Artware4: artists'texts

Benito Juarez - Rommel Romero
Michael Handsmeyer
Andy Lomas
Arturo Reategui - William Robinson
Umberto Roncoroni
Bogdan Soban

Benito Juarez - Rommel Romero

Emergent Lima is the city in emergency, this is how to show us more and more, amazing and rebel ... woven in the uncertainty... always against all attempt of control [planning]... an increasing city in complexity, whose levels of information exponentially increase in proportion of the number of interactions between their diferents components... interactions that generate structures [ complexes orders ] in constant fluctuation between the degradation and the daily renovation...
Emergent Lima is also the creative city, managing multiple mechanisms of urban survive, systems of cooperation in the middle of the shortage, flexible and stable systems in the middle of the convulsions and fluctuations, adaptive and evolutive in the middle of the degradation... Lima shows diverse endogenous phenomena to us… like the conglomerates of quarters, their vehicular chaos, the combis and the “dateros”, the multiple use of spaces public, the herds of autoorganizated street travelling workers…
What possibilities do we have to read it, to take part of it or to think about it? What kind of new tools means? What kind of changes of perception and action we have in face to us?
These dynamic ones challenge to us to observe our surroundings from new perspective. To relate to unstable enviroment and the uncertainty to us implies a humility act and lost of control on the reality turns out unreal the idea to understand (to make an optimal analysis) a determined context, since it implies to process an infinite amount of information), once again its invites to introduce us in the beast with the only torch that grants the certainty to us possible to undress its complexity from observing the simple laws of interaction between its components. Patterns of conduct that govern the dynamic ones of our context and that construct changing and evolutionary physical structures from interminable curls of feedback... to observe the tendencies without falling in the determinism.
New ways to approach the reality [city] also mean new ways to understand the use of the tools... toos are not neutral, each software imposes limits and rules to the user that determines the result. Functionalities that can be observed in different design software has been implemented by software designers and programmers to satisfy needs and standardized uses. But computers are not just a new tool, its a new medium, it generates new possibilities. This makes us to ask: which new things computers possibilities us as a new medium and which should be the function of the computer in architectonic design and urbanism?
To confront the reality with more information of the one than we can be able to absorb and to process implies we opened to the handling of emergent instruments... the tool, not like an agent passive, but - we can call- propositive... where the roll of the creator, composer change to something more similarity to an agriculturist who interacts with the climate, or to an escultor that controls and determines all the aspects of his work. Becoming that one who identifies the connections, reads the guidelines, recreates the surroundings and stimulates the interactions [but even so nothing guarantees its harvest...] of a work in constant construction.
The course EMERGENT ARCHITECTURES [ ] is generated like an exploratory space on the possibilities of the computing programs like applied tools to the reading and conception of new city-planning and architectonic projects, nourishing itself of theories that are emerging from the fields of Biology, artificial life and evolutionary computation; emergent systems, cellular fractals, robots, autoorganization, autopoiesis, etc. and explore on the ways in which they can be used like part of the cognitive process of neon urban designer.
From CHAOS to the COMPLEXITY. The enviroment’s perception as a goal of complex architecture
Its introduced to the complex conception of the reality identifying diverse isomorphes structures in multiple scales, substances or systems _ that relation can exist between spider fabrics, broken glasses, I deal urban, growth of districts, distribution of tables in coffee bars, tails in the tellers, growth of bubbles…..?
DECODING Searching of the emergent order in the complexity. The pattern’s architecture
Fundamental characteristics or properties are identified [ intrinsic qualitative and quantitative patterns ] in the interactions between components and that gives to origin to the complex structures [ attraction (15 u), repulsion (10 u), bifurcation (x3 * 0,5), friction (5 k), reproduction (x2)...].
RE-CODIFICATION Re-programming the genetic code. Trans-systémic architecture.
The behavior’s patrons to the programming language are translated [] From this digital reconstruction, software allows us to modify characters of configuration of the patterns being generated multiple variables of the same structure – multi formality and simultaneously allows us to such explore the application of principles that operate in a scale given for the neon conception architectures in other scales or systems – trans systems -.

Michael Handsmeyer

The philosophical discourse of determinism versus free will has existed for centuries. Recently this question has surfaced in connection with the interpretation of generative art. Those who are deemed determinists believe that all output of an artistic process is determined exclusively by the creator. In contrast, the proponents of a free will theory suggest agents created by artists can ultimately act autonomously to generate art.
Determinism in a theological context implies a creator that initializes all processes and foresees their outcome, which means he is omniscient. In generative art, the creator also controls all inputs and initializes all processes, yet may not be able to predict the precise form of output. This will more likely be the case the larger the number of components of a system and the number of iterations are, as this increases combinatorial variability. This lack of predictability may create an illusion of autopoiesis, yet all inputs and processes as well as any agents and their environments were determined uniquely by the artist.
Being convinced that all generative art is purely deterministic, one challenge to me is to develop transparent and simple processes that nevertheless produce complex, heterogeneous, and perhaps ultimately beautiful output. While more elaborate processes can lead to similar results, simple processes have the added advantage that their output is more predictable and can therefore be more easily refined through subsequent adjustments. In these simple processes, gradual changes in parameter values often lead to gradual changes in output, avoiding jumps and paradigm shifts.
The images I present in this exhibition are subdivisions of a square. A subdivision process is applied iteratively for eight generations. This process has the following attributes:
• Reduction in number of variable parameters – to as little as four division ratios
• Very simple input data – in this case a square
• Limited number of iterations with a pre-defined ending to the process
• No use of endogenous randomness, no conditionality, no Boolean logic
In its best case, this process offers an unanticipated heterogeneity, while providing means of control and refinement. It is this seeming paradox that I seek to develop.

La serie agregaciones
Andy Lomas

These studies are driven by a desire to explore the aesthetic and incredible intricacy of organic forms. The generated structures are created using a process of digitally simulated growth by aggregation.
Complex relations of symmetry and asymmetry exist on many levels. The simple rules used to generate the simulations are inherently symmetrical in nature but this symmetry is spontaneously broken by random processes in the growth algorithms. Radically different forms can be created by introducing small modifications to the generation rules, and biases to the ways particles flow before they deposit on the aggregated structure. The intricate sculptural shapes created have what appear to be large scale symmetries and similarities but when examined in detail it is apparent that no part is ever repeated and nothing on a detailed level is in fact symmetric.
The base algorithms used to generate the forms are variations on Diffusion Limited Aggregation. Different structures are produced by introducing small biases and changes to the rules for particle emission, motion and deposition. The growth like nature of the process, repeatedly aggregating on top of the currently deposited system, produces reinforcement of deviations caused by small forces applied to the undeposited particles as they randomly move. This means that small biases to the rules and conditions for growth can produce great changes to the finally created form.
The rendered structures are implicit surfaces composed of many millions of particles. Simulations can run from many days to weeks, with the final generated forms typically having between 30 and 50 million particle primitives. All the software used to simulate the structures and render the final images was written by the artist in Visual C++.

Arturo Reategui - William Robinson

Memescape is a computer-generated simulation, connecting a virtual system with people in a given physical space that generates diverse architectural forms and structures as a result of the interactions. The project explores the idea of architecture as a type of artificial life and how it evolves according to the dynamics that emerge in this artificial ecosystem.
It takes inspiration from the natural world and the constructive principles of living organisms and their processes. It exhibits characteristics found in biological systems such as metabolism, auto-reproduction and mutability. The creative power of natural evolution is emulated, exhibiting a virtual model that responds to the changing environment determined by the interactions between people and the virtual agents.
In this artificial ecosystem, the architectural forms that are generated are defined by trees of mutable function calls (the DNA, genetic code). This genetic information (memes) is propagated as ideas carried by the agents. It auto-replicates and mutates as a result of the interactions between agents and humans: when multiple agents from different original forms come into contact and the conditions are right, they will make a building. These agents are then destroyed and the building spawns new agents with mutations of its genetic code. The people can destroy agents, eliminating their genetic material from the gene pool. Using these methods, we explore a myriad of new architectural shapes and forms, we selectively evolve preferred genetic makeups to create forms through a complex adaptative system.

Generative Software
Umberto Roncoroni

The appeal of generative art begins in the emergent and dynamic quality of complex systems. Many digital artists argue that the self organization and the chaotic indeterminism that belongs to natural complexity can be digitally simulated and reproduced using techniques such as Cellular Automatas or L-Systems. On the contrary, I believe that every artificial model is deterministic: for each set of data, axioms and rules is always generated a same output, even if the artist tends to think otherwise due the great variety and complexity of the chaotic structures that recursion and random functions can actually produce.
True emergence, in digital generative art, can be obtained opening the parameters of the processes: it is the user or the programmer who, manipulating data and variables, actually brings the emergent qualities that algorithms don’t posses by themselves. For this reason, in my opinion, the true value of generative art does not consist in the infinite variety of forms or sounds that is possible to generate, but in the design of the process itself, that is software, which goal is to let its users take advantage of its ideas to enhance their creativity. Thus, generative art beauty is the same that the beauty of the man/machine interaction, that is, it seems to me, something deeper than a new artistic flavour.
Obviously a generative tool needs constitutively to deal with its software’ knowledge and information, and to find a strategy to transmit and share them. Inside the framework of this program, because the idea is that the user/public should design and implements his generative tools rather than use something already developed by others, the interface is the critical component of the system, because can enhance or inhibit creativity.
For this reason I tried to design an information architecture optimized to facilitate the access to software’ knowledge, something that include libraries, functions, data, algorithms and controls and all the stuff that allows the user to create, implement and test his own languages and grammars and to edit and modify the appropriate rules to analyze, improve or combine them with ease and speed. Linguistic and creative liberty, after all, is the true goal of this generative application.

Teaching generative art
A didactic experience at the FIA-USMP, Lima, Perú. 2006-2007
Umberto Roncoroni

Because of the greatly complex and interdisciplinary contexts involved, it is really difficult to properly design and implement a didactic methodology for generative art. Furthermore, there are very few pedagogic experiments, even in the international domain, that can be capitalized (maybe with the exception of Bauhaus and the Italian artist and designer Bruno Munari, to which workshop To Play with Art I had the good luck of being invited).
To begin with, the main question that is always asked by all the participants at generative art classes or lectures is about its practical utility and applications. As an answer, the works of Gaudì or Gehry are often proposed as specimen or examples, and usually the importance of digital design, manufacturing, the properties of new hi tech materials that allows the development and construction of new typologies of organic and fractal structures is stressed well enough.
But generative design is something more than just a design method or technology to build more or less automatically weird and fantastic forms: rather, it is a new way of thinking art, design and creativity, something that can definitely change the way we can be architects, artists or designers. In this sense, it is never stressed enough that a generative process is not always committed to build some finished object in order to be considered useful and efficient, because it is more properly an heuristic tool, a try and catch procedure, that allows to explore the different layers of design with an open minded and sensitive approach, say, capable to listen to users, to social and cultural contexts and to ecological demands in terms of materials and resources. All these models, elements and factors form the information system that every generative approach should consider and evaluate, because it is the true and visible new value that the whole systems generates from its separated elements and parts (a sort of business intelligence or decision system applied to creativity).
So to teach generative art, in my experience, doesn’t mean to train the student with some specific software or technology, but consists in the transmission of knowledge, theories and scientific models and techniques (Cellular Automatas, Generative Grammars, Biology, etc.) and to teach how to build a network of these topics inside an original and solid framework that each student must design and build by himself, following the specific needs of his project.
Referring to the exhibited artworks, developed by almost graduated students as the final work of the Generative Design course, the main goal was to explore the significance and the properties of procedural design, the artistic use of programming, the aesthetic meaning of algorithms and rules and to experiment with the possibilities of the generative approach as a whole.
Linking technology, data, information, knowledge and theoretical issues, students have tried to develop an integrated process to design a generative skyscraper. At the beginning the computational tasks and the steps of the generative algorithm have been developed by hand (with the help of 3DMax and POVRay to render the designs) in order to understand more deeply the actions performed at every single step. Certainly the whole process could be completely automatic, but at the loss of the direct contact with the nitty-gritty, and of the understanding of the true meaning of what is actually happening, in other words, with the loss of the advantages of generative design over a more traditional approach. Besides, as every computer scientist knows, any block of code should be explored by hand step by step… it is only after this hard and tedious hand work that is possible to implement a process inside an interactive application, without missing the possibilities that only well chosen parameters can generate. The artist closes the generative loop selecting the best process’ instances, eventually modifying with other tools the results and distributing or sharing the designs (a task that is much more easy when designs are in digital form of algorithms or data bases) to let others contribute with new ideas, changes or improvements. This topic, the distributed, hipertextual and collaborative framework, is another essential characteristic of generative design that is very important to highlight in the classroom.

Bogdan Soban.
My art is entirely computer generated and based exclusively on aesthetic ability of mathematics. All my works are results of the autonomous processes supported by computer programs developed by myself using Visual Basic. The method is known under the name “generative art”. The algorithms are my creative challenge and the emerging images my artist’s emotion. The concept was born more than twenty years ago when working in the area of informatics I often have been thought about the other possibilities of information technology having in mention the creative abilities of machine.
It was a daring idea until I didn’t begin with experiments developing my own programs that have created simple images. I started with geometric abstraction using pragmatic programming approach. Introducing algorithmic concept and Visual Basic programming language the pictures as the results of autonomous generative process were much more interesting and perfect from the artistic point of view. My basic concept is not to interfere with the machine creative process. My author’s creation is program code where I apply a huge number of mathematical formula and expressions with not to know in advance neither the type nor the style of the image could be created. The first image appearance of the new algorithm could be the real surprise or disappointment. After different adaptation of the code I can improve the final result but all the same all images created with my programs are absolutely unpredictable and not repeatable.
Naturally that all created images are not good from my personal aesthetic criteria. The selection of the created images is my second important role in the whole generative approach in digital art. I believe in the aesthetic abilities of mathematics. More complex formulas are used better are results. Combining fractal concept with my previous mathematical approach was very good idea to obtain more complex and harmonized pictures. From my personal point of view the pure digital art has to be created using proper program code. Using Photoshop or other graphic software is nearly the same as using colour palette, brush and canvas and the computer plays the role of the wonderful tool. I agree that could be a silly idea to treat a computer as a creative partner. It sounds unbelievable but it works. A good question is now where or when I feel the artist’s emotion as the most important part of human creativity. I feel it in the magic moment when new image begin to appear on the screen. The experience is much more intensive in the case of the first run of the new algorithm.

Artworks description
Presented artworks are result of a recent developed generative program based on two principal approaches: image decomposition, integration and immersion into a new created image in order to extract interesting details. The input into the process is an existing image or photo and the output is a new image with a very different similarity level in comparison with the input. The decomposition process is realized by the program algorithm, which reads an existing image and breaks it in elementary pieces – so called pixels.
The next step is the integration in new image, which I usually call image-mother. Decomposition and integration algorithms are based on multilevel deformation of the Mandelbrot calculus. The multilevel deformation algorithm has the task to partially eliminate the main fractal property – the self-similarity. From the same input the program can creates an immense number of different images-mother. Practically the input image plays the role of outside coloring palette. This procedure is absolutely autonomous without any kind of my intervention except the selection of input image - coloring palette - and the selection of the result. Having a good example of image-mother, the program proceeds with the immersion into image – mother. This phase of the process is partially interactive because the selection of exploring area is realized from the outside. Clicking on the interesting locations the program opens a new view of image located somewhere in its depth. Saving selected detail means to create an image-child.
By changing of the mouse-location and magnification parameter is possible to explore the third dimension of the image. The operation is not a simple zooming because the inclusion of the magnification parameter into the deforming algorithm causes the elimination of fractal self-similarity. The approach gives a real feeling of “traveling” through the image as a virtual world and making “photos” of interesting and beautiful details. The next trip through the same landscape opens other views and gives an occasion to make new photos. Exploring the same image-mother and creating images-child is possible until the program is alive respectively until it is running. The exit from program means to loose for ever the image-mother and the next start offers a new “landscape” to explore.