Not so long ago an architect had the choice of whether to embrace the computer , whether or not to participate in its advant ages and disadvantages. But our culture has now been engulfed by the computer. It lurks behind any creative or constructive process. It is ubiquitous. Architect s can no longer distance themselves. Pan - tone is nearly extinct, Rotring pens are rare and drawing boards very cheap in a bloated second-hand market.
W e all use computers to represent our architectures and even attempt to seduce our client s with walk-through or fly-through voodoos. But at the same time, the price of computational power has tumbled to the point where we can now embed intelligence directly in a building s fabric. This will liberate the mobility and responsiveness of architecture from the virtual spaces of the computer and drag these attributes into the real.
This unleashing of the digit al into the, until now , stupid husk of architecture has been the subject of much theory and experimentation. In 1969 Gordon Pask wrote an article entitled The architectural relevance of cybernetics in Architectural Design . John Frazer , another pioneer of architectural cybernetics, writes of the paper ...The argument rested on the idea that architect s were first and foremost system designers who had been forced to take an increasing interest in the organisational properties of development, communication and control. Pask identified a significant vacuum in architectural theory and claimed cybernetics as ... a discipline that fills the bill in so far as abstract concept s of cybernetics can be interpreted in architectural terms (and where appropriate, identified with real architectural systems) to form a theory (architectural cybernetics, the cybernetic theory of architecture...) .
The first significant attempt to introduce this new-found theory into real architecture was Cedric Price s 1976 Generator project for the Gullman Paper Corporation in Florida. Generator was a proposal to create changing and adapt able surroundings for between one and 100 people. It would allow them to reconfigure their environment manually or with the aid of a crane.
Formally it was modular , consisting of a series of cubes with interchangeable faces which could be partially or completely recomposed. Philosophically it s concept was close to Price s previous Interaction Centre in Kentish Town and the early 60s proposition of the Fun Palace. Y et it was more radical than any of these schemes in that it could disappear.
This was probably the birth of the intelligent building. Structural elements were anchored with a grid of foundation p ads, and wet services were confined to the east side of the site. It s re-articulation could be rapid as its components were light and of human scale.
What is remarkable about this project is its cybernetic system. The planning of various configurations was controlled by electronics and each of Generator s component s was to be fitted with a logic circuit which in turn was linked to a computer which would control location and usage.
The design team, which included Pask and Frazer, developed a range of programs that would have enabled it to interpret the suggestions of users and contribute it s own ideas.
Three other programs were developed to augment the system. One took the form of a perpetual architect constantly drawing and redrawing the con - figurations and their changes and co-ordinating the trajectory of the crane driver . Another kept track of component s in terms of their location and quantity . A final program would dream up, in the event of a lack of activity with - in Generator, unsolicited plans for adjustment. Here I believe we have for the first time an architecture that is in some sense out of human control, responsive yet operating with another digital impera The moment one start s to design an architecture that depends on digit al processes, one is immediately dealing with a different aesthetic. This can be called the second aesthetic ; it is the aesthetic of the algorithm. The first aesthetic is what architect s have been trained to do, that is to give definitive form to object s and enclosures. The second aesthetic derives it s potency from possibility and numerous-outcome problem-solving.
In computer programming, the algorithm is the diagram that one set s down to map the various stages of solving a problem. It will involve various process - es dependent on it s input data. The algorithm can also be made to attempt to achieve optimum criteria; this type of algorithm is called a genetic algorithm.
Steven Levy defines the genetic algorithm as an ...expeditious formula, a sort of recipe, a key to solving a problem, this particular algorithm is based on genetic principles.
Therefore the genetic algorithm could perform to criteria based on scaling fitness landscapes to achieve optimum performance . This means it can be programmed to aim to achieve an architecturally desired goal.
Two approaches are possible. One is to be able to create evolving architectural form, the other is to be able to create evolving architectural behaviour . This suggest s that architecture becomes an ecology of interactive elements, each re-articulating it self relative to varying spatial criteria and its relationship to it s peers and users.
A good example of this notion is Usman Haque s genetic algorithm surreal floor , featuring interactive talking and light-emitting element s. This piece was created under the tutelage of Professor Stephen Gage, Peter Silver and W ill McLean in the Interactive Workshop at the Bartlett School of Architecture. I will never for - get the sight of Odile Decq (Haque s external examiner) dancing through the installation s strange fields and it talking to her as it sought it s optimum performance.
This level of interactivity is cheap and easily constructed. Many dramatic effects and dynamic systems have been created by the Interactive Workshop, which seems to have taken on the mantle of architectural cybernetics in the UK. Other work includes Jo Plett s virtual pool that ripples when one step s in it, a beautiful and soothing piece; or Tom Holdom s amazing twitching robotic wall pieces that get bored and fall asleep if not entertained by passers-by. These works are merely the beginning of the creation of dynamic building element s that operate in architect-designed machine ecologies.
This is not pie in the sky . All these examples are all achieved with very cheap and simple computational techniques. Architectural cybernetics will be at a wall, ceiling or floor near you soon. The second aesthetic s time has come.
In the next issue of Architech Neil Spiller will consider the imp act on architecture of virtual reality