[ 2024-05 ]
{Originally published as part of my RISD thesis project One of One of a Kind}
Speaking about craft can be difficult. It’s one of those English words like “beauty” or “value” that can hold many meanings, which is a benefit in terms of its depth, but can make it hard to discuss with clarity. Among any group of people off the street it may conjure images as diverse as popsicle stick puppets, Martha Stewart, tie-dye shirts, or Brancusi sculptures. The curator and craft scholar Glenn Adamson poses craft as “the application of skilled making to the world around us”. A definition broad enough to hold all these varied manifestations, and yet precise enough to be useful. The word “application” implies an agent, a person capable of not only making but also perceiving where and how the results of their making belong. “Skilled making” is descriptive of the type of work we normally associate with craft, and yet open enough to include all manner of tools, materials, and techniques used to do so. And “the world around us” expresses a rootedness; to place, to context, to community, and opens up craft’s cultural, environmental, and social implications.
Of the three aspects just mentioned, it’s the idea of “skilled making” which seems to cause the most confusion. What constitutes skill? Who decides where the threshold is between crude and well-made? What kinds of tools can be used before you cross the line from making to manufacturing?
A popular understanding of craft is that it produces things by hand. This immediately becomes problematic, as only a few objects can really be said to be made only by hand — crude pottery, baskets. We can broaden the scope and consider the use of hand tools, but what does that really mean? Tools that don’t use power? Most tools considered hand tools have electrified versions: potter’s wheels, drills, hand planes. Tools that allow the hand to freely influence the design without any sort of guide or jig? Even a chisel, as it begins to carve, creates its own jig with the material, influencing the possible forms and surface finish.
It’s this ambiguity that prompted the writing of industrial designer and craftsperson David Pye’s The Nature and Art of Workmanship. In it, he makes a distinction between what he calls “the workmanship of risk” and “the workmanship of certainty” (he admits these are awkward terms). In the workmanship of risk, “the quality of the result is not predetermined, but depends on the judgment, care and dexterity which the maker exercises as he works”. The base example of this would be writing with a pen. If you asked me to hand write a letter for you, you would provide the content but the visual design of the letter would quite literally be in my hands — and my choices in the process of scribing for you could either enhance or degrade your original intentions for the letter. In the workmanship of certainty, on the other hand, the quality of the outcome is predetermined — there’s no risk of the design being ruined during the making process. The follow-on example to the letter would be modern printing.
The Luddites saw the Jacquard loom — that was made to quite literally imitate their skilled hand movements — as an affront to their agency in the weaving process. John Ruskin, the most influential writer associated with the Arts & Crafts movement, highlights the crudeness and imperfection of Gothic ornamentation as a sign of the creative freedom of the craftspeople. It’s this risk of crudeness, of imperfection, of spontaneous variety that at the same time is almost universally admired in the artifacts of the past and considered unacceptable when bringing an industrial product to market.
With this framework of creative risk, popular ideas about craftsmanship are freed from romantic notions of only working with particular tools, in particular mediums, or with particular techniques, and instead something deeper and ultimately more important is foregrounded. The seed of this entire exploration is the idea that generative design, when used in this spirit, is a new kind of creative risk, and that embracing it can be an avenue for a kind of aesthetic and cultural richness that echoes the past and yet embraces the future.
What I’m referring to as generative design goes by many names (another way in which it parallels craft); procedural, algorithmic, computational, parametric. These are all jargon that refer to the use of the algorithm — a set of instructions formalized in code — as a medium, with each discipline from engineering to architecture to art using their own nomenclature. Casey Reas gives the essential definition, speaking from the context of generative art: “the artist makes the system and the system makes the art”.
There are various approaches to this as well, that involve more or less risk by the designer. In the 90’s, the hot trend in product design was “mass customization”, which is a corporate version — letting the customer adjust the parameters of a system for, say, the design of their razor handle (Gillette did this), allowing them to get a custom product that they had some small part in designing. While giving up tight control on a part of the design does technically involve risk, it’s a watered-down version that was mostly used as a marketing tactic.
Another approach is the one that has been increasingly used in the generative art community over the last few years, which artist Tyler Hobbs called “long-form generative art” in a 2021 essay. This approach usually involves creating a generative algorithm and writing it to a blockchain, making it permanent, unchangeable, and also public — anyone with an internet connection can view and interact with it. Collectors run the script and mint outputs as NFTs which are transferred directly to them with no involvement from the artist. The artist completely delegates the curation step to their community of collectors; said another way, they accept the risk of not controlling what the final collection looks like. In choosing to keep the output of a generative system fully open-ended, expressed through a real commitment like publishing the code on a blockchain, and allowing emergent community curation over the internet, the artist and collectors together are engaging in a new, digital version of David Pye’s workmanship of risk.
Engaging in this type of long-form design has implications on how an algorithm is written. It means your goal shifts from producing a single good output to building a system that produces many good outputs — say hundreds of thousands or hundreds of millions. Or when the parameters of the algorithm are continuous rather than discrete, creating a practically infinite set of outputs.
This is where randomness becomes important. By using a mathematical function that generates random numbers and mapping them to the variables in an algorithm, a large output space can be explored without having to generate every possibility Sol Lewitt-style. This keeps the whole space of possibility open, yet manageable.
It’s also key to what people seem to love about generative art: the surprise. When the outputs are randomized, each time the script is run, it’s a true surprise — neither the artist nor the collector know what will appear beforehand. This surprise is part of the risk, and it’s something not welcome in traditional uses of generative systems for design (product design, engineering, architecture). “Random” is a taboo in these fields, and the intentional curation of the designer is seen as a very important part of the process. And it’s understandable, as the stakes are much higher on a building than a JPEG. But between building and JPEG are objects that could benefit from the variety and surprise that generative randomness can facilitate.
And in the case of physical objects, the importance of randomness extends to the making process, opening up new possibilities there as well, especially with the use of digital manufacturing tools — 3D printers, CNC milling machines, robot arms. This new generation of tools, unlike those that came before, are highly compatible with a design process which demands variety. A 3D printer, for example, can print a different shape every time it runs with no marginal cost. This is wildly different from something like injection molding, which requires an expensive and time-consuming mold for each unique shape.
Objects made prior to the industrial revolution were, out of necessity, made in a way that gave the maker consequential agency in the process. This agency resulted in variety as each maker expressed their own skills, taste, and preferences in the act of creation. Collectively, this resulted in an emergent richness of visual environment. We now have the tools and techniques to achieve this richness once again, but by our own design choices rather than out of necessity. The field of industrial design, because of its wide scope of practice, its familiarity with the range of tools and techniques, and its human-centered orientation, is positioned to take on this work.