Human–AI co-creativity is often understood through a tool-use or collaboration frame, obscuring the relational dynamics of artistic practice with machine learning (ML) systems. To advance the field, we recast such artistic practice through Material Engagement Theory (MET), emphasizing how creative and aesthetic processes emerge in sustained relations between artist and system. We examined 18 contemporary artists’ engagements with ML from 54 documents using a qualitative, interpretive, and framework-guided approach. Our findings show how artists and ML systems co-constitute creative and aesthetic properties by mutually discovering and attuning to each other’s affordances, where creativity unfolds along a control-chance continuum driven by creative and aesthetic thinging, enabling a mutual aesthetic becoming through sustained interaction. Key factors include developing artistic intuition for ML, internalisation, and the evolving role of the meta-artist. Herewith, we contribute MET as a framework for enriching our understanding of human–AI co-creativity as emergent and relational.

AI is rapidly becoming enmeshed in our professional and private lives. The ubiquity of such technologies raises a host of ethical questions, value clashes, and unforeseen consequences that must be confronted. Developments such as Ai-Da and DALL-E 2 are exciting in that they present robust new capabilities in AI and creativity. However, the futures such technologies unlock are also unpredictable. Given the speed with which such technologies are emerging and becoming adopted, the need to engage target audiences to weigh in on possible AI futures is critical. Our pilot project, Artistic Process Futures and AI, seeks to explore the role and potential implications of AI technologies with artists. In this paper, we show how participatory speculative design processes might be channeled into a public statement, or manifesto, regarding possible and preferable AI futures for supporting the artistic process, and how our workshop exposed uncertainty at the core of such deliberation.

Detecting fraud during online exams using proctoring software comes with substantial privacy challenges. Previous work argues students experience heightened anxiety and have privacy concerns. However, little is known about which specific aspects of online proctoring cause these concerns. This study contributes such insights by using the Contextual Integrity (CI) framework to discover how students (N = 456) rate the acceptability of 1064 proctoring information flows with varying information types, recipients, and transmission principles. We find that the acceptability varies considerably depending on the context. Besides exposing obvious privacy violations, we find that, under certain conditions, students consider it acceptable to share data with teachers - despite their lack of involvement in proctoring. Also, the acceptability of sharing highly sensitive information - which should under no circumstances be shared - sometimes increases. We discuss the implications of these and other findings and provide concrete recommendations for educational institutions using online proctoring.

Social robots can be designed to support the facilitation of co-design sessions. Facilitators regulate group dynamics to promote effective collaboration among stakeholders. Group dynamics are sensitive to mood expressions: Positive mood expressions by the facilitator promote cooperation in the group, whereas negative expressions promote conflict. However, whether mood expressions by a social robot facilitator also influence human group dynamics is an open scientific and practical question. To learn more, an experiment (N = 98) was conducted where small groups engaged in a co-design session led by a social robot facilitator. The robot displayed positive, neutral, or negative mood expressions throughout the session. The results showed that positive robot expressions, compared to neutral or negative expressions, increased perceived robot valence. Perceived robot valence increased cooperation and decreased conflict in the human groups. These findings contribute novel insight into how social robots can be used to innovate how co-design is facilitated.

Creative idea generation, the generation of original yet effective ideas, can be supported in virtual environments when interactions are facilitated via avatars with an appearance similar to their user, i.e. self-similar avatars. However, it is not known how self-similar avatars support creative idea generation. We propose that self-similarity supports the generation of original ideas, because it (i) increases the identification a user has with its avatar, which (ii) increases positive affect, and (iii) influences the positive affect-original idea generation link. To test this conjecture, an experiment was conducted where people composed their own avatar to be either self-similar or non-self-similar, which they then used to engage in two idea generation tasks presented within a custom virtual environment. The results suggest that using a self-similar rather than a non-self-similar avatar positively influences the generation of original ideas; and that this depends on the influence of self-similarity on the link between identification and positive affect. Thus, this paper contributes a mechanism that explains how self-similar avatars support the generation of original ideas.

Metacognitive awareness enables people to make conscious decisions about their own cognitions, and adapt to meet task performance goals. Despite the role of metacognition in task performance, technologies that effectively augment metacognition are scarce. We explore a novel approach to augment metacognition based on making the eye's pupil dilations, which associate with a variety of cognitions, audible via sonification in real-time. In this exploratory study, we investigated whether pupil dilation sonification can elicit metacognitive awareness. Our findings suggest that correlations between a variety of cognitions, e.g., attentional focus and depth of thinking, and sounds generated by the sonification can emerge spontaneously and by instruction. This justifies further research into the use of pupil dilation sonification as a means to augment metacognitive abilities.

Interfaces that subvert, substitute, or augment the relationship between acting and sensing make interesting tools for experimentation in artistic research. Amongst them, tactile vision interfaces, which enable people to experience to "see with their skin", have particular appeal. However, surprisingly few such interfaces are used by artists. In the present work we take a first look at how the artistic potential of tactile vision interfaces could be unlocked. On the basis of related artistic work, we argue that an accessible, generalizable and learnable interface is needed to enable artists to explore the possibilities of tactile vision. Subsequently, we present ongoing research that revolves around the development of such a tactile vision interface. A case study is presented that informs the development of a new prototype of this interface that overcomes two key limitations in its purpose to unlock the artistic potential of tactile vision.

Positive emotions can enhance the ability of people to generate original ideas, and its intensity can determine the degree to which people are able to think originally. How to design a technology that can be used to hack into this link between the intensity of positive emotion and creative thinking is, however, still an open problem. To address this we have conceived, developed, and experimentally evaluated a proof-of-concept interactive system that generates believable computational feedback about the originality of a user's own ideas in real-time. This system can manipulate this feedback to make a user's own ideas appear more or less original than people would typically judge them to be, and can also vary the order of this manipulation over time. This has enabled us to test experimentally that: (i) the order in which the positivity and negativity of the feedback is varied can be used to condition people's expectations, which (ii) can be used to later determine an intended intensity a positive emotion that a user experience, and which (iii) subsequently influence the degree to which the user is able to generate original ideas. The findings demonstrate that an interactive system can be designed to determine the type and intensity of an emotional response, in a manner that enhances the people's ability to generate original ideas.

Creative ideation can be enhanced in 3D virtual environments by manipulating the appearance of a user's avatar so that it primes a creative stereotype. However, not much is known about the factors that influence the effectiveness of using avatars to enhance creativity. In this study we investigate experimentally whether the degree to which users identify with their avatar moderates their actual ability to generate creative ideas. The results suggest that a non-creative avatar (office worker stereotype) diminishes creativity, but our creative avatar (artist stereotype) does not augment creativity. The similarity in appearance between the user and its avatar positively moderates, whereas its perceived embodiment with the avatar negatively moderates the ability to generate creative ideas. However, the study also suggests that self-similarity might be an even more effective way to support creative ideation than priming creative stereotypes. Therefore, this study (i) contributes that self-similarity and embodiment moderate the effectiveness of using an avatar to prime creative stereotypes to enhance creative ideation, and (ii) points toward a novel way to enhance creative ideation in virtual environments, by using avatars that look just like yourself.

Creativity thrives when people experience positive emotions. How to design an interactive system that can effectively make use of this potential is, however, still an unanswered question. In this paper, we propose one approach to this problem that relies on hacking into the cognitive appraisal processes that form part of positive emotions. To demonstrate our approach we have conceived, made, and evaluated a novel interactive system that influences an individual's appraisals of their own idea generation processes by providing real-time and believable feedback about the originality of their ideas. The system can be used to manipulate this feedback to make the user's ideas appear more or less original. This has enabled us to test experimentally the hypothesis that providing more positive feedback, rather than neutral, or more negative feedback than the user is expecting, causes more positive emotion, which in turn causes more creativity during idea generation. The findings demonstrate that an interactive system can be designed to use the function of cognitive appraisal processes in positive emotion to help people to get more out of their own creative capabilities.

Positive emotion can help augment human creativity. To utilize this potential in an interactive system, we propose that such a system should be designed to regulate the emotions that are caused by a creative task. We argue that this can be done by hacking the function of motor expressions in emotion regulation. To this end, we have conceived and made an interactive system that is designed to regulate positive emotion during an idea generation and an insight problem solving task. The system regulates emotion by letting users interact using arm gestures that are designed based on motor expressions, choreographed in a way that enables emotion regulation. Using this interactive system we experimentally test the hypotheses that positive approaching, rather than negative avoiding arm gestures, used to interact with a system, can heighten positive emotion, and augment creativity. The findings demonstrate that an interactive system can be designed to use the function of motor expressions in emotion regulation to help people perform better on certain creative tasks.

There is a strong relationship between the mood one is in, and the way one performs creatively. Previous research has shown that this relationship is complex. In this paper we argue that this complexity partly lies in a faulty conceptualization of mood. We will argue that an appraisal tendency perspective on moods will help to further clarify the relationship between mood and creativity. To support this argument we will highlight some inconsistencies in previous research, and use the appraisal tendency perspective on mood to develop predictions that help explain these inconsistencies and develop new directions for mood-creativity research. Future research is required to assess the accuracy of these predictions.