Computational Design Needs a (Design) raison d’être
In 2005, my mom spent all of her savings so that I could do my internship abroad. I was lucky enough that Office dA in Boston and Plasma Studio in London accepted my application. Monica Ponce and Nader Tehrani were teaching at the GSD in Harvard, and Eva Castro and Holger Kehne at the Architectural Association. Apart from the Venezuelan connection with Monica and Eva, it interested me that these architects were using New Historical Materialism and computational design to design projects.
For them, the use of computation was consequential. It was caused by the theoretical or philosophical ideas they were pursuing. And even though the tools to apply these ideas were rudimentary (Eva Castro’s master thesis was drawn by hand), the translation that Jeff Kipnis — Castro’s masters tutor — had made of Gilles Deleuze’s Historical Materialism to design processes was clear and visible. Later, Kipnis and Ciro Najle cemented these ideas cat the AA Landscape Urbanism master. Eventually, Castro became the Masters lead a couple of years later.
…
I chose London in the end.
There I was, in 2005, in Oxford Street, disappointed by how low density the city was. I was expecting some Manhattan, I guess. Back then, the (computational) design scene in the British capital was boiling. After finishing uni, I came back to London to work for Plasma Studio in 2008. Afterwards, I followed them to their newly opened studio in China in 2011. In Beijing, Eva gave me the chance to teach Landscape Urbanism at an experimental unit the Tsinghua University had opened. Along with Federico Ruberto and Nicola Saladino, I had the opportunity to be fully acquainted with the urban computational scale and how New Historical Materialism (aka Manuel Delanda) could support the development of computational design at the urban scale.
Proto-Computational Designer
In my first semester in uni, I drew the only set of floor plans I have ever drawn by hand. My second design studio teacher, Pietro Scarpa, “forced” us to learn VectorWorks (back then, it was called Minicad). That was 2000. Since then, all my production has become digital, even files to cut the cardboard pieces of my models. And for sure, as BIM software that it is, VectorWorks forced me to do “layer cake” designs. Very limited and unexpressive. Boring.
During the internship, I learned to model in FormZ. Those zigzagging supports under the stair are pretty expressive, aren’t they? I am pretty proud of that solution:
I was asked to model the structure to the millimetre for this house extension. The 3D file was used by the contractor to cut and weld those black metal frames. No 2D plans were needed!
When I joined back Plasma Studio in 2008, they had started using Rhino. Grasshopper had just launched, and Delanda was quite still in the air, joined by Sanford Kwinter and Jeff Kipnis. This is what design was for me: Computers and New Historical Materialism. It was undoubtedly a great bubble I lived in since my uni days.
Bursting the bubble
Recently I learnt that I am quite the brown, Latino, gay, privileged architect. I have been working as an architect in Sweden since 2017. And the contrast between the local way of designing and my design bubble has been stark. Here, AutoCad is the go-to design tool. Normally 2D sketches. SketchUp is very popular to make a 3D version of such sketches. At White Arkitekter, where I have been working, Revit has replaced, to a large extent AutoCad. Sketches are made directly as BIM models. And even though the country is quite left-wing, the benefits of Socialism and a welfare state are palpable (Foucault even taught and lived here!), there is no discussion of computational design. Even less so is that branch of computational design based on New Materialism.
But this is not a Swedish thing. This is a worldwide phenomenon.
Raison d’être
There is a lot of history and current development in computational design here in the Nordics. Finch and Pablo Miranda Carranza (Army of Clerks) are based in Sweden. SpaceMaker is Norwegian. 3XN and BIG are Danish. Most of the Swedish architectural behemoths have their Ladybug-Urban-Generators with web displays. But mainly, these last ones lack design intent, expressivity, and, more than anything, a theoretical background. Most of these tools fit very well with developers. They accelerate the feasibility and test-fitting stages to an instant. But I wouldn’t consider those design tools.
Teaching Landscape Urbanism with Eva Castro, I learnt how GIS, SpaceSyntax, cultural, social and economic factors, design intent and expressivity can coexist in a 3D and digital environment. How can you make such specific cultural, social and economic characteristics digital? Well, that is precisely what Kipnis, Najle, Castro made possible. This is not Patrick’s Parametricism, where a “super” script solves problems. In Landscape Urbanism, the autonomy of the designer, the human one, is vital.
Delanda taught us that it is a fantasy to think we can build everything using a bottom-up approach. Delanda insists that the world is made both of mesh-works and hierarchies, that is, of bottom-up and top-down processes. Hierarchies are made of a collection of smaller mesh-works, just as those are made of smaller hierarchical pieces. It is a nested world.
In Landscape Urbanism, the superscript is broken down. There are 3 clearly defined phases. And the designer is meant to intervene, to use her aesthetic judgments to drive the project where she considers. I can still remember Eva asking me, “Do you like it or not? Do you consider it beautiful?” when I presented a design that I was trying to justify solely on data. Our scripts must allow us to make such subjective and aesthetic calls. This is augmented design, not automated design.
Future
I would love to explain the Landscape Urbanism design process or how New Materialism is vital and can contribute to giving Computational Design a path.
I will instead invite you to reach out. Let’s meet. Let’s talk about all these tools and processes (methods, workflows) that have been available for almost 2 decades already. We, designers, need to engage deeply in what computational design means and how we can maintain our role in the design process. We need to write the scripts that support our position.
My email: libny.pacheco@white.se or hit me on LinkedIn.
