Rebecca Allen visuals for Kraftwerk concert at MoMA // 2012. Rebecca Allen is the current creative director for Nicholas Negroponte’s projects. image courtesy of MoMA (source)
assigned reading: ‘Toward a Theory of Architecture Machines’ by Nicholas Negroponte , 1969.
Today, the widespread use of computer-aided design has created a paradigm shift in the way architecture is conceived, represented and fabricated. The expressive and constructive potential of the digital has been -and still is- thoroughly studied by theorists and architects who’s position towards the digital is one of admiration or skepticism or both. It is obvious that machines have enabled designers to overcome the traditional constraints of thought and set the grounds for a higher level of formal and programmatic complexity with numerous aesthetic, functional and performative implications. But even with these advantages, architecture still remains in the hands of the architects who carry the professional expertise to successfully translate the user’s needs and desires into design. This disciplinary enclosure is further accentuated by the growing sophistication of digital architecture. Often it deploys an unfamiliar vocabulary, emphasises the role of the “expert” and makes the process of architectural design increasingly inaccessible to the user.
To understand exactly how this paradigm shift has evolved, its important to focus the study in the ’65-’80 period, which were the years of active dialogues about the new issues of computer assisted design and user participation. Even though the technological expressions of the theoretical work of those times may be outdated, it offers a very interesting insight on the clarity in the way they manipulated the concept of computers and their users.
Post World War II cybernetics was defined as “the entire field of control and communication in machines and in living organism” (1947). This opened up a new field of research for artists, planners and architects who were looking to explore the blending of the organic with the technological, – such as: Kepes, Banham, the Eameses, TeamX, Fuller, Archigram, Ant Farm, Nicholas Negroponte etc.
The 1960′s were the years of Archigram’s Walking Cities, Nicholas Negroponte’s cybernetic Architecture Machines, RCA synthesisers and Asimov’s 3 rules of robotics.
In 1969, in his paper “Toward a Theory of Architecture Machines“, a prelude to his 1977 book “Architecture Machines”, Negroponte questions the way we use machines in relation to architecture and discusses the new possibilities that these machines can offer.
It is an investigation of the way the users communicate with these machines and how the machine interprets this input. Although his work can be extrapolated in many fields, he focuses on the machine in relation to architecture and how technology has enabled design to go beyond the habits of the past.
Negroponte uses an introductory example to raise our sensitivity towards the machine. He speaks about the “learning machine” in a manner one might speak about a very young child or a pet. The study given as an example is simple, about a machine that is given to analyse ten different cuboids and use its experience from this analysis to produce an eleventh on its own. On this background, he ask the questions : can it learn more? can we not teach it about architecture history, context and sociology? and with all that knowledge, can the machine not produce appropriate architecture?
Continuing this idea, he goes on to speak about a designer-machine partnership.For a better understanding of his work, it is important to repeat his view about the architecture machines subassembly. The architecture machine is described as a sum of five main individual systems: a heuristic mechanism (the strategies the machine uses), a rote apparatus ( a method of learning from experiences), a conditioning device (a way to develop conditioned reflexes), a reward selector (a possibility for the user/designer to provide input on the machines results in order for it to be able to evaluate, observe and set goals for itself. In other words, a way to “domesticate” the machine.), a forgetting convenience (a way to discard irrelevant data).By breaking down the machine into these simple sub-parts, he had unknowingly a prophetic view on todays computer assisted design, a tool we are so accustomed with and we so easily take for granted. Not only does he for-see the tools of today, but he mentions the need to create a natural language that enables the designer and the machine to communicate, the interface.
Negroponte investigates the machines expressive limitations, with a dose of healthy skepticism. His concerns are not only about the results of the machines creativity being bad or good, but also the copyright of this results. His direction wants to bridge the gap between the humanistic discipline of architecture and the technical rationality of computing. While parts of his subassembly are easily recognised in the softwares we use today, the “reward mechanism” is lacking completely. While parallel fields research are exploring the possibilities of machines learning, robots with artificial intelligence able to recognise human emotions and able to learn, the architecture field has yet to give way to such freedom for machines.
Negroponte’s discourse goes beyond this view of using machines as facilitation tools. He bases his entire argumentation on the assumption that the functions of communication, inference, understanding of the context and self improvement -in other words intelligence, will raise the machines to the level of valuable collaborators, not problem-solving artefacts, but problem-worrying partners of the designer, allowing him to manage inconceivable complexities and stand critically in front of his own work, with beneficial results both for him and the user. We cannot ignore, and neither does Negroponte , the problems of a historically defined technological utopia that considered artificial intelligence as a soon realisable goal. However, the demands that he sets for computer personalisation and idiosyncratic approaches to computer graphics are definitely very valuable contributions in the field of machines designed for user empowerment.
Negroponte’s beliefs do not go as far as Yona Friedman’s whose belief, as stated in “Toward a Scientific Architecture”, is that the architect should be removed completely from the equation of the creating architecture, and that the machine should provide such an interface that can be used directly by the beneficiary. In this theory, the architect as a mediator between human needs and machine, as a translator of languages between human desires and digital input, disappears, and with this, Friedman believes that architecture will no longer be subject to the architects obsessions and “repertoire”, but it will respond directly to the requests of a community. While not directly sharing Friedmans views, both their works intersect at ethical points when the question of intelligent machines is raised. It is interesting to see such concepts discussed in a clear detailed manner way before their time.
The issue of computer aided design and software design decisions and their strong influence on how architecture today is conceptualised and expressed is still a question of great importance. This idea that a computer can have a learning process and then develop its own idiosyncrasies is one that still has to be applied in the computer aided design softwares. More than that, a sci-fi exploration of man-machine interaction can give way to new communication models of high theoretical value. Participatory design, in the sense of a user programmable architecture machine, could be a good topic to explore. These early studies and works, like the ones of Nicholas Negroponte, constitute a good base and a direction for a productive criticism and further research.