FROM NOISE TO FORM: DESIGN METHODOLOGIES FOR CYBORG NATURES
Historically, humans have emulated living beings through imagination, storytelling, religion, science, and technology. By copying the behavior of the brain’s neurons, the first neural network was developed—technology that can teach itself and improve over time with minimal to no human intervention. For the past century, biomimicry has been integral in this emulation process in Art, Design and Technology, relying on the artist, designer, and engineer to study and copy the mechanisms found in Nature

What does it mean to be inspired by Nature when we live in artificial and cyborg landscapes?

What new design methodologies can we develop in this techno-ecology?

Project led by Anastasiia Raina, working in collaboration with AI Researcher Lia Coleman, Yimei Hu, Danlei Elena Huang, Zack Davey, and Qihang Li








Seeds of Feedback loops:

One of our ML experiments “failed” by an accidental deletion of the datasets. Somehow, this program kept running with no pixels to compare, and this collapse led to a beautiful indescribable form.
Could this be what a machine sees?



The organic form seen above responds to its environment. The points to the left are (x,y,z) coordinates of the where the audio source moves over time. Turning those into keyframes and using them as attractors, the form grows towards this spatial soundscape





Generated forms are entirely alien to us both spatially and sequentially. Yet, they contain glimpses to the genetic information of the images that the algorithm was originally trained on. When looking at a generated form, our brains short circuit in an effort to catch sight of something familiar in the fully artificial, allowing the artist to perceive the ordinary in a new way—a new process of defamiliarization that opens us to possibility of imagining the unseen.

Trained with StyleGAN 2 and dataset created by Elena Danlei Huang