(1) Get a random number from the roulette wheel and say it's the X or horizontal grid coordinate.

(2) Get another random number from the roulette wheel and say it's the Y or vertical grid coordinate. 

(3)  if there is room, stick down a luminous sticker at (X, Y), unless doing so would completely block another blob, in which case stick down a piece of street.

(4) Repeat. 

Paul Coates, The deep structure of the Picturesque

the chinese influence

Chinese developers have a long experience of accelerated urban growth and offer a system of housing construction that is unbeatable both in price and speed. Yet the Chinese development model relies on prerequisites that do not exist in Africa: the African continent is particularly divided 48 aspects, roughly 2000 different languages are spoken that transcend the 54 state boundaries that segment the territory and were, for the most part, defined by colonial powers. Not only is the forced displacement of populations nearly impossible in Africa but the social structures and dwelling habits vary a lot from place to place.



The human mind is linear; we can only think of one thing at a time. When looking at pictures, our mind is only able to focus on one specific scale of the image: if looking at the pixels, we can't see the shapes and if looking at the shapes then pixels disappear. The computer's vision differs from the human ones primarily because in today's paradigm; it is non-linear: the computer can see every pixel and level of shapes simultaneously. Moreover, the computer can see every interaction between those levels that constitute the image.

Because humans can only see linearly, architects design the road, the house and the field one at a time.

What if the Artificial architect could simultaneously design the house from within and its urban context?

In the lecture, we will present applied examples of artificially generated imaginary urban environments:

In Northern Africa, Ghana, the Tallensi is an extensive family of about 400 people all living in one compounded house. The giant house is an aggregation of individual rooms that are built over time, as the family grows. In the language of the Tallensi, people who reside there, the words for “lineage” and “house” are almost identical. On the other hand, car infrastructure organizes the American suburb; this results in a tree-like urban structure.

UEO uses an artificial neural network to redesign the suburb by simultaneously unfolding its roadway infrastructure and the very intimacy of the family house in a nonlinear process. This artificial architect provides us (human) with an unexpected vision of a home and a city.


To test these ideas, the thesis focuses on Tenzug, a sub-Saharan mud house that for now will either disappear or be frozen in time by the UNESCO. The Tallensi home is characterized by its ability to constantly adapt to its social structure: a new wedding will lead to the construction of a new room made from mud. The house results in a cluster of rooms that indexes the growth of the family on a yearly basis. This prodigious adaptability comes from using mud and wood as construction materials: the house is modeled by hand using immediately available materials on site. After a few months with no maintenance, the roofs collapse and the structures quickly decay. While the context of the Tallensi has barely changed over the past 2000 years; its access to global technologies has radically transformed its context and aspirations in the past 100 years.Urban