Generating

Overview

The act of designing involves the creation of something new, whether an object, an image, or an idea. I call the first pillar of the framework ‘generating’ as opposed to ‘creating’ to highlight how computation allows designers to work at a level of abstraction from singular artifacts. An artist working alone might create a painting or a drawing. If, however, they were to generate it, that would imply an underlying structure and logic capable of producing more artworks, each qualitatively different from the others but arrived at through the same mechanism. Generating also implies a complexity in the production of a generative process that is gradually insurmountable by human effort alone. In his 1968 essay “Systems Generating Systems,”1 American architect and theorist Christopher Alexander differentiates ‘systems as a whole’ from ‘generating systems.’ The former are complex assemblages, not singular objects, that are nevertheless characterized by some “holistic phenomenon.” The latter are also not individual things, but groups of objects and forces “with rules about the way these parts may be combined.” Set into motion, they produce complex outputs that are more than the sum of the parts: Generating systems generate systems as a whole. Alexander closes his essay with a call to action for designers and architects: “To make objects with complex holistic properties, it is necessary to invent generating systems which will generate objects with the required holistic properties.”

If we grant that the work of designers has become more complex — think not only of large software programs or buildings with demanding energy requirements but also the ways in which designed artifacts are situated in socio-technical contexts — then it is important for designers to work generatively. Through computation, this is intuitive for technologists; something that is hard-coded is inherently limited to a certain scope, whereas abstractions allow for more flexibility and multi-purpose reuse. Computational design brings certain key methods to aid the designer in generating systems. In particular, through _rulemaking _processes, designers can more easily work at the level of the system. By _exploring (or, searching) _they can make sense of the potentially overwhelming space of possibilities that is generated. Finally, in _automating _aspects of their work, the goal for designers is not so much efficiency or optimization as the ability to shift between various levels of abstraction.

1. Alexander, Christopher. “Systems generating systems.” 1968.

results matching ""

    No results matching ""