In order to progress as a science, Computational Creativity research needs more formalism in terms of how we evaluate progress in the software that we write. Such formalisms can involve assessment of the increase in value of the artefacts (paintings, compositions, poems, games, recipes, theorems, etc.) that are produced by software. Other formalisms can involve assessing how much creative responsibility a piece of software has, with the value of the artefacts produced being a secondary consideration. We are developing formal models of progress in Computational Creativity research which take into account both product and process, in addition to notions related to how people perceive software, and how software presents what it does and what it has produced. Here are a couple of talks by Simon Colton in which he addressed some of the philosophical issues raised by the study of creative software:
Below are some of the papers from the group which give overviews of the field, and address issues of evaluation, accountability and other philosophical issues, in addition to suggesting some general approaches to Computational Creativity.