Screen Space, Real Time: ![[MWM_ScreenSpace_Damjan_Jovanovic.pdf]] A revised and expanded text: Real-time simulations are an emerging format for architectural modeling. Going beyond [[drawing]], mapping and diagramming, this format allows for the coexistence of continuous scales within the same model space, dynamic, real time interaction with the onlooker, and potentially raises novel questions on the nature of the notion of [[model]] and modeling, in the disciplinary sense. In contemporary culture, simulations are present mostly through the medium of video games, which are defined by their ability to offer an interactive, playable, immersive experience of a fully modeled virtual space at any scale - from the apartment block of The Sims to the entire universes in the games like No Man’s Sky or Stellaris. In recent discussions (see Federico Campagna’s podcast episode on the work of Stefano Gualeni ), games are seen as a unique way of exploring and participating in a practice of multi-scalar modeling of internally coherent Worlds. Simply put, games are great at producing ever novel and ever more complex, fully self contained model worlds that can be interacted with. A novel name for this practice is emerging in recent years, often under different but related titles: Wordling, Worldbuilding, Worldmaking. All these practices are pointing in the same general direction, and expose the fact that we, as a culture of design, need a new vocabulary when talking about design on such a scale and resolution. At the same time, simulations are entering the art world as a new format of storytelling and expression. The work of the artist Ian Cheng has recently gained global visibility, and is now part of MoMA and other important cultural databases. In his recent series of simulations called ‘Emissaries’ , Ian Cheng is exploring the concept of ‘Worlding’, which he defines as the “the unnatural art of creating an infinite [[Game]] by choosing a present, storytelling its past, simulating its futures, and nurturing its changes.” The concept of an ‘infinite game’ comes from James P. Carse , and offers a window into some of the ideological motivations behind the work. Working with simulations privileges open-ended structure and presupposes a non-deterministic universe, as a liberation from fixed and finite models of thinking and making. Cheng’s work consists of a series of self-playing, open ended simulations of a world imbued with characters and objects. Cheng models the world as a static tableau, introduces the objects and characters into it, and then uses GOFAI models such as behavior trees to run decision making for characters within the world. In this way, Cheng is able to test the boundaries of storytelling and meaning-making, by deploying a non-linear, unpredictable narrative structure of a model world from within. Cheng’s notion of Worlding represents in many ways a narrative paradigm which hybridizes cinema (animation) and games, where the traditional narrative structures are put in friction and an adversarial relation with the open-endedness and the chaotic beauty of a simulation. The classical tropes of storytelling that have always governed how we understand and assign meaning are incapable of regulating the chaos of a simulation, thus breaking our own attentional structure and inviting us to develop new rules of engagement. By making a video game that plays itself, Cheng is able to mark the arrival of the simulation as part of the ‘official’ culture. In a game called ‘Everything’, the artist David O'Reilly has produced a contemporary version of the famous film by Ray and Charles Eames, ‘Powers of Ten’. The game presents an entire universe of things: objects, characters, animals, trees, rocks, buildings, islands, planets, cells, molecules, galaxies, and many more - all of which can be taken over and controlled by the player, in an act of continuous engagement and seeing through the virtual camera. By comparison, in the ‘Powers of Ten’, the camera was able only to zoom in and out, no doubt influenced by the available technical modes of seeing at the time - telescope and microscope. In ‘Everything’ the camera is completely fluid, moving in any direction and switching laterally between scales of things, zooming indefinitely in a kind of a dance of endless access, absolute continuous vision. In ‘Everything’, as in many other contemporary games (for this see for example, Alexander Galloway’s reading of Nintendo’s Metroid Prime), the notion of montage, understood as the bringing together of separate Worlds into the same frame of vision, is completely absent and replaced with a continuous movement of a dynamic camera within the frame. This is a fundamentally different, software-based technical way of seeing, which is only possible in a video game, or as we could say, through the format of a simulation. This is how simulations make available a true multitude of access points to a designed World, by fostering an ecology of seeing, which is the main point of entry into their significance for architectural design. To summarize, a simulation is an interactive, non-linear visual and narrative format that can change and evolve through time, and have potentially endless, different outcomes. ### Beyond [[drawing]] What do simulations mean for architectural design? Architectural design as a process of modeling is dominated by a monoculture of seeing, which we call the orthographic sequence. This monoculture promotes a highly specific, detached, and quasi-objective mode of looking at a World of the architectural project, with only a few, highly controlled access points. A plan is a totally curated image/model, where the access to the underlying space presents an extreme form of control. This architectural monoculture is itself, an outgrowth of what Martin Jay would call ‘the dominant scopic regime of Cartesian Perspectivalism’. It could be argued that, since simulations are deploying virtual cameras, which trace their lineage to perspectival construction, the true innovation of the format is in the types of movement and dynamic access, which blow up abstraction and push designers to ‘think in’ full resolution immersive models, instead of thinking in plans and sections. This is why working with simulations could be considered a radical practice - the language of abstraction that is so tied with the current monoculture of seeing is rendered powerless when faced with the infinitely deep and wide space of the simulation. The work of Lifeforms.io explores the simulation as a design format and the studio works exclusively through interactive, playable models. An interactive model can be seen as a new representational format produced through the use of real-time rendering. In this way, we can understand the format of the simulation as a collapse of two previously separate formats of architectural representation: the image and the model. In the same way in which, as Lev Manovich would argue, all historic mediums have become collapsed into software, architectural design mediums get collapsed into the simulation. In recent years, a number of authors have grappled with the problem of defining what kind of an artifact is produced through digital techniques, with very different conclusions. As my contribution to this debate, I would like to propose the simulation as the right format for understanding digital artifacts in architectural design. Consequently, in the simulation, the act of modeling can be collapsed with the act of rendering, producing a hybrid idea of work, one that combines previously separate notions of production and post-production. An example of this can be seen in The Cloud Garden, a recent project that the studio completed. The Cloud Garden is an architectural story presented through a [[medium]] of the video game, or in our terminology – a playable [[simulation]]. The game is open-ended, there is no fixed goal or ‘end state’.  The story centers on an interplay between two human and two non-human characters, which are controlled by ai, and which are watched over by the ever-present eye and hand of the player, who initially takes the role of the observer. When the game is started, the characters perform a kind of scripted but open-ended choreography that is designed with the use of behavior tree type of [[ai]] , while the player discovers the ability to intervene directly within the small world, taking the role of the designer. This World is designed as a neutral, delimited grid space, invoking both the utopia of Superstudio and the empty stage of [[software]], one which we are exposed to when starting a new file in a software such as Maya or Rhino for example. The World here draws a parallel with the idea of the [[default]] [[space]] of a digital design [[tool]]. In this reading, a software already contains a [[world]], a cold, mathematical uniform reality that is presented as a neutral background, and this world contains infinite possibilities for [[design]] while at the same time being biased towards specific kinds of outcomes. An empty space of a software [[tool]] always already contains in it the Universal [[Space]] of Modernity. The project’s main idea is to imagine what would happen if this world were inhabited by characters that go about their lives, and to imagine if the player acts as if they were a god, from the outside – being able to manipulate reality and affect the lives of these characters. The project attempts to comment on the realities of design production and the disconnect between tools and life that design software takes for granted. This [[World]] is further imagined as a very minimalistic house, that contains only one enclosed box, one bed and one table. At the beginning of each play session, these are randomly distributed on the [[grid]], which affects the characters behavior and their decision-making process. The characters go about their daily routines – the humans are mostly on their phones, the cat lazily moves from one place to another, the dog is always watchful, etc. The routines can be interrupted by the player input, but they cannot be fundamentally altered. The player can build pillars which support a roof-like structure over the entire site, and this structure serves to separate or connect the characters. This story contains another story within it – a user interface panel displays an extra-diegetic text which serves as kind of a manifesto for a post-internet design. The project was produced in Unity as a WebGL build and can be played at bit.ly/thecloudgarden. From this example, we could draw some preliminary conclusions on exactly what kind of things we talk about when we talk about simulations. To invoke multiple histories of technical objects, we could say that simulation is a specific kind of an image and a specific kind of a model at the same time. First, we can understand simulations as a continuation of the history of the moving image, and animation in particular. There are a few primary similarities between these two formats, but there are many more crucial differences. Simulations are interactive, real-time, and potentially endless - while animations depend on determinism, sequencing and scripted events. Animations are fully planned out and delineated, while simulations enable and promote randomness and uncertainty. On the other hand, animations can be easily understood as part of the history of narrative making, they easily lend themselves to (linear) storytelling; simulations have at least a contested relationship with the idea of a story, and rely on a different kind of narration. As with the animations, the significance of this format can also be understood when read as part of the history of drawing and rendering in architecture, or more precisely, with the history of modeling, understood as an umbrella term for all representational testing practices. As ever-shifting and potentially infinitely long aesthetic formats, where probability and [[randomness]] produce a formless, percolating and intractable pictorial space, simulations could be described as post-cinematic images, able to host multiple regimes of representation within the same World. In a simulation, different kinds of images, texts, objects but also whole different genres of representation are able to coexist within the same space. Simulations are pictorial spaces that allow for infinite blending of multiple, traditionally separate or opposing aesthetic regimes: realist figuration of the landscape can coexist with cartoony characters and the abstraction of geometric or textual objects, supported by filmic elements. Simulations present us with a complete collapse of traditional aesthetic regimes into a unified, real time pictorial multiverse. Another history to plug into would be the history of a model, both in terms of architectural models and scientific models. Since with any simulation we are dealing with a changing, dynamic World with many parameters, we could say that there is an inherent act of modeling involved, and a simulation always presents a Model, or more precisely, a multitude of nested models, from mental to technical ones. For instance, when a designer works in any kind of modeling software - they are interacting with a simulation, and working within the simulation model space. They are also constructing a mental representation (model) of that space. An act of design happens through the interaction between these two kinds of models. It could be also said that models made through any kind of digital process belong to the format of the simulation, even though they might still retain traces of older formats, such as collage, for example. In this way, we can say that a simulation is its own, specific technical format, and it can not be reduced to an image - on the contrary, any digital image is an outcome of a simulation, and thus belongs in its orbit. ### Beyond Fictions In recent years, the field of architecture has seen a normalization and acceptance of fictions as modes of thinking and design. In the works of such diverse practitioners like Design Earth, Liam Young, or the works in Benjamin Bratton’s Strelka ‘Terraforming’ program, fictions are deployed as means of capturing (and making sensible, in Ranciere’s term) of very complex phenomena, such as planetary-scale urban conditions and the effects of the Anthropocene. What is evident from these projects is a clear break with traditional means of architectural representation, which can not possibly describe or model such a level of complexity through their existing repertoire of techniques defined by its monoculture of seeing. This ‘storytelling turn’ might be seen as a means to connect architecture to the predominant visual culture of today, but it could also be read as an admission of the discipline’s inability to model in the contemporary context. The technical logic of a design fiction is the regime of compositing, and not that of a simulation. Design fictions inherit their vocabulary from cinema and animation, and thus produce results which follow the logic of storytelling and not modeling. This brings us to a crucial aspect of the format: **a simulation is not a fiction, or not merely a fiction.** In the book “Virtual Worlds as Philosophical Tools”, Stefano Gualeni traces a lineage of thinking about simulations to Espen Aarseth’s 1994 text called Hyper/Text/Theory, and specifically to the idea of cybertextuality.  According to this line of reasoning, simulations contain an element not found in fiction and thus need an ontological category of their own. “Simulations are somewhere between reality and fiction: they are not obliged to represent reality, but they have an empirical logic of their own, and therefore should not be called fictions.” This presents us with a fundamental insight into the use of simulations as the future of architectural design: they model internally coherent, testable Worlds and go beyond mere fiction-making into Worldmaking proper. Architecture’s power of engaging with complex phenomena through modeling can be recaptured through the use of simulations. This was the main idea behind a seminar called Games and Worldmaking that I have led at SCI-Arc in the summer of 2021. In the seminar, we worked with Unreal Engine to create a multi-scalar, interactive, playable model of a self-sustaining, wind and solar-powered robotic garden, set in a desert landscape. The simulation, called Planet Garden, was envisioned as a kind of reverse city builder, where a goal of the [[Game]] is to terraform a desert landscape by deploying different kinds of energy blocks, until the right conditions are met for planting and the production of oxygen. The aim of the seminar was to learn how to model dynamic systems and explore how to utilize game workflows as ways to address urban issues. The work was a result of a renewed interest in exploring how non-linear models are represented and imaged. One of the main ways in which video games achieve this is through techniques that come from System Dynamics, a discipline that deals with the modeling of complex, unpredictable and non-linear phenomena. The field has its roots in the cybernetic theories of Norbert Wiener, but it was formalized and created in the mid-1950s by Prof. Jay Forrester at MIT, and later developed by Donella H. Meadows in her seminal book ‘Thinking in Systems’. System dynamics is an approach to understanding the nonlinear behavior of complex systems over time using stocks, flows, internal feedback loops, table functions and time delays. Forrester’s ideas were explored by the game designer Will Wright in a series of games that he designed for Maxis, including SimCity, SimEarth and The Sims. This approach has since been adopted as one of the main ways to model game systems as it enables easier understanding of systemic depth and causality in time based scenarios. The approach was also used in a series of early 2000s software projects by MVRDV , as means of working with urban phenomena. In these projects, MVRDV worked actively on establishing interactive modeling as a valid and accepted practice for urban design, and produced a series of tools and applications that were used in real world modeling scenarios. One of the main outcomes of the seminar was an understanding of the need to go beyond cybernetics and learn from contemporary, probabilistic ways of modeling through the use of machine learning. Eventually, simulations could potentially enable an emergence of a new kind of architectural project, graced with a novel, AI-based take on dynamic systems approach to modeling and with a new legibility in its representational strategy. Simulations also open up a possibility for a project to become collapsed with its physical architectural counterpart. A simulation points to a possible future where the architectural project becomes a holistic, systemic collapse of image, model and physical space into a single time-based construct, which would dissolve the existing Albertian separation of idea and corpus into a new digital/physical continuum. **