BAD IDEAS IN CREATIVE DESIGN SPACE EXPLORATION

This paper re-examines the nature and role of design ideas beyond criteria of quality are commonly used to evaluate the output of experimental design tasks, to guide the search processes of computational generative systems. The key principle addressed here is that focusing on high-quality ideas is insufficient if we aim to understand and support design processes. Four dimensions of ideas beyond quality are discussed, showing specific advantages which their consideration affords to supporting design space exploration. We illustrate this principle using a design-and-build task. The paper concludes with a discussion of the implications for the assessment of creativity and guidelines for the development of computational systems.