But what an approach is it? What is creeping in the shadows, only making appearances by proxy? Undoubtedly binary in form and thoroughly regimented, but that is not saying much of its state of being. As mentioned before, digital tools can be used to create material objects. And although plans for a beautifully carved stone Corinthian capital can be produced from its virtual models, it goes against the grain of digital tools – requiring specialized skills, adapted modelling and most of all, virtual emulation of pre-existing material tools.
However when used with the grain, a level of emergence can develop. Far from being in any way inseparably fused with Architecture, it is merely point where the development of a project progresses further and faster when not limited to the speed and capability of the author. Most digital tools in effect simplify the work of the author by receiving a command and responding accordingly. There are a number of various types of such interaction.
An example is a contained command. It may take the form of drawing of a virtual line between two given points – this may result in a simple straight line on a plain at a given angle, a curved line on a flat plane, or a straight line on a deformed plain, or any other formal variant. In either case, the beginning and the end point of the desired process are defined, and the intermediate development is entrusted to the tool.
Another is a single open-ended command, where a starting point is defined by the author and a command (or chain of commands) is enacted. This is in fact an algorithm, where the process is pre-defined and the result is dependent on the material that was fed into the chain at the outset. After the individual steps have been performed, the process stops. Such an algorithm can be used, for instance, to deform an object in a specific repeatable way.
When such algorithms are grouped together in a meaningful way and their results fed back into the process, emergence can occur. The process can also lead to a dead end, a stalemate or a continuous recursive loop. In any case, when only the starting point is defined by the author and verges on becoming irrelevant in a few iterations of the process, and there is no set ending point, the process does in a way acquire a mind of its own. Although fundamentally predictable (unless made not to be), the results are unfathomable as the development process is in a way liberated from the pace and direction of the author.
It is this characteristic that is key to explorations in parametric architecture. In this way a creative act is not comprised of the designated A-to-B act, or an open-ended effect, but results in a scale-less Deleuzeian fractal. Although it may not exist in its entirety, its parts are editable by proxy, not unlike a very elaborate marionette.
The complexities involved and emerging patterns themselves are not limited to the digital in itself, as the Bézier curve was conceptualized long before becoming a widely available digital tool. And Escher’s interlocking patterns exhibit much of the same characteristics. As mentioned before, the digital virtual is an expansion of the concepts that have been in some way or another present for some time.
John Conway’s Game of Life is an excellent example as it straddles the divide between virtualities particularly well. Whereas it’s basic rules are clearly defined in very simple terms (either mathematically or through words), the results show a great level of complexity and a sort of intelligence or will of its own. In a sense it can be said it simulates some aspects of biological behaviour. The resulting ‘intelligence’ is not rooted in a predefined set of wisdoms bestowed upon the growing cells, but rather in their interaction with one another. Through self-organization such cellular automata harness the power of numbers to produce a result that is considerably more than the sum of its parts.
Craig Reynolds’ Boids are another pseudo-biological simulation. Not unlike Life in many aspects, they focus on group movements and have become a useful mathematical explanation of bird flocks. They are essentially virtual entities in three dimensional virtual space guided by basic movement principles. Although no predefined formations are embedded in their behaviour, they soon begin showing patterns of movement not unlike those of actual birds.
Both examples are focused on a particular strand of artificial intelligence focused on a bottom-up approach to complex developments. This is partly relevant to architecture because the growth of cities (and indeed many animal-built structures) has followed similar principles in the past. But primarily, its self-organization can be harnessed to produce unplanned solutions to potential problems – either by identifying trouble spots through simulations, or through iterative problem-solving techniques which would otherwise be out of the scope of analogue research.
Such virtual entities can be grown by importing concepts of an encoded set of abstract instruction (such as DNA), acquiring a capacity for adaptation, growth as well as self-organization. Although irrelevant to their inherent immaterial identity, they can nevertheless be declared to represent viruses, bodies or cities. But even if they show Darwinian evolutionary tendencies acquired through successive generations of improvement, the limits of the original creator can be shown to seep through within a fixed system.
This overt dependence can perhaps be diluted (or hopefully removed) through the introduction of simplexity 17, where there are no direct predetermined causal relationships. By using ‘weighted’ decisions and introduce a kind of learning into the mix, a cloud system begins to appear. In the sense that there isn’t a central decisive point, but rather numerous voting nodes, all contributing to the end result in some small interconnected way.
However, this can again be seen as merely the adaptation of a digital system to human concepts. The brain functions in a similar way, as do simulations of weather patterns or economics – in short, where the variables don’t quite meet up and a degree of ‘slippage’ occurs. Since the decision making process is diluted amongst a vast array of involved elements, the influence of a single author is significantly reduced.
It is perhaps because of these irreconcilable traits that architecture has appropriated these tools. As the role of the architect is not unlike that of a conductor, ensuring numerous disparate agents function in a synchronized and meaningful way, grooming the digital cloud into a meaningful entity is the sensible step forwards for architecture.
