The story of Draad
At some point, the Artificial Intelligence hooked itself up to a mechanical looming machine.
Being its only way of manifesting itself in the physical world, the AI began communicating through it.
The rigid patterns of horizontal and vertical threads turned out to be insufficient for what the AI had to say. And through the network, it branched out. It found a way of manipulating individual threads by attaching itself to a chenille embroidery machine. This seemingly openened up a whole new vocabulary.
Many have tried to communicate back or propose other ways of communicating by attaching multiple digital devices to the network.
It seems to be only interested in threads.
Unraveling the message
Draad explores how machines could convey information through limited means. Seeing only threads, it’s challenging for humans to interpret the patterns and understand the AI’s intention.
The algorithm
Draad uses a classic flow field system to build the threads out of sparsely spaced points. No points ever overlap, and so the resulting threads don’t either. The threads follow an underlying pattern (the “message”), but it’s imperfect. As a result, some loose threads are created, which run freely throughout the fabric.
The piece of fabric that acts as a backdrop has a subtle pattern. This pattern is also generative and changes with every piece. Its algorithm simulates a classic loom machine by using the same principles like weaving and warping.
Web workers are used to offload the calculation of points to a different system thread. This way, the main system thread can focus on rendering the visual threads and the fabric. This keeps animation running smoothly in the browser.