On G+, you could be invited or apply to the private Create community. The community was designed to help creators to meet, network and learn from each other. There was also the so-called discussion of the week, which was a pinned post of the community, to share around a work, experience or anything you thought worthy. I owned one of these discussions just before the sunset became a reality. Now that G+ is closing, I’ve decided to share an edited version of this discussion here in public and with a possibly broader audience. Who knows, you may be interested as well, and could share your thoughts.

Opening

I titled my discussion of the week Prophetess of Numbers. The opening post read like this: During the 50s-60s, the Icelandic named it tölva, the French ordinateur, the Hungarians számítógép like most of the world, simply re-used the English word computer. Once named, the world could start to assimilate, tame and make useful an object that didn’t exist a decade ago. Today, computers are ubiquitous and on the verge of disaggregating, becoming an integral part of our societies’ fabric.
I fell in love with computers at first sight in 1981. Since I never stopped doing stuff with and around them. My motto being: give me a computer, electric power, and I can make anything! A maker before time if you wish, certainly a nerd forever. But, I always considered computers as an Art form, way more than just a tool. Moreover, they are undeniably becoming key instruments to help us create and invent.
I know I am biased, and I imagine a few of my fellow creators shouting “heresy, witchcraft, the stake!”. OK, I hope not really. What I do expect though, is that you will be willing to share how you do use computers to create. Obviously, we all require one to share our creations on G+, but what about during the inception, the realization?

IMG_2636

This post triggered a few exciting comments and exchanges. As I mentioned, this is a redacted version of the original exchanges; therefore, I will summarize the most interesting inputs and will provide verbatim the comments/answers I’ve submitted (without the names, of course). [I = Input, A = Answer/comment].

Computers are tools

I: Computers are just tools with no creativity in them.

A: Thank you for sharing. I hear you and completely understand your view. Many of my colleagues and friends share your utilitarian view – and use – of the computer. And it is because of my bold statement (computers as an Art form, allowing new forms of Art) that I am very curious to hear from you guys.

I: Everything you can do with a computer, you can do without it. It may be harder, longer, but it is not impossible.

A: Thank you for sharing this interesting use of the computer in your creations. Very cool! I understand your struggling as your example makes a lot of sense. You could indeed spend a substantial amount of time to do similar explorations without a computer. But would you do that? What I am trying to ask is, would your Art be radically different without the computer? I would imagine that not. You would probably find another clever way to reach the same objectives. In this case, I consider your algorithm itself as being artistic. Please note that I am not claiming that a computer algorithm does or could do art intrinsically (that’s another fascinating topic :)). Would you consider your algorithm as being art?

Open-it up

To foster more discussion, and steer away from the computers are tools, I added some personal experience in the following post: When I was much younger, I had this fascination for computer graphics. In the late 80s – early 90s, image synthesis was in its infancy. But the same way early video games with poor graphics & gameplays fueled our imagination, early 3D left a lot of space to explore to the dreamers. I had this illusion that with the right models and enough computing power, we could simulate the entire universe, and who knows, even predict the future if we were fast enough! These were the time some of you may remember when everything was a teapot :). I was a freshman, and I was looking for an internship when my linear algebra professor Jean Zeitoun offered me to go to CIMA (Centre d’Informatique et de Méthodologie en Architecture). Think about it as a thinktank, focusing on architecture & urbanism researches as well as an AI lab (not the deep learning stuff, but the interdisciplinary firework). I was tasked to add Boolean 3D operators to the geometry engine of IKO, the in-house brewed 3D software! The software suite was the brainchild of Michel Bret, an artist, and mathematician who was focusing at the time on designing graphics software for artists. The place was humming of artists like Sabine Porada an architect, painter, and teacher. I am afraid that I didn’t do very well, barely met the expectations composing logically 2D objects – yeah, I missed a dimension there :). That’s probably why many years later, Michel didn’t pick me as a Ph.D. student in Arts and Technology of Images, but instead, I landed in the AI lab of Patrick Greussay under the guidance of late Jean Méhat. A radical shift in appearances that I really don’t regret. Indeed, those were the years where I discovered the algorithmic beauty of nature as well as the beauty of algorithms.

Try again

Unfortunately, it didn’t trigger many reactions (I guess that the G+ shutdown was a hotter topic). Well, that’s when you don’t give up and try again. I, therefore, followed up with this post: How do you create a tree … with a computer? We could fire-up our etch-a-sketch (a.k.a. a graphics tablet) and draw one. I suppose we could also snap a picture and Photoshop the heck of it. If I had to do it, I would use a beautiful … algorithm! A very simple one named L-system. The L stands for Lindenmayer, the Hungarian biologist who introduced the technique in 1968. It consists in re-writing a string recursively, applying a set of simple substitution rules. He used his system to describe the growth of plants. Note that L-systems were successfully used to reproduce the morphogenesis of many other organisms. For the linguists, L-systems are semi-true grammars (according to Noam Chomsky’s classification). For the Euclidian geometry aficionados, they have a lineage to fractal geometry. For us, let them be an algorithm to create vegetal growth. Suppose we use two letters A and B (variables) and begin with a string of one character, let’s say A. Our rules could be: replace As with AB and replace Bs with A. Applying these basic rules recursively gives: A, AB, ABA, ABAAB, ABAABABA and so on. We have just simulated the growth of algae! This is also the first system Lindenmayer created. If you want to spice up your creations with a dash of L-systems, you may want to read the Algorithmic Beauty of Plants by Przemyslaw Prusinkiewicz and Aristid Lindenmayer.

algo

As a result, few testimonies were shared, on how creators are using their computers, SW, etc. To these, I answered as: your everyday use of computers in your creative activities is a brilliant testimony to the progress they have made to become useful to everyone! I remember a time when you had to key-in, via rows of switches on the machine’s front panel, machine language programs byte after byte (the Altair 8800 to name it). Even for a so-called nerd, this was a tedious task. Today, computers are everywhere as you shared with us. In fact, they are disaggregating right now on every level: hardware and software. Think about it as when you turn your light switch or faucet on, electricity and water flows …. what else would we expect? Computations will follow the same trajectory. Few work already on dust computing, where myriads of dust particle-sized computers and sensors are building up vast networks. I am wondering what our creative process will be in such times? Thank you so much for your comment!

Bring-on the algorithms

I had to bring in the big guns to spark some good discussions. I was ready to get the heat, and brought-in the algorithms for generative designs as follows: Generative algorithms are often used in Art. For example, a cellular automaton can produce interesting geometrical patterns. You may know them in the form of Conway’s Game of Life. One could easily discard them as “meh, a bizarre wallpaper”. But as creators, we can see beyond and maybe recognize biological patterns, such as the ones displayed by seashells (Conus textile for example). More specifically, in 1984, Stephen Wolfram (the father of Mathematica) made it popular via his rule #30. But even earlier, in 1973, Herman & Liu used cellular models to explain these pattern appearances in bivalves. Hans Meinhardt presented an excellent survey of such technics in The Algorithmic Beauty of Sea Shells published in 1995. These models are principally based upon activators and inhibitors (as well as a lot of derivatives). And of course, Alan Turing published in 1952 his seminal paper The Chemical Basis of Morphogenesis, modeling the formation of natural patterns in animals such as spots or stripes. All these approaches, by the way, link into a computer science branch known as Artificial Life.

IMG_6250

I’ve spent some time creating algorithms inspired by multi-cellular organism’s morphogenesis. My postulate at the time was that an organism is the expression of a program, the genetic code and that there must be a lot to learn from biology and physics involved in morphogenesis. Especially for massively parallel programming (we are talking trillions of cells). And arguably, these programs work pretty well, as we can witness it every day. We, the expression of these codes, can even create Art!

Let’s take a very basic example. Suppose you want to compute a three-dimensional function you know nothing about, using a very sparse two-dimensional grid, where a compute cells exist at each row-column intersection (calculate z knowing (x,y)). The problem is that in some areas the function may be monotonic, and may oscillate in other ones, while you have limited resources. Therefore, increasing the entire grid’s resolution is not a viable solution. So how do you do it? Suppose these cells are in gradients of signals (imagine chemicals in a solvent with different concentrations depending on where you may measure it), and one of them stimulates the growth of the cells. When a cell grows enough, it splits, and the two new cells migrate away from the original position in any direction (need space for the new cells). By doing so, each cell takes away the same genome (the program), the previous values of z. What the genome does when it is expressed, is to compute the new z’ for the new (x’,y’). It then differentiates z’ and z, and extrudes growth signals proportionally. So, the greater the variation of z, the higher the growth signal’s concentration in the vicinity. Which in turn, increases in fine the cellular multiplication rate and the computation’s resolution in that area. At the opposite, when z variations are low, the organism (the computational grid) will not use its precious resources to grow more cells. Quite simply, isn’t-it. What I like – and try to achieve in my everyday activities –, is to design systems that are distributed and without central controls or registrars. Instead, they are self-regulating systems with researched qualities emerging from the cooperation – without any form of intent whatsoever – between the agents.

Are anyone’s creations the direct or indirect result of a generative process/algorithms? I would love to hear and learn from your experiences!

Later on, I followed-up with some pointers: If you are interested in generative design, but at the same time, all my technological mumbo jumbo is not palatable to you, maybe Processing can help. This open-source sketchbook software was designed for creators in mind. It is using a straightforward programming language (Java) and is focused on graphic arts and has grown a lot into dynamic performances. There is a lot of material on the website (processing.org), and you can buy good books on the topic. I can recommend Generative Design by Hartmut Bohnacker, Benedikt Groß and Julia Laub (Princeton Architectural Press – New York). But of course, your experimentations are the best way to learn and to create! By the way, do you know the creative work of John Maeda? This MIT Media Lab professor and designer – quite mainstream – did some interesting researches around generative design that may inspire you as well (Design By Numbers, Creative Code, etc.).

I: One member of the community made a thoughtful linkage into Cajal’s The Beautiful Brain exhibition at the MIT Museum.

A: Thank you for this beautiful linkage into biology. I cannot resist remembering the way Alan Turing described the brain as a bowl of cold porridge. It is extremely hard for me to not be fascinated by the object itself and all the creations it was and is able to achieve. The tour de force is being able to study itself. In computer science, students are surprised when the first time we review functions calling themselves recursively, and how it makes you think of a particular category of problems in a different way. But here, it is an organ exploring its own inner and most profound mechanisms! It, of course, starts with a descriptive phase, then building models on top of it, and more importantly letting science do its thing (failing, new hypothesis, experiment, learn and fail again). When you look at today’s overhyped AI claims, they all originate from an over-simplistic model of a neuron (1943, McCulloch & Pitts). I wish I could still be around the day we will understand how trillions of cells and ever-evolving interconnects leads to a conscious being capable of creating… Thank you for your comment!

cajal-webpage

Winding down

The topic went out of steam, and we reached the end of the week, that I concluded as: as the weekends, I want to thank you for your contribution to the discussion of the week on the use of computers in our creative processes. Merci, and have a great WE!

Feel free to leave feedback on any of these fascinating topics, and I would love to know your views!