| |
|
1.2 _ MORPHOLO OR DE ARS COMBINATORIA
|
| |
|
The Morpholo (by Thieri Foulc / Ou. Pein. Po) is a 'combinatoria' of
square tiles which can be arranged in different manners, as a game, as
art, or as architectonic component. The tiles can be juxtaposed at will,
yielding several billion unforeseen larger shapes. There is only one rule:
to match, on the edges, black against black and white against white. This
is a formal rule, comparable to the 'fixed forms' that are frequently
observed in literature, but seldom in the plastic arts. The 'edge contraint'
rule does not interfere with the working of the artist's imagination:
the shapes on the surface of the squares can be varied freely. Nor does
this rule guarantee a result: like the framework of the ballad or sonnet,
it is capable of bad or good. It all depends on the artist drawing the
tiles as well on the combinations that come into play.
|
|
Each side of the square may be divided in half, and each half-side may
be designated black or white. For each side there are 4 possible structures,
for the square there are therefore 44=256 possible strucutres. The tiles
may be numbered according to the order of their generation.
|
|
Applying the morpholo as a mural, for example, it is possible to create
a complete room (6.40 x 6.40 x 3.20 m) floor, walls and ceiling by distributing
256 tiles, mesuring each of them 80 cm per side. This potential application
has been achieved using Diasec ® squares which are backed with a film
of Ferriflex ® magnetised rubber. The magnetized squares are attached
to a metallic armature, and they can be moved around at will. In this
way the decor of the mural can be changed on daily basis.
|
| |
|
1.3 _ THE RNA GAME
|
| |
|
"Evolution as such is comparable to a learning process based on
reproductive acts of memory. We can simulate the basic principle of such
a learning process in a game" (Eigen, Laws of the games, 1981).
|
| |
|
Given the enourmous number of possible routes they could take, it is
astonishing how rapidly evolutionary processes make their way to the goal
of optimally adpted structures. It's possible to demonstrate how this
happens with the help of the RNA game.
|
|
Each player has a string with a total of 80 breads on it. The red, green,
blue and yellow beads, corresponding to the four buildings block of RNA,
are arranged in a irregular sequence. Since this is an 'evolution game',
we have a mutation die in the form of a tetrahedron whose four surfaces
are colored red, green, blue, and yellow respectively. In each round of
play, this die is used to mutate, in accordance with the coulor rolled,
a bead located in a certain position. The point of the game is, beginning
with a chance arrangement of the chain, to create as quickly as possible
a structure of folds that is charcterize by a maximum number of complementary
pairs. The roll of the mutation die applies to a bead designeted by the
player in advance, and only this bead can be exchanged for a bead of the
color rolled whenever it seems advantageous to do so. The following rules
must be strictly adhered to: 1.the steric rule: because the formation
of nase pairs between different sections of a sequence can be accomplished
only by folding the chain on a plane, loops will inevitably result; 2.the
rule of complementarity: if two beads of comkementary colors (red-green
or blue-yellow) are opposite each other in the folded RNA structure and
if the rule three is fulfilled, these beads are considered a pair and
are linked together; 3.the rule of cooperativity: the linking of complementary
beads in a pair can occur only if there is an unbroken sequence of at
least four red-green pairs, two red-green pairs and one blue-yellow pair.
These stable base pairs are considered 'selected' and are no longer subject
to the roll of the dice.All those rules are based on data established
in experiments and therefore represent the behavior of RNA molecules in
a thoroughly realistic way. It comes as no surprise, then, that the structures
that prove victorious in this game are the very ones that have also won
in evolution and that can be found in nature wherever the conditions of
stability necessary for their selection are present.
|
| |
|
1.4 _ SHEET METAL STRUCTURES FOR MILGO MANIFACTURE
|
| |
|
The project by Haresh Lalvani deals with software-driven fabrication
of sheet metal for architectural surface structures and is being carried
out at Milgo's manufacturing facility, testing the relationship between
an artificial 'genetic code' and the manufacturing process. All sheet-metal
structures were generated using a morphologically encoded algorithm, which
provides the possibility to generate endless 'variations on a theme' by
manipulating the code. As a result, no two structures need be alike, so
each individual in the world could have their own unique structure. A
procedure was developed whereby single continuous metal sheets could be
marked by computer-driven equipment and then folded (manually, for now).
The resulting structures are employable as both architectural and industrial
design products as well as pattern of complete environments. The algorithmic
approach permits the structure to be modelled, transformed and fabricated
with ease. The expectation is that the morphologic elegnce in the shaping
would also translate into an economy in building.
|
| |
|
"The concept of the genetic code… permits each of us to be unique
yet encoded by the same basic genetic alphabets (DNA bases). These sheet
metal structures are morphologically coded in a similar way." (H.
Lalvani)
|
| |
|
1.5 _ PRIMORDIAL LIFE 3.16
|
| |
|
"I wrote Primordial Life to capture the principles of evolution
in an interesting and visual way" (J. Spofford). Designed to run
as a screen server or in a window of PC, Primordial Life presents an environment
filled with artificially living creatures called biots. Like their biological
counterparts, each biot has a genetic code that serves as a blueprint
for constructing it. It is this blueprint which lays the foundation for
their evolution. The biots battle to dominate the environment. Over time,
new species will emerge while others may die out. It's a tough life, but
some "biots" got to do it.
|
|
The genetic code consists of 10 segment genes and one trait gene. A line
is made up of 1 to 10 segments, and a biot can have 1 to 8 lines. Biots
are made up entirely of lines. Each line has a unique color and a purpose.
Green lines absorb solar energy and convert it to biot energy. Without
biots with green lines, all biots would surely perish. While green lines
are extremely handy for capturing energy from the environment, they are
vulnerable to attack from the menacing biots with red lines. Biots with
red lines are considered predators, or perhaps herbivores. They steal
energy from plants by attacking their green lines. When a predator touches
a green line, the predator gains energy. In addition, the plant may lose
its line. If a red line is longer than a green line, the green line will
be "plucked" off the biot. If there were any lines that came
after the green line, they too are loss. Now, while this may sound tragic,
all biots have the ability to regenerate lines they have loss, provided
they have enough energy to do so. If a red line is shorter, the green
line just dims, and takes much less time to regenerate. There are also
other lines that are vulnerable to attack by predators. These are the
white and light blue lines. White lines are only visible when "Sexual
Reproduction" has been enabled.When a white line touches any line,
other than another white line, it injects the biots genetic code into
the target biot. In this manner, fertilization takes place.The last biot
to touch a biot before it gives birth is considered to be the father.
White lines, however, are just as vulnerable as green lines. Light blue
lines are little rocket motors that pull a biot around. They fire serially
at a frequency determined by the genetic code.They cause biots to rotate,
or travel in an apparently erratic path throughout the environment. Light
blue lines, are vulnerable to attack by red lines. Biots with blue lines,
or defenders, can shield themselves from a predator's red lines. When
defenders collide, no harm is done. A predator with a red line longer
than the length of a blue line can "pluck" the shield of a defender
and continue the attack. Likewise, a defender with a longer blue line
can "pluck" the predators red line and potentially halt the
assault.
|
| |
|
1.6 _ QUANTUM GAMES, the latest frontier of games
|
| |
|
"The best way to win a game like 'paper-stone-scissors' is to use
"quantum rules". Baffle your opponent by pulling out a hand that is simultaneously
'paper' and 'scissors'" (Phil Ball).
|
| |
|
Recently, three physicists, Eisert,Wilkens, and Lewenstein claim that
a game that typically ends in a kind of frustrating stalemate ceases to
do so if played with "quantum rules".
|
|
The game is called the Prisoner's Dilemma, and it pitches two players
against one another in a manner reminiscent of the playground game of
paper-scissors-stone. The players can choose to 'cooperate' or 'defect',
and receive a certain payoff depending on their choice and their opponents.
The 'dilemma' is that the payoffs imply that a rational player should
always defect; yet they both get a better payoff for mutual cooperation
than for mutual defection.
|
|
This game serves as a model for individuals selecting strategies in competitive
scenarios both in evolution and human social and political systems. It
appears to suggest that cooperation can arise only if we can overcome
our 'rational' instinct towards selfishness. But now the three physicists,
in an experimenl realized on a nuclear magnetic resonance quantum computer
have shown that there is a Third Way: to play using quantum strategies.
This, they say, creates a new best strategy that offers the reward only
the irrational, cooperative strategy secures in our 'classical' world.
|
|
How do you play a game with quantum rules? In the classical Prisoner's
Dilemma, one can make only a single choice: cooperate or defect. Each
player had a 'qubit'. This quantum bit, rather like a computer's zero
or one binary digit, but with an important difference: a qubit can exist
neither in one state nor another, but in a mixture a superposition of
both. In the paper-scissors-stone game, this would be analogous to whipping
out a hand that is both paper and scissors at the same time not half of
each, or halfway between the two, but capable of "collapsing" into either
pure state.
|
|
But the quantum game can be even more strange, because it is possible
in this case to make the choices of the two players "entangled", so that
each influences the other. When two quantum particles, such as photons
of light, are created in an entangled state, making a measurement on one
of them automatically fixes the corresponding property of the other. In
an entangled game, the researchers show that the best strategy of both
players is neither to defect nor to cooperate, but to offer some strange
concatenation of the possible outcomes open to them. No one player can
improve their own fate by changing their move alone. It is hard to give
any classical description of this strategy, other than to say that when
both players use it, they both come off as well as they possibly can.
In other words, this quantum strategy is not only the most rational but
also the most profitable. There is no longer any dilemma.
|
|
Of course, real creatures play the game in a classical world, where defection
or co-operation are the only options. But Eisert and colleagues point
out that some aspects of their quantum game resemble the interactions
that go on when information is exchanged quantum-mechanically something
that has been shown to ensure secure communication of coded signals. Perhaps
the greatest significance is philosophical, however: quantum mechanics
can dissolve deadlocks, or 'strange loop', that, according to classical
logic, appear insuperable.
|
| 2_
BUILDING SCENARIO
Scenario planning could be a very crucial
tool to help an architect design the correct solution for the future.
In the game of architecture, after making certain rules and defining the
roles of each element, you start creating different scenarios. Different
results are obtained when implying those scenarios to the architectural
game. That is the aim of scenario planning. This allows one to have a
better idea of where the future may be heading and, in essence, help design
a better architectural solution for the future.
Scenarios don't predict the future so
much as they illuminate it, preparing us for the unexpected. Scenarios
are multiple approaches to the future, and stories of the inevitable.
The best scenarios aren't necessarily those that come true; they're the
ones that subvert expectation, providing deep insights into the changes
happening all around us. The better scenarios are, the more they penetrate
to the deepest possible understanding of the present. (P.Mc Corduck and
N. ramsey, 1996 ,p.18).
A lot of companies are faced with the
dilemma of trying to make correct, confident decisions for the future.
A lot of individuals, also, would like to be confident in making the correct
decisions for their future. Parents, for example, ask what education they
should provide themselves or their kids, what stocks they should invest
in- and so on. This is where scenario planning comes in to play. It helps
us see where the future might be heading, possibly for different yet realistic
predictions and they're for help us today, in our decisions for tomorrow.
The decision parents are faced with regarding the education of their children
is the same decision an architect is faced when having to choose a particular
design solution for the future.
We must understand that scenario planning is not a crystal ball to the
future and is therefore not a precise picture of the future. It is more
a plan that under lines the driving factors or forces that my effect the
future one-way or the other. Scenario planning helps the planer recognize
these forces in order to help make better decisions today. A great example
is a tornado. The satellite pictures along with readings from inside the
tornado, that are being transmitted through equipment inside the eye of
the tornado, helps the weatherman predict which way the tornado might
move based on the forces surrounding the tornado. You can therefore see
the satellite reading and equipment inside the tornado as the scenario
plan that is helping the planner (weatherman) predict the movement of
the tornado. It's a very similar situation with the future accept tornados
are much faster.
Basically, using the scenario method means you write three or four in
depth stories
about the future, evolving around different combination of a number of
plots or logics. The stories have to be internally consistent and they
have to contain elements of present reality. All scenarios evolve around
the same issue, which you want to gain insight about and/ or decide upon.
Thus, you start with isolating and formulating a question or a decision
upon which you want to build the scenarios. This can be a question or
decision concerning your organization, a society or any other unit of
interest
Scenario planning originally was a military tool used by the U.S Air Force.
It was great in the prediction of war games. After World War II, the RAND
CORPORATION who further developed the Scenario planning technique with
the help of the Hudson Institute took up scenario planning. Herman Kahn
established the institute after his resignation from RAND. Kahn reworked
scenario planning in his famous book the year 2000 (1967), and applied
it to be used as a tool for the business industry. After the 1960s scenario
planning found its way into the corporate world and quickly became an
essential tool for many a company. In the 1980s Shell used scenario planning
to predict the future of the oil industry. They assumed there would be
another substitute for oil. Based on that and many other factors, scenario
planning helped Shell rise from fourteen to second place among the oil
companies.
Scenario can be successful in structuring uncertainty only when (1) they
are based
On a sound analysis of reality, and (2) they change the decision maker's
assumptions
About how the world works and compel him to change his image of reality.
(P. Wack,
1996, p.32 )
To build a scenario plan you have to identify the focal issue that is
related to your future or your company's future and then decide what are
the driving forces that will have a major effect on your scenario. In
general there are four major forces that have the most impact on scenarios.
The first force is the motivation for being socially responsible. This
force refers to the interaction of, among others, experiences, Idea, awareness
and interpretation of problems and developments, incentive, impediment,
perceived gains and losses which underlie motivations of individuals and
groups in their various- and often multiple- roles in society. The second
force -Glocalisation- refers to the simultaneous process of globalization
and localization. Glocalisation is a term, which includes on the one hand,
the development of a world economy based on capitalism, the expansion
of (free) trade, the growth and, there with increasing impact of multinational
corporation on global liberalization of finance. The third force is transfer
of responsibilities. This simply means things such as health care, social
insurance, and so on -traditionally held by government or public institution-
is now handed over to private business, and the NGO (non government organizations).
The NGO are usually single-issue and non- profit organizations. The fourth
and last force is technology including things like communication and information
technology and how fast or slow it's growing or expanding.
These same forces can be applied to help
create scenarios specific to architectural solutions. As stated earlier
for the world of architecture other elements and rules need to be defined
to help make a more suitable scenario. Perhaps things like population
growth and the demographics that go along with it. Will people need more
comfort in their habitats or will they only need the necessities. Other
forces can include such things as city constructions, planning and amount
of physical space available. All these forces along with the four main
ones explained above will help predict what architectural solutions need
to be designed to best fit the future.
After explaining how to build a scenario,
the important of making scenario planning,
And how will it affect the future? I want to show you two example of how
each
company or individual can build a scenario and determine the major forces
that can
affect his scenario plan. The firs example is how to build a scenario
that has been
published by Taylor Management Center in Boulder, Colorado (MG Taylor
Corporation, The Manual, 204,1983), and the second example using scenario
to
develop strategy for company that develops, manufactures and market test
and
measurement equipment in an industry that is changing rapidly.
First Example:
Purpose: To order over time your projections
of events in the future.
To create a map and model to guide you in your planning and decision making
process.
Method:
1. Decide on a general-purpose or specific
field scenario. If your scenario is oriented to a specific discipline,
watch for developments in other fields that impact your field. Look at
relationships networks of events.
2. Determine the number of years you will
forecast for in the scenario. Go beyond your particular need, i.e., if
you are looking at a field around a five year goal, build the scenario
for eight or ten years.
3. Start the process with the right place
to work, such as a Management Center with walls you can write on. Otherwise,
use long sheets of butcher paper tacked on the wall. It helps to have
an assortment of colored markers to write with. When your tools are in
order, divide your work surface into equal portions-years, weeks, days-according
to the time period you wan forecast.
4. Use each forecast as a springboard for
adding more events to the scenario, continuing to work from the present
to the future and from the future to the present.
5. Check each forecast for lag time. What
is the time lag between a new idea and its market place implementation
in your fields of concern?
6. Collect information and document your
work.
- Look for facts to support or deny your scenario
- Utilize Delphi methods.
- Clip from newspapers and magazines. Information in books is several
years old by the time it reaches the market place. The order in which
information appears in periodicals is important to note, for example:
Science News addresses the newest discoveries; Scientific American writes
about "known" facts; Popular Science gives product announcements;
and fortune provides business assessments.
- Update your scenario at regular intervals. This enables you to calibrate
the rate of changes your field.
7. Keep records of all your scenarios.
8. Always strive for a sense of synergy
in your scenario. The principle of synergy states that the behavior of
a whole system is unpredictable from the behavior of its parts. This premise
leads to corollary; once you understanding of future change, you can make
predictions the "experts" will miss. You become better able
to direct your personal destiny and destination of your company.
Second Example:
This is a five years scenario plan for a company called T&M, originally
started
as supplier to other business units in the multinational company, but
over the
years also developed a significant customer base outside the multinational
company and outside its original area of business.
After being divested, the management team of T&M felt it should rethink
its strategy.
This need was not felt to satisfy the new parent company, but because
they wanted to
get a better insight into the effects of digitization on their industry
and to get
more focus into areas of technology, development and marketing. The scenarios
served as a tool to articulate the different implicit assumptions that
were already
guiding the policies in the company.
The following three scenarios were developed:
Tin lizzy: Stability in a growing economy (scenario 1)
Happy, go lucky: Room for change (scenario 2)
World of Chaos: Turbulence in technology and politics (scenario 3)
The result of the scenario- to - strategy project on this company, it
changed from a
small, backward looking, company in analogue Test and Measurement equipment
into
three promising, mid - sized, experienced start-up companies in digital
multimedia
equipment
Conclusion:
Scenario planning is simply a way to help us deal with the uncertainties
of the future and be more aware of the driving forces that helps us spot
which future it will be. This in essence helps us make our today decisions
more confidently. This same thought process is easily applied to many
applications including the game of architecture. It can help the architects
of today build the correct designs for the people of tomorrow.
|
|
3_ EVOLUTIONARY ARCHITECTURE
|
| |
|
In the last chapter we've been trough how scenarios could predict the
future of architecture and the needs, in certain way, of new users. Architects
could also create scenarios to achieve an architectural concept, which
once located in its position will interact and have a better performance
with its future environment.
|
| |
|
But, how does architecture could be predicted? The growing world environment
is dominated by man-made Eco-systems. In this man-made nature, architecture
is a fundamental part. So from this point we should start considering
architecture as organic. This organic in architecture shall mean the fact
of understanding it as an evolving matter, just like organisms. In order
to predict a scenario for architecture is important to develop an Evolutionary
Model of Architecture. The formation, development, eventual extinction
of architectural concepts, could only be explained trough a general theory
of adaptive change used as well in the biological sciences.
|
| |
|
3.1 Genetic Language as a Code Script
|
| |
|
The Architectural Concept is created in the architect's mind in terms
of space, structure and form. These concepts can be expressed like genetic
language, which produce a code script used in form generation. When we
give a computer this "genetic language" of architecture we will
be able to set it in a predefined environment and accelerate or decelerate
the natural process of nature to see the behaviour of our architectural
concepts. Then we will treat architecture as an artificial intelligent
life responding to its environmental changes.
|
| |
|
Trough the computer this acceleration or deceleration of the natural
life of architecture in its new "created" environment could
be traced and taken to new stages where the crafty research through studio
work could almost be impossible.
|
| |
|
3.2_ Antecedents
|
| |
|
The theory of evolution was developed thanks to the work of two men:
Charles Darwin and Gregor Mendel. Darwin defined the mechanism of evolution,
natural selection. Mendel's achievement was to identify the gene as the
unit of inheritance.
|
| |
|
Alan Turing was in 1935 concerned about provability for which he made
a universal computing machine, the Turing Machine. The machine should
be capable of performing any computable process by following a set of
logical instructions on an endless paper tape. John von Neumann developed
the serial basis for a serial computer. His research leads him to develop
a theoretical framework for a self-replicating computer. Both of them
were basically interested in conceptual computers, generative processes
and the nature of living processes.
|
| |
|
John Koza laid the foundation for genetic programming (GP) in 1992 (Koza
92) and GP has since been applied to a range of different problems and
disciplines. Research by Peter Bentley into a generic design tool using
genetic algorithms introduces the concept of a multipurpose form-creating
interpretation of a genotype and library of fitness functions for different
purposes (Bentley 96).
|
| |
|
3.3_ Architecture as Biology
|
| |
|
Before going on with the following part its important to define what
does architecture means to the aim of its development as an evolving intelligent
living organism.
|
|
Lets understand architecture as a cultural system which adapts its forms
in order to represent the relationship between the different elements
coexisting in its environment. Architecture is the information which characterizes
the forms of buildings which themselves are material objects.
|
| |
|
Evolution is the continuous changes make by an organism in order to adapt
to the changes imposed by its changing environment. This environment is
conformed by a group of systems in constant communication trying to adapt
its behaviour to the state of all the others. The community of organisms
which genetic information is exchanged leading to a similarity of physical
characteristics within the community is called specie.
|
| |
|
Trying to adapt the theory of the biological sciences to architectural
terms, we should note that there's clearly no similarity between architecture
and living organism. What we should try to empathise is the analogies
which could be found to establish a parallel between architecture and
biology. This parallel will lead us to run as we explained at the beginning
a scenario where treating architecture as a living organism will give
us an evolved architecture (how architecture will be).
|
| |
|
We can make a catalogue of equivalences and similarities between architecture
and biology.
|
| |
|
In comparison the representation in Architecture is equivalent to the
biological concept of adaptation. Architecture as a system tries to represent
its environment with its architectural element. As its elements try to
match the functions, organization and basically cultural matters of its
environment which is then, represented in space and time. This also happens
in nature where living organisms adapt themselves to fit in its environment.
E.g. white bears and vernacular architecture in the Artic.
|
| |
|
When talking about the Style in architecture we could also be talking
about Species in biology. An architectural style is not a homogenous entity.
It defines a range of formal characteristics which are share by a large
number of different buildings. As in nature when species match their genetic
needs mating could be done between them (sharing so the genetic information).
Architectural style matches the communication between architects in what
we could call style. The typical elements of a style are equivalent to
the genetic complement of species both providing the heredity mechanism
necessary for an evolutionary system. Like the DNA molecule, the information
shared by different buildings of the same style gives a pool of possibilities
(elements and forms) mixing it up by chance, giving a success or failure
result. This communication and exchange process in architecture is equivalent
to the reproductive processes in Biological sciences. Where combining
the genetic complement of a previous generation of organisms produces
new organisms. Combining the typical set of a previous generation of buildings
produces new buildings. In both cases failures are not copied.
|
| |
|
As we know buildings are the basic units for architecture while for the
biological species is the organism. They are indeed evolved elements which
are product of exchanging processes. And as well they are unique combinations
of shared information found in theirs systems. Organism is to Species
as building is to Style.
|
|
Architecture lives and develops itself in a cultural environment while
biological environment holds living organisms. Both are formed by systems
that interact and communicate one to each other in a series of events.
Cultural environments are formed basically by institutions while the biological
environment are a group of interacting species.
|
| |
|
Natural selection in Biology occurs when the surviving potential of an
organism has not enough resources to survive in its environment. A failure
in its genetic combination will not be transmitted to the next generation.
In architecture since one building could become an imitative source for
many other buildings, its adaptable success or failure determine the degree
to which this particular combination could be repeated again.
|
| |
|
Finally emerge of new species and extinction in biological systems could
be compared to the emergence or collapse of a style in architecture.
|
| |
|
3.4_ Evolutionary Approach to Architecture
|
| |
|
A. From the Game to the scenario
|
| |
|
Since the researches of Turing and Newman, a great evolution has been
made in terms of computing. Actually they settle down the basis for the
development of today's computers. The display of architecture as an evolving
system in a computer allows us to see the evolution of the system as an
intelligent life. The games we analysed at the beginning of this essay
tried to reproduce an evolving conduct to represent in some times; life.
The evolving creatures and world game of Jason Spoffort shows the creation
of life in a computer. The amazing discoveries of Tom ray in its "Virus
Game", which we will see afterwards, defines a coming reality in
programming design.The scenario we create in a computer is the environment
where our artificial life will live. In the game of life we could create
a self-evolving architecture which will develop in what future will be.
|
| |
|
The Life game
|
| |
|
The Life Game was developed in the sixties by John Horton Conway. This
game simulated a life like behaviour. What he develops was a two dimensional,
two state cellular automatons on a square grid with simple transition
rules. These cells obey rules in an anthropomorphic language. The cells
could be death or alive. One cell will die if it had less than two cells
or more than three cells around it. A death cell could come back to life
if three living cells where around it.This game does not explain the rules
of evolution but it shows how does simple rules applied to a scenario
could produce complex emergent behaviour.
|
| |
|
The Universal Constructor
|
| |
|
John Frazer built this model of a self-organizing environment in 1990.
It features a baseboard which is called the "landscape" and
a series of cubes called "cells". These cells could be staked
vertically on specific locations on the landscape. Due to its cube form,
these cells could represent either structure or environment. Each of them
contains an integrated circuit that could communicate with the cells above
and below them as well as the location of the cells is communicated to
the controlling processor.
|
| |
|
Each unit has an identifying code which is displayed by eight light,
the system knows what and where is each part. The cells also display a
interactive red light which serves to communicate with anyone interacting
with the model. The state of each cell could be mapped to a graphics output
device where it is represented by colour or a geometrical transformation.
|
| |
|
B. Intelligent Buildings
|
| |
|
"If a building was conscious, then it could get depressed and choose
to terminate its existence by demolishing itself"
|
|
John-Paul Frazer
|
| |
|
We should understand intelligent building not only as a building whit
provision for information technology or a building where all the environmental,
security, lightning, energy saving, etc. systems are controlled by computers.
An intelligent building must be a building where all its systems have
a learning capability. These systems should be able to determine their
own arrangements for the user's benefit, more than the user himself. Intelligent
machines, as well as buildings computer systems should learn from experience
during use. This learning during time means that they should be capable
of evolving.
|
| |
|
The Generator Project
|
| |
|
In 1979 the headline of several newspapers used the term "intelligent
Building" to explain the machine that Julia and John Frazer had just
created, the Generator. This machine was the proposal for the Gilman Corporation.
It was a series of relocatable structures on a permanent grid of foundation
pads on a site in Florida. It was a computer program to organize the layout
of the site in response to the changing requirements. In addition it suggested
that a single chip microprocessor should be embedded in every component
of the building, to make it the controlling processor. This would have
created an intelligent building which controlled its own organization
in response to use. If not changed, the building would have become "bored"
and proposed alternative arrangement for evaluation, learning how to improve
its own organization on the basis of this experience. (John Frazer, Evolutionary
Architecture, 40).
|
| |
|
The Interactivator
|
| |
|
Dataspace
|
| |
|
Dataspace is the raw material of the informational possibility on which
both environment and seed can be mapped. It consists of points in space,
lines - the joining of points together, planes - the joining lines of
together, objects - the joining of planes together and theoretically the
fourth dimension - joining objects together.
|
|
The Dataspace is an infinite field of points in space, all 1 unit apart.
Each point in space has 12 immediate neighbours.
|
| |
|
The layout of the points in the Dataspace is based on the geometry of
13 spheres packed as closely as possible together with the centre of each
sphere being a point in Dataspace. It has fractal geometry as every part
is similar to the whole and it is scale less.
|
| |
|
Environments
|
| |
|
Environment is the information space in which seeds interact with the
world around them. When the right conditions exist they replicate themselves.
The environment, like the seed is mapped onto the points of Dataspace.
|
| |
|
There can be 1 to 12 spheres points around a central core point. This
gives a possibility of 4096 different configurations. All configurations
can be described with a 12 digit binary code, such as 0010X1100X11 where
the 0's show an un influencing position, the 1's an influencing position,
and the X an un decidable position. The un decidable position is influenced
(switched on or off) by the seeds requirements.
|
| |
|
Seeds
|
| |
|
A seed is a group of cells mapped onto the points of Dataspace. There
can be 1 to 12 cells grouped around a central core cell. This gives a
possibility of 4096 different configurations, which can be described with
a 12 digit binary code, such as 001001100011 where the 0's show an unoccupied
position and the 1's an occupied position.
|
| |
|
Growth
|
|
|
|
Rules for growth are embedded in the seed and influenced by interaction
with the environment. Complex behaviour can emerge from simple rules.
When the environmental conditions match the requirements of a seed it
becomes an active participant in the 'world'. It will breed with other
active seeds in the 'world', self-replicating themselves. This is done
by genetic crossover - each cell position of a seed is compared with the
cell positions in its breeding partner. If a cell exists in the same position
in both, their offspring will have a cell in that position too.
|
| |
|
If a cell does not exist in the same position in both, their offspring
will also have a vacant space at that position. Where cells are different
in each partner it is randomly decided whether the offspring will have
a cell there or not.
|
| |
|
Evolution
|
| |
|
Genetic algorithms determine the ability of certain seeds to improve
their performance in a particular environment. This is achieved by producing
successful offspring. Unsuccessful offspring will slowly die out. There
is an opportunity for mutations to occur. These are small random mistakes
that can occur when breeding.
|
| |
|
C. Artificial Evolution
|
| |
|
"Artificial evolution is the end of engineering's hegemony. Evolution
will take us beyond our ability to plan. Evolution will craft things we
can't. Evolution will make them move flawless than we can. And evolution
will maintain them as we can't"
|
| |
|
Tom ray created a tiny hand made creature called 80. It was called like
that because it had 80 bytes of code. When he released them in his computer
they started to reproduce rapidly fulfilling all his computer memory.
This was indeed a virus but which couldn't get out from his computer.
|
| |
|
At the beginning this 80 bytes bugs started to reproduce them, but after
some time they flipped a bit and became creatures of 79 or 81 bytes. Hours
later this evolution soup had evolved to nearly a hundred types of computer
viruses, all in a battle of survival in this computer world. "After
a year of reading computer manuals, I sat down and wrote code. In two
months the thing was running. And in the first two minutes of running
without a crash, I had evolving creatures".
|
| |
|
A single creature he programmed by hand populated the world Ray has created:
the 80. This creature reproduces itself by finding an empty RAM block
80 bytes big and then filling it with a copy of itself. But this was not
a coping machine. Ray added two features that make this world actually
evolved; this program occasionally scramble the digital bit during copying
and he assigned his creatures a priority tag for an executioner. In short
he introduced variation and death. This executioner was caller the "Reaper".
But the reaper passed over the rare mutants that worked allowing them
to keep on mutating and form new creatures. Ray also put his bugs an age
tag so they will die when becoming "older".
|
| |
|
After few hours running the program less bytes mutants achieved to survive
more than bigger bytes creatures because they needed less cycles to regenerate.
So creatures actually emerge to compete for computer cycles. But in the
process something very strange happened, he found a creature with only
45 very efficient bytes which overran all other creatures. Having a closer
look to this 45, he discovered that the 45 were indeed a parasite. This
new creature borrows the reproductive section from an 80 and copied itself.
But as if there were too many 45's then there will be less 80's to supply
their reproduction section.
|
| |
|
Ray started running his ecological experiment by supplying more 79's
to his world, which were immune to 45. But as the 79 keep on reproducing
a new parasite evolved, the 51. This parasite was indeed a genetic event
which transforms the 45 into a 51. New creatures evolved in this living
soup like hyperparasites, parasites that steal from other parasites. And
ray found creatures that surpassed the programming skills of human software
engineers.
|
| |
|
"I started with creatures 80 bytes large, because that was the best
I could get and I thought that with evolution they may come out with 75
bytes or so". Ray let the program run overnight and the next morning
he found the big surprised that there were creatures 22 bytes long. Actually
this was not a parasite it was a self-replicating creature. He couldn't
believe that a creature could replicate itself with only 22 instructions
without stealing them from any other creature. "What humans begins
can't Evolution can" (Kevin Kelly Out of Control).
|
| |
