Punch Card Collective
Who we are
We are Annie Blazejack, Peter Boyer, Geddes Levenson, and Alex Rubinsteyn.
We are two artists and two computer scientists. We are two pairs of husband and wife. We are long-time friends. We are a kind of chosen family. Collectively, our interests span art, art history, textiles, biology, engineering, computer science, graphics, and machine learning. We wanted to work together on a project that would combine many of our interests and skills.
Crypto is a challenge to use, so it’s hard for the average artist to make an NFT. Conversely, many programmers aren’t knowledgeable about art history and art theory. Working as a collaborative allows us to close these gaps.
We decided that we wanted to make an artwork that was very aware of its relationship to the rest of contemporary art and art history. We arrived at weaving because weaving is the ancient mother of computers. Artists called weavers invented the first computers in order to facilitate the production of more complex fabrics. It seemed reasonable (even essential) that computers should return to the problem of designing cloth.
Making art as a collaborative
Difficulties
If one artist can make one artwork in one day, how many days will it take four artists to make an artwork together? That’s a trick question, the answer is forever! Reaching consensus about the innumerable sub-problems that go into making an artwork is really really slow.
Other major difficulties we encountered were (a) We didn’t want to make ‘art by committee’ like some half-baked Hollywood movie script, and (b) We wanted to continue taking risks throughout the artmaking process. If each change is carefully negotiated and hard-won, it can become exhausting to make big changes later in the project.
Collaboration techniques
We found that dividing into pairs (artist + coder) helped us work more efficiently. (We are two artists and two coders, so we can split into four possible pairs of artist + coder.) Working in pairs was most useful once we had already built the guts of the project, especially as we designed color schemes and honed trait combinations.
We also had frequent review sessions in which the four of us gathered around our largest screen to look at the state of the project. We asked ourselves “What works? What doesn’t work?” and agree upon next steps. This was an iterative process, with more-or-less weekly reviews.
We created separate controls for our variable traits so that we could look at each trait in isolation and multiple traits in combination. Encoding complicated combinations of ‘this trait can occur with this but never that and with the other thing only very rarely’ is a kind of individual brushwork, a glimpse of the artist’s hand.
We settled into a ‘Yes, And’ approach to decision making. Luckily for us, the world of weaving is practically limitless. There is room within our problem space for innumerable solutions.
We divided the project into four ‘archetypes,’ four different tunings of the algorithm, so that each one of us could have creative control of a portion of the project. Annie got to perfect Hypnos, Peter worked on Maze, Alex designed Control, and Geddes captained Mythos. Dividing the project solved the problem of ‘art by committee,’ because it allowed each of us to have a decisive individual voice.