Building my learning machine, and the building as a learning machine

This idea spawned somewhat spontaneously while I was thinking back over a quote that's stuck with me for the better part of three years. It was delivered with such colloquial, intelligent profundity that I couldn't help but wholly absorb it, as well as the rest of the information delivered with the speech it came from (shown below).

"If you look at the interactions of a human brain, as we heard yesterday from a number of presentations, intelligence is wonderfully interactive. The brain isn't divided into compartments. In fact, creativity -- which I define as the process of having original ideas that have value -- more often than not comes about through the interaction of different disciplinary ways of seeing things." (Ken Robinson)

There are two aspects of this quote that stand out to me with respect to this final year studio and architecture. Firstly, that an original idea that has value "more often than not comes about through the interaction of different disciplinary ways of seeing things", and that this would be a useful way to build my learning machine (ie, the processes I will follow for research and development) for this studio. Secondly, that "intelligence is wonderfully interactive", and that a brain "isn't divided into compartments", and that these points can be used to conceptualise a building as a learning machine.



The Building of My Learning Machine

The first aspect is how I want to work through this master's year studio - the interaction of different disciplinary ways of seeing things to generate original ideas that have value. I will select three disciplinary ways of seeing things which seem to have promising potential when interacting with the disciplinary way of seeing things that is architectural design. Currently, the three disciplinary ways of seeing things I have chosen are:

1. Biology,
2. Evolution, and
3. Computing.

The Building as a Learning Machine

This train of thought fits best under the discipline of biology, but still has some strong relationships with evolution and computing.

The idea of an intelligent, interactive building is an exciting one, and seems to be cropping up more and more these days (specific examples to be searched up later and added retrospectively so I don't break my train of thought). Most of the time, no one part of the building's operation and usage is wholly distinct and unaffected by the other parts. Thus, according to the definition Ken Robinson uses, it could be said that a building is like a brain. Historically, humans have been creating braindead buildings. Sometimes beautiful buildings, but braindead nonetheless. They operate, they breathe, and everything like that - but they're vegetables. They cannot respond to us. Or if they do, it's only in obvious terms, through wholly controlled interactions. Recently, however, technologically-minded interventions have introduced the capacity for reactions in the built form. It's as though some buildings and building elements are now recovering from a coma, and are starting to be able to autonomously respond in complex ways to interaction.

Though Ken's speech was explicitly about human learning and education, it is my contention that interactive, responsive architecture - occasionally stylistically classified as "high technology" - would benefit from such a conceptualisation. The building as a learning machine is an interesting, exciting idea. Due specifically to technological advances, built forms are capable of being dynamic, and adapting to their purpose. Through analysis of occupational usage - which in the case of this studio would likely be emulated through data obtained via social website analysis - buildings can be programmatically designed to modify themselves to suit occupational use.

The conceptualisation of the building as a learning machine can extend beyond this specific example. More generally, the building as a learning machine is capable of responding to its environment and, moreover, learning the trends. Conceptualised as a finite state machine, the learning machine's state would change in response to the input from its environment, with subsequent states depending on the previous states as a form of trend-learning. In a way, having the building as a learning machine relates somewhat to forming a practical application of phenomenalism.

There are many directions this idea could go, so to keep the project feasible, I will need to restrict myself to researching and developing one or a few specific examples.

You can find other speeches and presentations from TEDtalks here, or through the TEDtalks YouTube channel.