Editorial

Computational creativity is a relatively new field which brings together elements of artificial intelligence, cognitive science and philosophy to build and study computational systems which model or display potentially creative behaviours. The First International Symposium on Computational Creativity took place during the convention organised by the Society for the Study of Artificial Intelligence and the Simulation of Behaviour (the AISB) at Goldsmiths University of London in April 2014 when the convention was celebrating 50 years since the founding of the AISB and 60 years since the death of Alan Turing, founding father of both computer science and artificial intelligence. Following an auspicious inauguration, the symposium has since carried on to celebrate its fourth edition during the AISB convention in April 2017 at the University of Bath, bringing together researchers to discuss practical work in the area, theoretical approaches to computational creativity and philosophical questions raised regarding the potential of non-human creative agents.

Research on computational creativity, whatever claims it makes, often attracts heated debates about whether machines can ever be creative. A similar debate has long surrounded research in the field of artificial intelligence, where the notions of weak and strong AI have been at the heart of philosophical arguments for decades. The surge in the use of digital tools in creative production adds to the significance of the debate. On one hand, some researchers in the field sidestep these issues and focus on pushing forward the technical state-of-the-art. On the other, ostensibly creative computational agents provide a new lens through which philosophers, technologists, creative practitioners and the general public can explore notions of creativity and artificial intelligence.

To highlight the success of the symposium and draw attention to new developments in the field, we have edited this special issue for Connection Science. In this issue of the journal, several of the leading researchers in the field present their ideas and argue for the relevance and significance of these ideas in today's computational landscape.

The themes that emerge in this issue will resonate with the current trends, both theoretical and practical, developing across the field. In particular questions are raised about how the broader context of creativity, including its social setting and collaboration between creative agents, be they human or artificial, should impact the study of computational creativity. In addition to theoretical and evaluative work, novel systems for generating creative artefacts are described, again with a tendency towards awareness of the context in which these types of systems are perceived and the importance of understanding human perceptions of creativity in modelling and evaluating computational creativity.

In “Is Deep Dreaming the New Collage?”, Boden suggests that while Google's deep dreaming methodology can combine and transform images in surprising ways, being based in deep learning, it is not analytically interpretable. This approach is contrasted with the conscientious appropriation and recombination of culturally and politically charged signifiers that has characterised the human practice of collage. In the context of collage, then, Boden argues that deep learning will not be able to generate truly creative artefacts until the paradigm can be guided in a disciplined fashion.

In another work “Swarmic Autopoiesis and Computational Creativity”, al-Rifaie et al. from Goldsmiths, University of London address the philosophical concept of autopoiesis and autopoetic artist in the two modalities of Gluttonous Autopoietic Artist and Contented Autopoietic Artist. In the first, as the decoding process of the input sketch unfolds, it leads to a less complex structure which ultimately results in an empty canvas; therein reifying the artwork's “death”. In the latter mode, by refocusing the autopoietic artist's reflections on “meaning” the autopoietic processes become more open-ended in their creative endeavour. Therefore, the relevance of autopoiesis to the field of computational creativity is emphasised, and its impact is discussed and visualised using two swarm intelligence algorithms, Dispersive Flies Optimisation and Stochastic Diffusion Search.

Researchers from Spain, Delatorre et al., in their work “A Computational Model of the Cognitive Impact of Decorative Elements on the Perception of Suspense”, explore the role of suspense in perceiving a story. They argue that existing computational systems are mainly based on narrative theories, and in general lack a cognitive approach involving the study of empathy or emotional effect of scene objects; following on this, they propose a computational model of the influence of decorative elements on suspense.

In a related article, “Sparks Will Fly: Engineering Creative Script Conflicts”, Veale and Valitutti explore the concept of humour. These researchers from University College Dublin argue that with a computational framework in place, it is possible to begin to explore the fundamental differences between, on the one hand, script blends that are relatively seamless, and, on the other, those that create sufficient friction to be viewed as humorous. In their analysis, they use Twitter style character constraints to generate both ironic remarks and short narratives.

Jordanous, from the University of Kent, critiques the Beyond the Fence project, which used computer software as much as possible in order to create what was dubbed as “the world's first computer-generated musical”, a production subsequently performed in London's West End for two weeks in 2016. Beyond an analysis of the production itself, the challenges of evaluating this complex and collaborative project are investigated in this article titled “Has computational creativity successfully made it ‘Beyond the Fence’ in musical theatre?”

In “Towards a Computational- and Algorithmic-level Account of Concept Blending Using Analogies and Amalgams”, Besold et al. provide an intermediate report on work towards the development of a formal computational account of concept blending. In their work, they present the theoretical background as well as an algorithmic proposal combining techniques from computational analogy-making and case-based reasoning, and demonstrate the feasibility of the approach in two case studies.

This special issue journal is the second in as many years publishing extensions of work originally presented at the AISB Symposium on Computational Creativity as well as invited novel work from active researchers in the field. Following the success of the first symposium, the Second International Symposium on Computational Creativity took place at the AISB Convention in Kent in 2015, the Third Symposium at AISB 2016 in Sheffield, the Fourth Symposium at Bath in 2017 and the Firth Symposium will be held in Liverpool in 2018. We look forward to participating in the continued evolution of Computational Creativity as a field.

We would like to thank the authors of this volume as well as the reviewers, whose meticulous comments and suggestions enhanced the already strong submissions and assisted in presenting a rich mix of theoretical discourse, applications of state-of-the-art developments and philosophical debates on computational creativity.

And last but not least, we would also like to express our gratitude to the editor-in-chief of the journal, Tony Prescott, and his team for all their support in publishing this special issue on computational creativity, as a topic with increasing relevance and significance in this day and age.