The Architecture of Spontaneous Order | Patrik Schumacher

The kind of freedom we're craving for and asking for and the degrees of economic freedoms we need now, we never needed them so badly. The role of the architect is not to impose a formal order pre-conceived but to congenially articulate this emergent network of relations. We have the computational revolution, the flexible specialization, network society, and therefore I think to express that we should also express the frontier 21st century everything. And the 21st century everything we need degrees of freedom which would never have existed.

Yeah, that's absolutely right. So, first of all, how I became a libertarian through the 2008 crisis. And a lot of people shifted left, became anti-capitalist, but I had been there from the mid '80s into the mid '90s and moved out of that. So, I wasn't going to be tempted back into left-wing ideology. So, when I discovered the Austrian School of Economics because their explanation of the crisis made a lot of sense to me. And people like Peter Schiff predicting it and somebody like Tom Woods explaining it. And then diving rapidly into picking up the right readings and writings, and I discovered the Mises Institute and the works of Rothbard, but I've already been familiar to some extent with Hayek's works. So, that's where I started. And it applies beautifully to the built environment, because in the development process, urban development process, we don't have much market process left. It's very heavily controlled and regulated and politicized. And to the detriment of development, basically leads to some kind of paralysis, and in the end the housing crisis, for instance, we have affordability crisis. It can be traced back to that. And it's all rent-seeking, rather than proper competition, entrepreneurship, discovery process, none of this exists in the built environment in the advanced countries, for instance, Britain. Now, so I was very intrigued when I became a full-on libertarian. Also understood that it's very important in the urban development process to liberalize.

I just heard about Liberland being founded. I was intrigued and more or less immediately in the first year of founding, I flew to Prague and visited Vit Jedlička, the president and founder of Liberland, and we had a whole afternoon on his roof terrace debating, discussing. And so, that's where I connected and then I went to also visit Liberland. I mean, you couldn't go there, but there was a Liberland conference close by in Serbia. And Eric Brimman picked me up, was there at the time consulting Vit and trying to help Vit negotiate deals. And so I met with Eric and we also learned about Prospera at the time and through that also came in touch with Titus. And also he had just published a book around that time, free private cities, which I was eagerly gobbling up. So then I got involved with Eric. We were in parallel working with Vit. We helped him to organize some architectural competitions and then we developed the metaverse for Liberland. I went to several of his conferences and I also connected up with Eric and Titus. Visited Prospera and we did studies, actually development studies for Eric. And then, also, more detailed housing proposals with a very new concept of market exploration through co-creation, which is something you couldn't do in Europe because of the planning restrictions. So that's where I got into it and I was at the time also speaking at many libertarian conferences, Students for Liberty in various places in the US, Liberty Con and so on. And met with people and became part of a network.

So I coined the phrase parametricism, recognizing that there's something very profound and fundamental, I called it a new epochal style for the 21st century. And that was my role and also kind of canonizing to some extent what has been achieved over a number of years. By that time it was nearly 15 years in some kind of wave, a whole generation of architects investing in these new computational design concepts, but also the physiognomy, morphology of these compositions and urban plans became kind of radically different and new, much more adaptive, much more organic through the techniques of integration, adaptation and this coincided with the idea of the so-called urban renaissance where the period in the mid-20th century, there was this kind of suburbanization, greenfield factories, sleeping cities, not much communication required in the city because everything was command and control planned through big corporates or the state. And you have this kind of flow back into the city with the beginnings of the knowledge economy through the computational telecommunication revolution. And so that in architecture meant more dynamism, more complexity, layering new onto old and that developed into some kind of new transition, what I call transitional styles. Postmodernism came out of that, so-called deconstructivism, and I was involved with that with Zaha Hadid, who was one of the major forces of so-called deconstructivism, and out of that developed something which I later, in retrospect, called parametricism, which is quite distinct, but I think much more long-lasting. So postmodernism had 10-15 years, deconstructivism 10 years, and then parametricism we're now in literally 30 years in.

But the 2008 crisis took the steam out because what was great in the period up to that there was much more dynamism development, there was a re-marketization of a lot of development processes and you had phenomena like Dubai and in Europe a lot of booming cities and that energized the whole field and it gave younger voices, new voices work to do. And the 2008 was a general moment of retrenchment, where you had the anti-capitalist flourish, where there wasn't work, where the schools of architecture also became kind of more politicized and anti-capitalist. And so that in a way took the steam out of this movement parametricism, unfortunately, because throughout the '90s and up to 2008, 2009, 2010 it was really the field seemed to be converging on this. And I mean we kept going and there's a number of other protagonists, a small handful of protagonists kept going and we kept rolling out and maturing and developing, but it didn't deliver that kind of takeover of the field similar to what the modernists who started in the '90s and '20s and into the '30s, then there was a disruption of the Second World War, but then post-war was rolling out throughout the '40s, '50s, '60s and transformed the physiognomy of the planet all over. And so that didn't happen with parametricism. So we got this kind of strange retrenchment going into retro styles, going into quiet inconspicuous and the field, the discipline itself started to talk more about social justice and sustainability and criticizing growth. There was some kind of woke takeover of the field. Which hampered the kind of optimistic, progressive in the positive sense, technology fueled advancement of urban life.

Yeah, it's a version. I mean, postmodernism injected some kind of degree of complexity and there was this idea of collaging motives and integrating new over old. And it used traditional motives. A kind of pop collaging with traditional architectural motives, but you get something which is more lively and rich and varied than the modernism which is this paradigm of endless repetition. Yeah. And so, postmodernism started to be more playful and versatile. But with historical motives. Then deconstructivism picked up the kind of versatility and complexity build up, but no longer with historical motives, but with more abstract geometric elements. And so, that was I thought an advance. And they had additional compositional opportunities which postmodernism didn't have. And out of that, though, this couldn't wasn't in my view sustainable because it became very quickly too complex and illegible. These compositions they became kind of visual chaos nearly. So, parametricism comes out of this ordering that chaos by having the intricacy, versatility, adaptiveness of what we try to aim for in deconstructivism, but with the new tools, computational tools, we could do something which was more coherent and more legible.

No, okay, I can resolve that. First of all, on a more general philosophical level. We looked at process of self-organization, spontaneous order, complexity theory, chaos theory, the way a complex formation of eddies and flows happens when water flows across some rocks. These are lawful, but they're quite complex and rich and adaptive. And they're not invented. You don't impose. You allow the various parameters and influences to come together and have a kind of self-organization. And that's something which Hayek could also use that phrase because he was interested in these processes of self-organization, complexity theory, like cybernetics. So I think there's the connection right there. Now, when it comes to the spontaneous order of the city. What I'm arguing is this. We need to allow entrepreneurial freedom. We need land use should not be prescribed by any planning regime. I think that the people come together into the city by seeking out very particular connections and the multitude of connections. And they have to find each other. There needs to be a degree of freedom to associate and self-organize what I call the co-location synergies. That's what you need maximal free entrepreneurial freedom. And then the role of the architect is not to impose a formal order preconceived but to congenially articulate this emerging network of relations by making these underlying connections, affiliation, what belongs together more or less legible by articulating that, organizing that spatially and articulating this. So, when you come into a site and you say I want it to be here because I want it to be close to some of these neighbors. And so you establish then not you come in with a preconceived form. You say you have to be adaptive because I want to make my entrances into my building line up with the entrances of the neighbors I want to connect to. I also want to formally affiliate to the buildings perhaps they have. There's a certain geometry, a certain path hitting my building. I allow this to continue in that direction so that you see the building forms out of these relations. And parametricism allows for that. And that's the spirit. So, it is some kind of spontaneous order, but it needs to be articulated. The profession as has been was not able to do that; it had a very rigid vocabulary and insisted on proportion, symmetry, a certain orthogonality, but often sites are not orthogonal. So, you can exploit with a pliable amoeba-like kind of approach, you can fit yourself much better and also if they have different levels coming to your building, you can mediate. So, these moves mean that we can trace and make visible the spontaneous order which emerges and we don't impose. Our view is that we have to trace, follow, articulate the affiliations and connections because that's why people come into the city – to have these connections, to be close to things they care about, but also to work together with things they choose to be next to. And sometimes if everything is kind of visually chaotic, you don't understand what functions together. But if you formally affiliate and pick up some of the materialities, some of the geometry, some of the alignments and directions, you can intuitively read this environment better. And that's our idea.

And also to give, don't stuff everything which needs to be accommodated into a preconceived box. Internally preconceived. No, you allow the requirements to push the envelope, to structure the internal relations without preconceived geometries.

Yeah, that's right. There's always frameworks, but not only in parametricism. The point of parametricism is that the frameworks are more versatile, playable, they have more input parameters, but also I can choose to have many potential frameworks I could bring to bear. Compared to older styles, there's only one or two or three with inverted commas frameworks to start with, and they have less pliability. So, it's always comparative. You can't get away from what I call the formal a priori, the kind of framing set of categories even when you describe something. You come with a language. The question is your language more versatile, expressive, subtle, adaptive or is it more crude, simplistic, etc. So, my argument is parametricism has a more advanced, meaning sophisticated, pliable, expressive language, also more tools to pragmatically adapt, not only a question of expression, it is about accommodation to pragmatic conditions, which are more complex.

I mean, the previous in the old world, you know, life in various institutions was more static. 9:00 to 5:00, you come and sit always in your same cell. At lunchtime, you go to the canteen and after lunch, you go back. And then everybody streams out and you don't need to have inter-awareness, inter-visibility, continuous re-calibrating. You have maybe a few scheduled meetings and that's much more dynamic now, particularly knowledge economy. So, every day you need to make sure you take note of what all the other moving parts are you plug into. And then you kind of roaming around, you often have not a fixed desk. You're roaming around in different settings, detail-oriented space, for instance, and you have meetings with internal, external conditions. You also right after work and in between work, you have lunch meetings, which are business meetings, maybe outside networking. After hours, you go to the pub or go to the exhibition opening or conference. So, it is much more dynamic and interactive and therefore the settings you need to emphasize different things. You need to emphasize that you accommodate all these flows and relations, that people also have more eligibility of all the other things going on, which might be relevant to them. That's why we have opening up the section. We have atria, inter-visibility, inter-awareness and we have more kind of fluid in different settings. You can quickly shift from one setting to another. It's not closed rooms. Once you're sitting in the closed rooms and then you come open the door, there's an empty corridor and you have to know where you're going. But if you're in an open bureau landschaft over many levels, you get a sense what's going, who's coming, who's going. And what might be relevant to you. And if you go to a Google campus, that's what it's like. Actually, at Google, people have 20% of their time free used to have to freely associate. They set up teams, set up projects. Because you can no longer command and control from above. It's moving too fast. Knowledge is dispersed. Work is self-directed. So, we are creating the environment for that. And it's not corporate headquarters and factories on the green field, which just follow a blueprint, so it's continuously rewriting the blueprint. And that's what we want to accommodate. And express that and facilitate that. And of course, you always have some kind of systematic framework, but there are many competing ones and they update quickly. So there's not one frame which we stuff everything into. It's like the endless species of nature, you know? None of them is arbitrary. They start arbitrarily, but then they adapt, they evolve. They exploit a niche. And things in nature they collocate and synergize and feed off each other. And that's the way I see the world of the spontaneous order of the knowledge economy.

Yeah, you definitely have to let the market come in, but also you need to seed it with something. Something which already has the rough general features. I wouldn't start with a grid of 100 equal blocks. I would start with something which has a little differentiation or range, economies of scope, so you can have small, medium, large enterprises, but you also allow them to connect. So you can fuse blocks to allow for bridges, etc. Yes, you can fully invent on the drawing board. But when you come to seeding a city and making early first sketches, you need to invite people to something which they can imagine. But then there's something already you know there will be updates and transformations and flexibilities. You have to be very agile on that. And we can do that because we don't destroy the figure by making changes. If you have a very rigid neoclassical, symmetrical, axial plan, if you adapt something, it will be out of whack and shows as a mistake. That's not the way we would design. So it's inherently flexible. We thought this through also with respect for instance something like Liberland or Prospera. So we're saying we can start in parallel with some kind of curated center. It remains adaptable and we have other districts where we try to say, 'Come in and we let the group of owners who started to invest, they can arrange amongst themselves how to curate some rules because you might need some rules to contain negative externalities. And then but then you also should be proposing a free for all where you come with a risk. I'm placing something here. I'm unrestrained, but I therefore cannot control what happens next to me. But I would trust that somebody next to me will might be here because they want to have something to do with us. So we will be synergistic. But you can't predict. So there's a bit more risk, but also more opportunity in the free-for-all. So I was opting for a kind of mixture of regimes in parallel. A central curated where managers and owners, let's say the free cities investors make propositions, and it will be more predictable that you come in here and we will roughly in this environment. But then of course you also rely on the fact that it is being filled out by the planned kind of functions. And you wanted to say there will be a conference here. You have to actually find somebody who's hosting conferences. Unless the city developer wants it himself to become a conferencing company or exhibition company. So there are pros and cons. And the way we looked at it, we're saying we start on three or four fronts simultaneously, and then let's see what catches on. And then remain agile to adapt as we go along. So yeah, it's definitely something which you can't do overnight.

What is very important though is that you start with a critical mass. Everybody waits for everybody else to be there to have the network starting. So the network effects like with any company which relies on network effects, a city is relying on it. The difficulty is to bootstrap that out of nothing. Particularly in a remote location. So that's why I think it's very important to have things like a metaverse version or to build a community ahead of settling. So there's relations, there's expectations, there's plans which have been shared and recycled. And we developed this tool of a co-creation virtually, a digital twin evolving, which you can already use for communications, for conferences, like Liberland. The Liberland metaverse, for instance, played that role and wanted to play that role. And I think that's important because the difficulty is, as I said, it will always start relatively small, but it has to be rapidly grown to a certain extent that you're not isolated. The good aspect is that you also can be plugged into telecommunication network, but you also want to have things happening at site. And even if it's just to share certain facilities, but also to have collaborators. So, that's why I also thought it's important to try in the curation to target certain interest industries first, where you feel and you know, when you open a new building, it's the same. You can say, 'I'm calling this a media center because I wanted to not randomly have any tenants which have nothing to do with each other. I want media companies here.' So, you have to play a little bit. I want to kind of reserve these for media companies. Try to find and bring some media companies. That will then bring other media companies. And then you have something you don't need to be very large to have critical mass of co-location and synergy. That's I think a strategy which I would emphasize. You need to niche or have some kind of in the world division of labor, some kind of specialization so that you don't need to have 10 different seats which each of them lacks critical mass. And relevance of these to each other is insufficiently strong. So, that's the tough task. And sometimes you have to hold a little bit longer rather than giving up too quickly and it becomes a kind of random sampling of anybody who somehow feels to... Of course, so that's the kind of thing and I think Prospera's been doing that with respect to for instance crypto and certain things which make sense there like innovative health tourism and health innovations. So I think that's important and that's an entrepreneurial task and we are very alert to that and we don't have the support that as architects we're not writing these briefs and programs. But I would highly recommend and your point is very true that it will have to emerge and if for instance the health tourism takes on much more than the crypto which might have different requirements then you need to also be able to pivot or recalibrate the priorities and let this emerge because we know blueprints into long future are vulnerable particularly in such dynamic times.

Well, what I think a lot of people appreciate. First of all, yes, because the look is not the immediate intent. The look is an outcome of a rational and adaptive process. And it's interesting, you can recognize it, but it's at the same time more versatile than all the other styles put together. It's a bit like you always will recognize whether a form is an organic form or a natural form, rather than a mechanical form. Even though there's so many different natural emerged forms, you recognize this isn't an invented mechanical piece of equipment. So that's sometimes the puzzle where people think, 'Oh, it's recognizable. That's why it's all the same.' On a very abstract level, it is all the same. It remains recognizable, but it's like the infinite forms of nature. So it's much more versatile and rich in variation, although recognizable has one label. And it's not actually recognizable as Zaha Hadid. Not to the extent, I would say, then I wouldn't be proud of that. Recognizable as Parametricism, I'm proud of that. And there are many others who are coming in since we are such a dominant mothership of this movement, we'll consistently work that. It is often conflated and associated, but there are many other authors. So I would say yes, there is a language in that broadest sense. And also it means that it gives a certain identity to a place.

So what people dislike today is that there seems to be a lack of identity in different places. They all start to look the same, but if you look closely, it's not the same, but you can't get the gist or essence of the place. I call that garbage spill urbanization. If you spill out various cans of trash, they all have quite different ingredients. But when as a field of trash, they all look the same. Because it is amorphous. It has no form, shape, order, logic you can hook onto. Even though the one has a Coke can and a milk carton and a crumbled paper, and this one has plastic pieces and smashed glass bottles, in the end we respond to them as all equally white noise. Disorder. So that is where we have to reflect what we think is sameness, what is difference. And we crave for identity and with parametrics, it's very clear that we can build up not preconceived. Every town would evolve in a path dependent way because it starts with certain condition and there's certain material, certain climatic conditions. There's something which gives it a particular identity. And if you work in parametrics, you intend to affiliate as things come together because they wanted to be together. And they have a certain look and feel and sets of relations. This builds up and this might also include ingredients which were there. Whether it was a valley, a river, an island, whether it's tropic climate or northern Nordic climate. All of this will imprint. So you end up with recognizable places. They're similar in a way to the vernacular where you had to through the constraints and parameters and with less blueprinting but evolving village forms, town forms. Often they're concentric, not because somebody invented a concentric plan, but because you start with some kind of fortification and then the first ring of houses around that, second ring, third ring. And then the whole thing is locked onto a river on one side so it has this asymmetry and orientation to one side. So these things evolve. And in a way that's Parametricism does it too. So the vernacular settlement structures, they have not been planned, they accrue, but they have character and logic because there are certain regular parameters which imprint themselves and generate unity and the right variety relative to conditions. And that's the paradigm of course in today's world, we're operating at a different scale with more versatility but the principle of self-organization, spontaneous order, it is still there.

My response to this is: they're quite right. If you look at the different environments we pervade, there's something a sense of place, a sense of coherence, versus that jumble of incongruent buildings which have no rhyme, no reason, no relation. Because in the current world, it is actually failing of our discipline, but as well as the kind of randomness of a land use plan where you have to only build what is prescribed here. And therefore no immediate synergies, but also the planners prescribe too much, so you can't articulate and architects try to bring their own relatively self-indulgent ideas and this one is quite fragmented. I would also choose the historic city. As I said earlier, there's the vernacular, but also then the later centuries where architecture was involved, but mostly involved only in the primary buildings. And the bulk of the building was still traditional home building, vernacular. These are more lovely, more places with more identity. And also the convergence back into the historic centers, to the cities, this kind of more dense life is also better adaptive now. And also this idea of pedestrianized cities where you have to connect up and be in many places quickly than the green field and mechanically mass-produced sleeping cities. So, I would the intuition is right. And if you had to choose with what's available either this kind of concrete mega building or the garbage spill cluster, I also would choose the historic city center, and that's where postmodernism started. But, it is not, it also wouldn't be if you now compare what we could deliver a kind of spontaneous order which is more adaptive to the actual needs because with the historic center you have preconceived elements. They're nicely ordered, they create a nice ensemble and city. But, they're still the buildings are too small. The buildings didn't have enough variety between buildings. And we can deliver variety between buildings without collapsing into garbage. We can make larger variety, but have gradual transformation from one type to another or make them similar in interesting ways. So, basically the attraction is going to the best we've got. But, we have something much better we can deliver. That's my response.

Oh, yes. I mean, look, when I would make a new place, we also learn from some of those features of the historic city. For instance, the focus on the void spaces, on the plazas and places and streets as spaces rather than only at the buildings and the in-between has no shape. No, we also want to shape the urban space, and not only the buildings as object. And sometimes the buildings as object have to retreat into a framing surface to focus on the space. There's lessons we can learn. Also the variety and the different types of cities, more monumental and more intricate with small alleys and small surprising plazas, and we can learn from that. But again, as I said, maybe we have to do it on a different scale. We have to do it more variety. So, there's no lack of learning and that was the whole period of postmodernism was about. But there are things like lodges where you have between the street surface and the outdoor and the indoor, you have a covered buffer zone for instance. Features like this we can pick up on and when we did the city for practice in Italy and Sicily, you also learn some of the features of the historical buildings that are better adapted to climatic conditions. In the modernist period, we could just do a glass box everywhere and inject mechanical power to heat or cool, whereas in the traditional architecture, you had to try to develop comfortable conditions without having mechanical burning through huge energy because you couldn't do that. So, you make things like wind towers to ventilate out the heat. You have shading, covered places, make the streets very narrow because they're self-shading. And all the things we in the modernist period a lot of these subtleties, adaptive subtleties were left behind, not required. Now, with this sustainability consciousness they're coming back, but also they give character and identity. And therefore, in different climates, you have always these kind of roofs, roof overhangs. And often you don't even have enclosure. You're just covered and let the wind breathe through freely. And so we can work with these conditions, but now on a different scale and combine them with different requirements. So it is understood that there's this draw to these all these patterns and we can integrate them. So I would have this research paradigm learning from the vernaculars. Because we can learn how there has been hundreds and thousands of years of incremental adaptation to climatic conditions. And it's a bit like biomimetic research where engineering solutions go out into nature because organisms have developed intricate solutions to problems which we can learn from, and similarly we're learning from the vernaculars and the historical architecture because we have to relearn it because these were intuitively discovered through trial and error and they're not written scientific analyses.

But of course we can't take it over one to one because as I said we have slightly different requirements now, particularly on this level of scale. But I feel and what I have the responses we get with our work is that the kind of love and sense of comfort and energy people get when they walk into a historical city, we can solicit a similar response with our buildings and cities.

Yeah, they were solutions to problems from a natural scenario as well. The human mind is a natural scenario, isn't it? It's like, okay, we want a slanting roof, we want an overhang, right? We need to prop it up with something. We work with slanted rooms. There's drainage. There's also vaulting. A lot of vaulting. And the interesting thing is that when we do vaulting, in the old days, it was only possible if you have a full symmetrical condition. Because if you vault, you have the forces you rely on an equilibrium. If it is more of a freer form, an irregular form, then traditionally they wouldn't know how to equilibrate this. And you have differential settlements, you wouldn't know how to dimension the foundation, the whole thing would crack up. Now, we can make use of vaults, and we can make them more adaptive to site conditions, to the requirements of spaces and rooms. We don't need repetition and symmetry to the same extent. But we like vaults. Vaults are very efficient. Because everything is in compression, and material is much stronger in compression. And so we can that's contemporary research. So parametricism moved on to what I call tectonism, which is still parametricism, but it is absorbing many more, engineering computational logics and science, and more subtly integrate that as well.

The logics because they bring out typical forms which have a coherency. And also they relate back in a lovely way to some historical models, for instance a vault. And when you place the stones, there's a certain pattern, or if you have reticulations or ribs, like in the Gothic vaults, this is actually quite efficient to make the overall structure lighter but still equally stiff. And these ribs pick up a certain pattern and nearly give an ornamental character to the space. So when we in the stage of tectonism, it affiliates even becomes even more akin to traditional structures because what connects them is this idea of an adaptive discovery process, except that now we have more degrees of freedom and versatility. And there were more constraints in the old days, but the Gothic cathedral is a wonderful structure and the ribbing pattern is beautiful. But the base repeat, you're constrained to that. What if your church opens up and each bay and ribbing becomes larger and the profiles become deeper and the columns become heavier. The forces to the outside become larger. That's the way we would do. And when you look back at a project like this, you see it's as beautiful, maybe perhaps more beautiful than the Gothic, because you can see the more intricate adaptive structure. The organism is more complex. It's about if you compare a worm to a snake, the worm has an elegant form, but it's segments, you don't know where's the front, where's the back, top, bottom. And the snake has this nice tessellation, which is very different on the bottom, the sides, the top. There's a front, a back, and there's more intricacy and also superiority. If you put this snake into a contest with a worm, we know who wins. And that's the way I would compare a more simple version of architecture and a more sophisticated one. It looks the way it looks because it adapts better and is more optimized. For instance, if you look at the traditional vaults and arches, they're semicircles because easier to install. But a paraboloid arc, in inverted commas, not an arc, form is a much more optimized form.

Yes.

It has problems. It's easier to make. That's why the compromise then the kind of pointed arch. If you go from the semicircular arch to the pointed arch, you approximate more closely the paraboloid. But it's still not the optimal form. Do you find the optimal form? You should take a chain and you hang it. And then you invert it.

And that's not a semicircle. So we work with these more optimal forms. Now, they have a subtle elegance, you know, they're more optimized and they're more subtle. They're very coherent. They're mathematically organized. The physics generates them. And we're not hampered by restrictions of the builders who might have been able to do a paraboloid, but it was also not, how do you draw that? The semicircle, you just have a compass. They didn't know. And also, in the semicircle, you can make each of the stones the same. In the paraboloid, the angle keeps changing. It's harder. And so on and so forth. So that's where we have this kind of we're moving from the worm to the snake. And when you look back at the Korean design, there is this kind of Sagrada Familia or the work of Gaudi and some of his later works, just this character of the most sophisticated. You have to imagine stripping down all the ornament and crustations. These are more sophisticated versions of let's say a kind of basilica form, already first glimpses.

Well, yes, there are different types of AI. Of course, we worked with algorithmic and computational design. That was the biggest revolution. But what you can and we had a lot of tools of form finding, the kind of physics simulations. As I said, the chain model inverted, but not only in the simple arc of three-dimensional chain network. So we had all this. Now when it comes to AI, the image AI, everybody was so excited about that part. Yes, you can accelerate but you rely on the body of work to train up the AI. And some of the most sophisticated versions, which I call tectonism, there isn't enough out there for the AI to pick that up, but we have trained our own AI feeding it with these examples and then it can accelerate that. Yes, you can shortcut and invent the form finding. And it is also lovely and creative when you can interbreed different prompts and different references and images and allow it to compose something more complex. But yes, it is relying on what has been out there. And so we also have to target and can train up the networks so that it can learn the most sophisticated versions. And then of course it speeds up enormously.