Kyuha Shim designs typography, though you could argue that he doesn’t do it all on his own. The PhD candidate at the Royal College of Art leaves much to algorithms of his own creation. As a researcher in the world of computational graphic design, Shim is exploring how we can use data and parameters to create visual systems, increase efficiency and solve certain nagging problems in the world of design.

Shim received his MFA in digital media from RISD before heading to MIT’s SENSEable City Lab where he worked with the team on data visualization. This is where he realized that the same parametric design principles being used to design buildings and jewelry could be used in the graphic design field as well. “Data has been my favorite subject for sometime, but I am specifically interested in how data can help practitioners make better decisions in terms of the production process in design, rather than pure representations of data,” he explains.

Algorithms can have practical applications like increasing efficiency by automating small, repeating details. It can allow graphic design elements to respond to external inputs like sound and motion. By tweaking a single parameter, you can impact every element of a layout, making the process far less labor intensive. “It takes a while to build a system which solves specific problems, so it is impractical if we use it for one time,” he says. “However it becomes useful if users revisit the issue again and again.”

Algorithms also have the ability to be a creative force. We’ve seen this in Hoefler’s Obsidian typeface, which used algorithms to generate the shading of the font. MIT Media Lab has long had a shape-shifting logo, first by E Roon Kang and Richard The then by Pentagram’s Michael Bierut. Computer scientist Donald Knuth built Metafont, a software that algorithmically designs typefaces, and Shim’s own projects similarly explore form created by bending to a certain parameters.

In Performa, for example, Shim built an algorithm that uses typing-input data to create typography in real time. Press a key and the corresponding character will stretch, thicken and elongate until you remove your finger. The longer you hold a key, the more extended the letter becomes. The harder you push the key, the thicker the stroke width. In GEN Shim explores how complex code-driven typography can get. The morphing image you see is the result of a “turbulence noise algorithm” and randomly automated crops of images. “While the user controls the density of image particles on the outline of texts, the system randomly choses most of the other parameters (e.g. decay, size, position of the fragments from the image and so on) that dictate the rendered appearances,” he says.

It’s an interesting paradox: For all the parameters attached to projects like Performa and GEN, they lack a certain measured control often found in typography. But like painting or drawing, these projects have an immediate expressivity. In a recent blog post, designer Rune Madsen argues that we’re at the beginning of an entirely new paradigm in design. Metadesign, as he calls it, is a natural continuation of the contemporary designer’s skill set. Like sketching, building a framework through code will become expected of designers. It’s easy to see why. Parametric design is certainly efficient, and it can be a means to a creative ends. But it also presents the opportunity to push the limits of a given design practice. If no tools or systems exist to enable what you want to create, you just make the tool yourself.

In that way, generative design is more about the way in which you arrive at an outcome rather than the outcome itself. It’s an exciting thing to consider. As Pentagram’s Natasha Jen puts it:  “Parametric design is a technique, just like water color is a technique or oil painting is a technique. It doesn’t have inherent value. It’s not good or bad.” As it is with most creative endeavors, it’s what you achieve using that technique that matters.