Abstract

This paper argues that a too-expansive view on creativity is unhelpful at best and deeply misleading at worst. As with “representation”, the word “creativity” comes value-laden in ways that researchers cannot lightly get away from, if they can escape at all; simply claiming that one is using the word in a technical sense is not a solution. Neither should one take an overly narrow view that takes advantage of a priori arguments to deny creativity to classes of agents or putative agents solely by their membership in those classes. The paper proceeds by offering a definition of creativity meant to prejudice neither human being nor artefact; then setting out the conditions for a putative creative agent to be a creative agent, concluding that no existing artefactual agents appear to fall into this category; finally, addressing the question of why computers, computer programs, robots, and related artefacts have nevertheless had a profound – indeed, transformational – effect on human creativity, taking creativity to places that neither human beings nor artefacts could have gone on their own. It ends with a discussion of the person I see as one of the key early voices on computational creativity.