Figure - available from: Frontiers in Robotics and AI
This content is subject to copyright.
(A–C) Q1. Self-reported years spent drawing. (D–F) Q2. Drawing education. (G–I) Drawing practise: participation in collaborative or collective drawing (Q3), in drawing communities (Q4) or as a professional drawer (Q15). Abbreviations are in bold. Participants selected all answers that apply, except for Q1 where participants selected only one category.
Source publication
This article presents the outcomes from a mixed-methods study of drawing practitioners (e.g., professional illustrators, fine artists, and art students) that was conducted in Autumn 2018 as a preliminary investigation for the development of a physical human-AI co-creative drawing system. The aim of the study was to discover possible roles that tech...
Citations
... Another theme is the importance of human intervention in working alongside GenAI. While GenAI can generate content, it is the artists' role to steer the GenAI to ensure the outputs align with their creative vision and the contextual requirements (Jansen and Sklar 2021;Liu et al. 2024;Tsao and Nouges 2024). For instance, Tsao and Nouges (2024) found that students actively intervened to include their own direction and style into the GenAI-generated content. ...
... Additionally, artists tend to prefer creative outputs where they have invested more effort, likely due to feelings of ownership, competence, or effort justification (Mehler et al. 2024). Jansen and Sklar (2021) found through qualitative interviews that artists generally prefer cocreative GenAI and remain sceptical about fully automating creative work. In line with this, Zhou and Lee (2024) noted that while GenAI can initially boost creative productivity, the novelty of GenAI-generated content decreased as the process became more automated and standardised. ...
Generative artificial intelligence (GenAI) has recently attracted attention from literature and organisations, especially due to advances in machine learning techniques. However, research on GenAI in creative contexts remains in its early stages, with few attempts made to assess the current body of research or synthesise the existing knowledge in this area. To address this gap, this paper employs a systematic literature review of 64 studies to identify methods, research trends and key thematic insights shaping the current understanding of GenAI in creative contexts. The findings of this systematic literature review emphasise the rapid development of research on GenAI in creative contexts. The analysis highlights key factors influencing the adoption and impact of GenAI in creative processes, as well as the implications for creative outcomes and industry practices. From this analysis, several potential directions for future research emerge, including the long-term effects of GenAI on creative processes, socio-economic implications for creative industries, and frameworks for ethical use, and perception of GenAI-generated content.
... This phenomenon is similarly observed in activities such as art and digital fabrication [27,51]. In co-drawing with robots, physical touch and textures of drawing materials made the artists prefer tangible mediums (e.g., pencils) than digital tools (e.g., tablets) that fall short in these respects [67]. ...
Robots extend beyond the tools of productivity; they also contribute to creative activities. Although typically defined as utility-driven technologies designed for productive or social settings, the role of robots in creative settings remains underexplored. This paper examines how robots participate in artistic creation. Through semi-structured interviews with robotic artists, we analyze the impact of robots on artistic processes and outcomes. We identify the critical roles of social interaction, material properties, and temporal dynamics in facilitating creativity. Our findings reveal that creativity emerges from the co-constitution of artists, robots, and audiences within spatial-temporal dimensions. Based on these insights, we propose several implications for socially informed, material-attentive, and process-oriented approaches to creation with computing systems. These approaches can inform the domains of HCI, including media and art creation, craft, digital fabrication, and tangible computing.
... To be considered a creative drawing, the style of representational drawing must differ significantly from traditional drawing (Chen et al., 2020;Jansen & Sklar, 2021). Based on this psychology background, Torrance has two significant tests, such as verbal (TTCT-verbal) and nonverbal (TTCT non-verbal), also, this test consists of two types of stimuli, such as figural (TTCT-figural) and verbatim (TTCT verbatim) (Alabbasi et al., 2022;Lee et al., 2024;Torrance, 1966). ...
The Torrance Test of Creative Thinking (TTCT) is commonly used to assess creativity in pupils, but research is limited on its effectiveness in identifying cognitive levels and how factors like gender and brothers count influence creativity. This study investigates the interaction between gender and the number of brothers on creativity as assessed by the (TTCT-Figural: Circles test) among students aged 7 to 13. A cross-sectional study design was employed, involving a sample of 11,636 students, including 5571 males (47.9 %) and 6065 females (52.1 %), selected from public and private schools across Sudan. The findings demonstrated an interaction effect between gender and the number of brothers, revealing significant gender differences in (TTCT-Figural) performance , particularly across the domains of fluency, flexibility, and originality, with females outperforming males in all three areas. Notably, no gender differences in flexibility were observed at ages 10 and 13, nor in fluency at age 13, and originality showed no gender disparity across all age groups. Additionally, TTCT-Figural performance did not vary significantly across different age groups. The majority of students exhibited very low performance at age 7, with slight improvements noted at ages 8 and 9. However, after age 9, there was a decline in the number of students performing at a moderate level. Moreover, these results emphasize the importance of understanding the role of gender and familial factors, such as the number of brothers, in creativity development. The implications of this study suggest the need for targeted educational programs and policies intended at enhancing (TTCT-Figural) performance, particularly for pupils, to adoptive creativity further effectively.
... Porcu et al. (2024) defines it as "the freedom to experiment, to be creative, and to investigate radical possibilities". More recently, the concept of human creative autonomy is used to assess the sense of control and authorship experienced by artists when interacting with a technological device (e.g., Jansen and Sklar, 2021). Another developing line of research (Vinchon et al., 2023) investigates the creative autonomy of hybrid systems where humans and AI collaborate and share or alternate their control in the co-creation of a (creative) output. ...
... It can be suggested that the advanced functions of generative models could bring this "exploration" even further, by suggesting new ideas (e.g., new melodies, new artistic styles) that have not been considered by the human user yet. Moreover, as other artificial tools (Liapis et al., 2016;Jansen & Sklar, 2021), generative models could benefit human creative autonomy by optimizing time and resources, providing inspiration and stimuli for new ideas, as it seems supported by some initial studies (Doshi & Hauser, 2023). ...
... For instance, Liapis et al. (2016) suggest that computers can improve human creativity and lateral thinking either by providing a stimulus for reframing humans' routines and mental associations, or by using human input to circumscribe the search space for specific and novel solutions. One study (Jansen & Sklar, 2021) reports that artists and illustrators are skeptical about the presence of an AI collaborator which can interfere with their artistic work, but at the same time they recognize the benefits of optimizing resources and providing further inspiration. This suggests that AI is perceived as capable of both obstructing and supporting creative autonomy in the artistic process. ...
In the wake of recent advancements in the field of AI, this paper investigates the impact of recommender systems and generative models on human decisional and creative autonomy. For this purpose, we adopt Dennett’s conception of autonomy as self-control. We show that recommender systems can play a double role in relation to decisional autonomy: as information filter, they can augment self-control in decision-making, but also act as mechanisms of remote control that clamp degrees of freedom. As for generative models in AI, we show that they can be seen as a powerful system of selection and suggestion (similar to standard recommender systems) but also as an instrument for information production. We suggest that the latter perspective opens new possibilities in terms of creative autonomy. Additionally, for both systems we propose a distinction between “extrinsic” and “intrinsic” mechanisms and effects. Through Dennett’s theory of self-control, this paper offers new insights into the relation between AI and human autonomy by framing it in terms of remote or self-control and by addressing the impact of generative models on creative autonomy.
... Furthermore, if robots are to support humans in the creation of artwork, it is important for the robot to have flexible styles of strokes for the user to specify either through choice or demonstration. Many artists do not wish to automate the artistic process [3], [4], but some are open to co-creative assistants [5]- [8]. The work operates on the assumption that giving more creative control to a user co-creating with a robot over the style of the image, allows them to feel more ownership over the artwork that they create with the robot. ...
A painting is more than just a picture on a wall; a painting is a process comprised of many intentional brush strokes, the shapes of which are an important component of a painting's overall style and message. Prior work in modeling brush stroke trajectories either does not work with real-world robotics or is not flexible enough to capture the complexity of human-made brush strokes. In this work, we introduce Spline-FRIDA which can model complex human brush stroke trajectories. This is achieved by recording artists drawing using motion capture, modeling the extracted trajectories with an autoencoder, and introducing a novel brush stroke dynamics model to the existing robotic painting platform FRIDA. We conducted a survey and found that our open-source Spline-FRIDA approach successfully captures the stroke styles in human drawings and that Spline-FRIDA's brush strokes are more human-like, improve semantic planning, and are more artistic compared to existing robot painting systems with restrictive B\'ezier curve strokes.
... However, as Schön [31] points out, in many professional and creative tasks, the outcome is usually not something that is already determined at the start of the project -finding out what the end goal is and reflecting on the process are aspects that happen during, and not just before the activity. This is particularly the case for sketching and drawing as part of ideation and design processes [18,34], as well as for learning maker skills [36,37]. ...
... Thus, in a co-creative activity, a machine needs to understand what the human wants, including the multi-modal ways in which that is communicated. The activities that collaborative machines could perform in the drawing process have been proposed to be corrective drawing, predictive drawing (including labor-saving drawing), scene completion and artist-block mitigation [18]. While in the overall interaction the human is certainly "in charge", the initiative for sub-tasks is divided to a certain degree between the two parties -it becomes a mixed-initiative interaction (see Horvitz [17]). ...
... Jansen and Sklar [18] review the main works on (physical) drawing interaction in literature and art, dividing this space into the categories of artists drawing alone, collaborative drawing, robotic systems drawing alone, and co-creative drawing systems. Classic examples of human-to-human collaborative drawing interfaces in the HCI literature include VideoDraw by Tang and Minneman [35] at Xerox PARC (1991), which is a shared drawing tool that allows two users to not only draw together but also gesture towards each other in the process of drawing. ...
... Evidence has been gathered specifically regarding perceptions of AI in the art realm. Studies have focused on the extent to which participants have identified AI-generated art and distinguished it from human-made art (e.g., Gangadharbatla, 2021;Schubert et al., 2017), whereas other studies have delved deeper into perceptions of AI-generated art and the factors that influence the perceptions of the AI system or the art it generates (e.g., Bellaiche et al., 2023;Hong & Curran, 2019;Hong et al., 2021Hong et al., , 2022Jansen & Sklar, 2021;Lima et al., 2021;Tubadji et al., 2021;Wu et al., 2020). In general, evidence from these studies points toward people not always being able to recognize AI-generated art or to differentiate it from human-generated art and toward people tending to value human-made art over AI alternatives, although not in call cases. ...
... Studies focusing on human perceptions of AI in art varied in their methodology, including an experiment (Lyon et al., 2021), survey experiments (Epstein et al., 2020;Gangadharbatla, 2021;Hong & Curran, 2019;Hong et al., 2021Hong et al., , 2022Lima et al., 2021;Tubadji et al., 2021;Wu et al., 2020), and mixed-method studies (Jansen & Sklar, 2021;Schubert et al., 2017). Lyu et al. (2021) found that participants with an art or design background could recognize various artistic styles of visual arts (Fauvism, Expressionism, Cubism, and Renaissance) even after AI-created changes. ...
... music (Schubert et al., 2017). Along with identifying AI-made art, researchers have focused on participants' evaluations of AI-made art (Gangadharbatla, 2021;Hong & Curran, 2019;Hong et al., 2021Hong et al., , 2022Jansen & Sklar, 2021;Lima et al., 2021;Tubadji et al., 2021;Wu et al., 2020). The studies dealt with artistic value, monetary value, quality, and appreciation of AI-created artworks. ...
... Other factors also influenced assessments of AI-produced art. They included AI's perceived moral status (Lima et al., 2021), cultural differences (Wu et al., 2020), and the context in which AI was used (Jansen & Sklar, 2021). Lima et al. (2021) showed in two experiments that participants' evaluation of AI's agency (AI's ability to create and experience art) was not changed by the process in which they were exposed to AI-generated art. ...
... The HCI and design research community have long appreciated collaborative forms of speculative experimentation and art-based activities [2,20,21,22,23,66], while valuing drawing in particular as a form of knowledge production and sense-making [6,18,27,34,35,38,39,41,42,45,48,49,54,57,61,62,66]. Alongside this body of work, emerging more-than-human approaches [12,13,20,32,33,46,51,53,64] acknowledge relationships between humans and non-humans as interdependent, where non-human perspectives can point to novel insights and new design opportunities. ...
... Today, computational models such as Generative Adversarial Networks (GANs), Convolutional/Recurrent Neural Nets (CNNs/RNNs), and Generative Pre-Trained Transformers (GPTs) appear to have imaginations of their own as they newly contribute to creative processes by adding layers of interpretation and generativity [3]. This capability of humans and artificial entities to imagine and perform alongside one another has given rise to new concepts and theories of creativity including computational creativity [8,9,11,15,16,26,38,50], postanthropocentric creativity [52,55], and digital craft-machine-ships [1,3,10]. These theories focus not on whether or not the AI is intelligent but rather on how including more-than-human agents in creative processes may give rise to new approaches for creative expression and sense-making in art and ...
In this pictorial paper, we present a series of drawing conversations held between two humans, mediated by computational GAN models. We consider how this creative collaboration is affected by the hybrid inclusion of more-than-human participants in the form of watercolour and artificial intelligence. Our drawing experiments were an extension of our search for new ways of seeing and telling, which includes a reflection of the extent to which more-than-human elements took part in our creative process. We discuss our tendencies to form strange interpretations and assign meaning to the unpredictable and ambiguous spaces we created with them. We further speculate on the characteristic material agencies they revealed in our interactions with them. Finally, we contend how such collaborations are already and always embedded and embodied in our ways of seeing and knowing in design and creativity research.
... We are witnessing human-as-collaborator applications in a wide range of art, ranging from a painting to a film. Collaborative drawing and painting workflows utilize AI and robotic arms to co-create with humans (Jansen and Sklar 2021). In AIBO, the opera performer's spoken words were used to generate texts from GPT-2 (Pearlman 2021). ...
Artificial intelligence (AI) is increasingly utilized in synthesizing visuals, texts, and audio. These AI-based works, often derived from neural networks, are entering the mainstream market, as digital paintings, songs, books, and others. We conceptualize both existing and future human-in-the-loop (HITL) approaches for creative applications and to develop more expressive, nuanced, and multimodal models. Particularly, how can our expertise as curators and collaborators be encoded in AI models in an interactive manner? We examine and speculate on long term implications for models, interfaces, and machine creativity. Our selection, creation, and interpretation of AI art inherently contain our emotional responses, cultures, and contexts. Therefore, the proposed HITL may help algorithms to learn creative processes that are much harder to codify or quantify. We envision multimodal HITL processes, where texts, visuals, sounds, and other information are coupled together, with automated analysis of humans and environments. Overall, these HITL approaches will increase interaction between human and AI, and thus help the future AI systems to better understand our own creative and emotional processes.
