Ausra Cerkauskaite

MA Graphic Design and Typography

Everyone has a unique handwriting. Hand - pen - paper is arguably the most accessible and convenient way to write a note without losing one’s personality behind it. When we type our message on a keyboard, we are in control of how it appears on the screen. That is, we have a definite number of choices of fonts, sizes, weights, etc. to make regarding the display. However, fonts on the messaging platforms are designed to function, not to represent a person. What if a computer (machine) could select a unique font for us based on our appearance or mood to represent us?

This project is about a collaboration between a designer, a machine and a user that led to a prototype website, where the user can write a message in a font selected based on their facial features. Initially, to broaden a creative space, the machine as an independent and unbiased agent has been invited to generate unique glyphs that could potentially be assembled into character charged fonts. The results of the collaborative efforts between the machine and the designer be divided into two groups: legible fonts and miscellanies. Both groups serve as a database for the machine to select the specific font for the user.

Machine learning is getting more and more interest in the creative fields, including typography, and now is the best time to explore the creative aspects it gives to the type design, before it produces absolutely perfect letters in a perfect setting.