
Speaker 2
So completely questioning the things that we know we have done successfully and thinking, are there better ways out there? And I think that is what, applied innovation, research and development are doing for this industry and every industry.
00:28:26:00 - 00:28:47:08
Speaker 1
I love that expression “dreamers of the impossible,” that's, that's that's what I'm going to hold on to. I think there's, and lots and lots and lots of applications these days. We need we need design to be dreamers of the impossible. Martha Tsigkari, senior partner and head of applied R&D group at Foster and Partners, thank you so much for joining us on Future of XYZ today.
00:28:47:08 - 00:28:51:00
Speaker 1
It was very special.
00:28:51:02 - 00:28:57:12
Speaker 2
Thank you, Lisa, I really enjoyed. Thank you so very much for giving me this platform to speak about the lovely work that we do.
00:28:57:14 - 00:29:18:04
Speaker 1
And for everyone watching and listening, you can, of course, just Google Foster and Partners to find out some of the amazing work that's happening. You can also download the iF Design Trend Report 2025 to find lots of trends to, six mega trends with two sub trends under each and one of which, Martha, has a deep dive on some of this work.
00:29:18:06 - 00:29:34:20
Speaker 1
You can do that at ifdesign.com who is our presenting sponsor. And, otherwise, thank you for watching. Thank you for listening. Thank you for subscribing. Thank you for leaving a five star review on your favorite podcast platform, and we will see you again in two weeks time. Martha, thanks again. Speak soon.
00:29:34:22 - 00:29:36:03
Speaker 2
Thank you. Bye.