Architecture & Style is a project by Stanislas Chaillou that makes use of machine learning algorithms to explore the possibilities of generative design in the creation of architectural floor plans. Chaillou created a generation methodology for the plans using Generative Adversarial Neural Networks (GANs) that would learn from various precedents and then be able to compose adaptive floor plans based on different styles. The project shows clear artistic sensibilities as the program is learning from established architectural styles such as Baroque, Row House, Victorian Suburban House, & Manhattan Unit in order to then generate its own designs influenced by these styles. I found the project to be very interesting as it demonstrates how architectural practice can make use of new advancements in artificial intelligence.
Unnumbered Sparks is an interactive community artwork made by Janet Echelman and Aaron Koblin. The artwork is a 745 feet sculpture in the sky controlled by visitors in real time using their smartphones. Unnumbered Sparks is essentially a large canvas projecting a Google Chrome web browser that is over 10 million pixels in size. The project was coded using a programming language called Go and the visuals were rendered with WebGL. WebSockets was used to connect the mobile devices to the sculpture.
Unnumbered Sparks Janet Echelman and Aaron Koblin.
This project grabbed my interest because it is an interactive piece of artwork that requires the cooperation of a large group of people. Aaron Koblin often creates interactive art and programs that utilizes data input from multiple people at a time. His works like Bicycle Built For Two Thousand, 10,000 Cents, and New York Talk Exchange are examples of artworks created with the collaboration of many people in real time. Interactive artwork maximizes potential when many people interact with it at the same time, which is why I admire this project.
The project I admire is Silk created by Yuri Vishnevsky in 2012 (with music and sounds by Mat Jarvis). It is an interactive introduction to generative art, in that it allows users of the website to create their own images. The website uses the user’s mouse coordinates to create lines, which is then dictated by the program’s algorithm to manifest in a fractal-like manner. I’ve often used this website to relax in the past, as it is therapeutic to see the colorful lines of light dance across the page.
An image I created using only “black” in Silk, with the rotational symmetry set to “6”.
From what I can understand, the code takes the mouse’s coordinates and then uses them to generate lines from one point to the next. Then, that line is being duplicated and angled in a way that we get “silk” threads coming off of each main line. This produces a tangled effect similar to that of a silk web. The creator’s artistic sensibilities definitely come off in their algorithm, as there were specific choices that were made to create the weave-like effect of dancing light. This website would have given off a very different feeling if the creator had chosen to just use simple straight lines. Furthermore, the addition of rotational symmetry in the algorithm allows users to experience surprising effects as they move their mouse across the screen. Shapes that seem like stars, webs, hexagons, and flowers are prone to appearing as users explore.
Another image created using Silk, this time with different colors being used.
“Artificial Arcadia,” created by Fragmentin in collaboration with KOSMOS architects in 2019, utilizes a 3D spatial contraption to raise questions surrounding how the natural, digital, and artificial shapes contemporary landscapes. The installation features white blankets propped up by poles—reminiscent of famous Swiss mountains—and changes and falls as the mountains themselves begin to melt. What I enjoyed about this piece was how viewers were able to interact with it and walk through the space; in my experience, incorporating space into artwork creates an immersive experience and a more profound impression on the viewer. From what I can guess, the program must calculate the heights of each pole at a given time, so perhaps there’s a section of code devoted to each pole. Also, the program should be making these calculations every few minutes, so there must be a part of the code that regulates the time frames in which these calculations are made and the code repeats itself in a loop. I think that with their narrative surrounding landscapes, the artists rightfully chose to program something that involved 3D space (supported with a 2D graphic) rather than creating an algorithm for 2D work; I really admire how they explored the intersection of coding and art in order to make an important statement about climate change.
This project by generative artist Kate Compton is called Flowers. It is an interactive genetic algorithm that was mainly created using JavaScript, as well as other algorithms to draw the flowers. I find this playable interesting because it incorporates a sense of complete randomness but also makes use of similar and repetitive geometrical shapes.
A quick, non-interactive video depictingFlowers
In this project, the flowers have a 26-dimensional parametric ”genotype”, which generates L-system flowers in a continuous function. This continuity helped in creating a smoothly morphing animation. However, it is not fully continuous since the leaves and petals have been given integer values. This means that when the flowers animate from one form to another, there is some popping as leaves and petals appear.
A special characteristic of this project is that it uses a custom parametric generator, which controls color and shape. Initially, a variety of flowers are shown and the user can click on their favorite one. The selected “”favorite” flower spawns several closely-related and similar-looking plants and the user can continue to evolve their flower.
Initial variety of flowers
Variety after choosing the right-most flower
A link to the interactive version of Flowers can be found here.
This is a generative art piece called The Historical Spiral by designer Ramiro Perez. This project stood out to me because of its presence in three-dimensional space. Perez’s ability to introduce light and shadow to an object that is so evidently mathematically made is mesmerizing to look at and in many ways, makes me ponder how he did it. It seems like Perez used fractals to create the spiral form, but what makes me go “wow” are the micro details that are in the form. If one compares ring by ring, he or she can see that the orientation alternates from laying down to upright. Within each spiral, you can see mini spirals as well, which shows that scaling in and scaling out really just reveals the same mathematical pattern at a different size.For me, the attention to detail is what convinces me that much tought went into the design of this structure.
Kate Compton is a self-described “crafter of twitching generative bots.” Her work perfectly encapsulates concepts of generative art; it is bright, geometric, whimsical, and most importantly, random. A great example I found of her work is “Flowers,” a generative art piece that randomly spawns a sequence of graphically-generated flowers that sway in the wind. If you click “replant,” you’re able to regenerate the sequence of flowers, and the “evolve” button causes the regeneration to occur rapidly.
I thought that this was a fascinating example of generative art because you can almost see the principles in action. According to Free Code Camp, some of the core foundations of generative art is that:
It’s random
It’s based on algorithms
It’s geometric in nature
The “replant” and “evolve” buttons make it clear that there’s an element of randomness to Compton’s work. Further, she’s very upfront in her bio that she works mainly with JavaScript (and very possibly, p5.js!). Thus, it’s likely that Compton wrote an algorithm in JavaScript and allowed it to reign freely over her art. Finally, the graphical style of the flowers is clearly geometric and shape-driven, and this works wonderfully with her work because of how flowy and limber the flowers feel in the wind.
It’s clear that Compton’s artistic capabilities are embedded in her algorithm because of how she utilizes color palettes and combinations as part of the “random” element. Every time the sequence of flowers evolves, so does the color palette, but Compton’s clear knowledge of color theory keeps the colors vivid, bright, and fun. My guess on how the algorithm functions is that with every mouseClick on “replant,” a random color is selected for each of the different flowers. Not only that, but a new random sequence of shapes is generated as well. Then, for the “evolve” buttons, the algorithm emulates a continuous sequence of mouseClicks to keep the colors and shapes evolving and changing. Kate doesn’t explain the secrets to her work, but it’s that element of randomness and surprise that keeps fans coming back for more.
Kate explaining how she creates generative art with Javascript at the JSConf in Iceland, 2018.
![[OLD FALL 2019] 15-104 • Introduction to Computing for Creative Practice](https://courses.ideate.cmu.edu/15-104/f2019/wp-content/uploads/2020/08/stop-banner.png)
