Augmented Creativity: Using AI Collaborators to Make a Music Video Dancecult: Journal of Electronic Music & Dance Culture 17(1): 102–128 ISSN 1947-5403 ©2025 Dancecult http://dj.dancecult.net https://dx.doi.org/10.12801/1947-5403.2025.17.01.06 Augmented Creativity: Using AI Collaborators to Make a Music Video Kirsten Hermes University of Westminster (UK) Abstract This article explores the use of AI tools in the DIY production of a music video through a practice-led case study involving Midjourney, Runway and Plask. The project demonstrates how AI enables faster and more stylistically ambitious visual content creation, especially for music producers without formal visual training, while also reshaping the artistic challenges for the maker. Through critical reflection, public feedback and an industry interview, the study examines how AI functions as an incubator and technical executor, aiding in the preparation and elaboration stages of the creative process. Drawing on creativity theory, this article argues that while AI can produce novel and valuable outputs, it lacks intentionality and must be guided, interpreted and curated by humans. The research supports emerging concepts of AI co-creativity, suggesting that the future of content production lies not in AI replacing humans, but in augmenting human direction through collaborative frameworks. Keywords: AI, music video, co-creativity, MidJourney, motion capture Kirsten Hermes is an interdisciplinary researcher, book author, senior lecturer and audio-visual artist, bridging scientific and creative domains in her work. She tours internationally under the moniker Nyokee. Kirsten holds a PhD in sound perception from the University of Surrey (UK), which was funded by the EPSRC. Kirsten regularly publishes interdisciplinary book chapters and academic articles, combining scientific and technical knowledge with creative practice. She recently published the book “Performing Electronic Music live”. Kirsten also works as a freelance sound designer and video creator, for example for ROLI. Feature Article https://dx.doi.org/10.12801/1947-5403.2025.17.01.06 Hermes | Augmented Creativity 103 Artificial Intelligence (AI) refers to systems that can learn from data and adjust their behaviour without explicit programming, with generative AI now capable of producing creative outputs such as text, images and audio. While creation is never ex nihilo, being built on the shoulders of predecessors (Sarmiento and Stahl 2008), the rise of generative AI has sparked debates around ethical use and the potential replacement of human creatives. Supporters of Human-Centred AI advocate for approaches that support, rather than replace, human capabilities (Browne 2022; Shneiderman 2022). All creative industries are already experiencing the integration of AI into various aspects of the production process. In music production, more research is needed to understand their influence on the creative thinking of musicians. Rather than pursuing full automation, this article focuses on a co-creativity setting, exploring both the affordances and limitations of AI tools in a collaborative artistic workflow. Specifically, it examines how co-creativity between humans and AI can be better understood through the lens of music video creation. This is explored through a case study in which a music video was produced using three AI tools: the generative image platform Midjourney, the video platform Runway and Plask, an AI-powered motion capture tool. In addition, a semi-structured interview was carried out with an AI industry leader, using the music video as a springboard for discussion. Further, a documentary was produced around the creative process and shown to online communities on Reddit, YouTube, Facebook and Instagram, to receive feedback from the public. Via this experiment, critical reflection was used as a research method. The findings are grounded in existing literature around creativity. The first section is a brief introduction to the field of AI and tools relevant for musicians. The second section is a summary of existing creativity research. The third section details the practice-led case study. The fourth section discusses the findings from this case study in the context of the existing literature and the last section is the conclusion. A Brief Introduction to Creative AI Tools Artificial intelligence (AI) refers to machines learning from experience and adjusting to new inputs instead of being explicitly coded. The term AI was coined in Dartmouth in the fifties, when pioneer scientists attempted to develop a “fully intelligent machine” (Wingström et al. 2022). However, to this day, intelligence is not fully understood and therefore difficult to replicate (Wingström et al. 2022). Narrow artificial intelligence is designed to perform highly specific tasks, as found in many commercially available AI tools. General artificial intelligence refers to AI that could replicate human intelligence, which is currently not possible. Oxford philosopher Nick Bostrom (2014) defines artificial super intelligence (ASI) as “any intellect that greatly exceeds the cognitive performance of humans in virtually all domains of interest”, a type of technology reserved for sci-fi movies. Within AI, the field of machine learning enables machines to carry out tasks which are non-routine and require creativity (Susskind 2020). Neural networks are a special case Dancecult 17(1)104 of machine learning. Generative AI is a category of artificial intelligence algorithms that generate new outputs based on specific training data, such as images, text, or audio. There are now dozens of AI tools available for music creation, spanning composition, production, mixing, mastering and more. Platforms like Suno, Udio, Stable Audio, Boomy and Mubert can generate full audio tracks with customizable styles, while tools such as Orb Producer, AudioCipher and Captain Chords focus on MIDI generation. AI-assisted music production is also supported by platforms like LifeScore, Aimi, BandLab and Logic Pro, which offer features for AI collaboration. Mixing and mastering can be automated with tools like iZotope, Sonible and Landr. Audio separation tasks can be handled by LALAL. AI or Splitter.ai. Beyond music, AI tools are increasingly used in adjacent creative domains. Text-based tools like ChatGPT, Claude and Gemini can generate lyrics, promotional content, or concept development. For visual assets such as album artwork and social media content, platforms such as Midjourney, Google Imagen, Stable Diffusion and Flux are being widely adopted. In video, tools like Runway, Sora, Stable Video Diffusion and Kling support the creation of music videos and short-form visuals. Some platforms, such as Plask and Deepmotion, offer motion capture from video and emerging research is also exploring text- to-motion generation. AI-generated sound effects for video can now be created via tools like ElevenLabs SFX Generator, SFX Engine and MMAudio. As AI continues to evolve, so do its potential applications across the entire creative pipeline, from idea generation to post-production. Many creatives are concerned with AI replacing their jobs (Susskind 2020) and there are urgent issues surrounding copyright infringement, particularly the unlicensed use of artists’ work in AI training datasets (European Innovation Council and SMEs Executive Agency 2024), as well as the risk of homogenisation and standardisation of creative outputs due to the tendency of generative models to reinforce cultural bias (Zhou et al. 2024). These developments threaten to marginalise underrepresented voices, erode creative diversity and devalue human authorship. In response, this article explores the potential of AI co-creation, that is collaborative processes between humans and machines (specifically for a music video), as a more ethical and creatively fulfilling alternative. In order to understand how humans can collaborate with AI to produce creative outputs, the state of the art in creativity research is briefly summarized in the next section and will be used as a framework for the discussion going forward. Researching Creativity Creative ideas are novel and valuable combinations of known ideas (de Mántaras 2020). As such, creativity is agreed to be a twofold concept, requiring both originality and effectiveness (Kaufman 2016; Wingström et al. 2022). As Becker (1982) states, “we are interested in the event which consists of a work being made and appreciated; for that to happen, the activity of response and appreciation must occur”. Similarly, Bourdieu (1993) posits that works of art exist as symbolic objects only if they are known and recognized, that is socially instituted as works of art. For an output to be perceived as creative, it must be new and Hermes | Augmented Creativity 105 surprising, a definition that often traces back to Margaret Boden’s work (2004). But where does creativity emerge? Traditional romantic and inspirationist assumptions saw the individual at the centre of the creative process, whereby a moment of insight or inspiration led to creativity. As such, early studies into creativity focused on the individual and their biological makeup (Thompson 2015). The earliest known study into creativity, carried out by Sir Francis Galton (1869), focussed on “hereditary genius” (Thompson 2015). Nowadays, it is commonly agreed that several factors are required for creativity, and that cultural and social circumstances are as important as the individual (Csikszentmihalyi 1988 and 1999; Sawyer 2012). Csikszentmihalyi posits that “creativity is the property of a complex system and none of its components alone can explain it” (1997: 56). Creativity as a Systemic Phenomenon In his systems model of creativity (see fig.1), Csikszentmihalyi (1988 and 1999) specifies that a creative system needs a “domain” or culture that consists of symbolic rules, guidelines and practices; an “individual” that can bring novelty into the domain and a “field” of experts who recognize and authenticate that novelty. Figure 1. Revised systems model of creativity (Kerrigan 2014). Image credit: Kirsten Hermes. Dancecult 17(1)106 The ongoing interactions that occur between the domain, the field and the individual, do not have distinct stages or a specific start or end points: the relationships between each of the elements are “dynamic links of circular causality” (Csikszentmihalyi 1988: 329). As such, systems are non-linear, complex and interconnected. Communication in the “field”, in the form of collaboration and social networking, plays a crucial role for creativity. Many definitions of creativity contain the elements of actor, process, outcome, domain and space (Wingström et al. 2022), particularly the first three (Amabile 1996; Boden 2004; Csikszentmihalyi 1997; Rhodes 1961) and existing studies show that all five are still relevant in the context of AI (Wingström et al. 2022). Actors produce creative outcomes within a domain, by engaging in a creative process. Space refers to a network of actors, artifacts and environments that interact in the creation processes (Wingström et al. 2022). Some researchers consider creativity to be distributed into artifacts and the environment (Wingström et al. 2022), which becomes especially relevant in the new field of AI. Creativity as a Process So far, creativity has been described as emerging from systems or networks. It can also be studied as a process with different stages. Csikszentmihalyi (2013) distinguishes the stages of preparation, incubation, insight, evaluation and elaboration. In the preparation stage, a problem definition takes place, sparking curiosity. During incubation, ideas are processed in the subconscious, eventually leading to insight. During the evaluation stage, an individual decides whether the insight is valuable and worth pursuing. Lastly, elaboration occurs, where the work is implemented and realized. Notably, these stages are not linear but recursive and the number of iterations depends on the depth and breadth of the issues dealt with (Csikszentmilahyi 1997). In this way, the subprocesses involved in creativity recur over and over in complex sequences (Eindhoven and Vinacke 1952; Mumford et al. 1991; Runco 1994). The phases can also co-occur when a person engages in preparation for one aspect of a problem and incubates for another aspect of the problem (Lubart 2010). Some older models have four stages (Lubart 2001). In his presidential address to the American Psychological Association, Guilford (1950: 451) noted that there was “considerable agreement that the complete creative act involves four important steps”, that is, preparation, incubation, illumination and verification. Amabile (1996) distinguished problem or task identification, preparation, response generation and response validation and communication. By separating the creative process into distinct stages, AI creativity can be studied in a more nuanced way. The findings and discussion section refers back to these stages from the perspective of the critical reflection. Creative Personalities Some studies are concerned with understanding creative personalities. The well-known OCEAN model (or five factor model, Furnham 2008), refers to the parameters of openness to experience, conscientiousness, extraversion, agreeableness and neuroticism (Costa and McCrae 1992). Another often discussed parameter is divergent thinking, which humans Hermes | Augmented Creativity 107 do to imagine novel ideas and to explore multiple solutions to a problem (Runco and Acar 2012). Divergent thinking is followed by convergent thinking, where thoughts are combined in a structured way to arrive at the best-suited solution (Wingström et al. 2022). Beyond these innately human personality criteria, de Mántaras (2020) posits that creativity is an advanced form of problem solving that involves memory, analogy, learning and reasoning under constraints: all traits that can be replicated by means of computers. He claims that creativity can be investigated, simulated and harnessed for the good of society. The findings and discussion section analyses to what extent AI could be seen as a creative agent. Interactions between individuals and their social environments have a large impact on creativity (Csikszentmihalyi 1997), as discussed in the following section. The extent to which outputs are both novel and useful has been categorized in different ways. Boden (2004) distinguishes P-creativity (psychological) and H-creativity (historical). P-creativity relates to outcomes that are surprising and novel to an individual’s mind and H-creativity relates to revolutionary outcomes that remain significant throughout history. In this way, creative outputs can not only constitute eminent work, but also more “everyday” levels of creativity (Lubart 2001). Similarly, definitions of “big C” creativity are concerned with exceptionally creative geniuses, while “little c” refers to the everyday creativity of a given individual (Simonton 2013: 71; Stein 1987: 420). The rank ordering of creative outputs is socially constructed and prone to change. As such, creativity cannot be reduced to formal properties ( Jennings 2010). Beyond the Human-Centric View Distributed creativity theories (Glǎveanu and Petre 2014) and posthuman creativity theories (Harris and Holman Jones 2022) challenge more human-centric views and instead emphasize the interaction between humans and “non-humans” — for example, artifacts and technologies. Bruno Latour, Michel Callon, Madeleine Akrich and John Law devised actor- network theory (ANT) in the 1980s. Here, groups of human and non-human elements form networks and affect each other via connections in these networks (Latour 1996). While ANT is often seen as conflicting with the systems view, both models highlight that all human outputs involve complex landscapes of many parameters. From a posthuman perspective, knowledge and creativity can be co-constituted with artifacts or technologies (Ihde 1990; Latour 2013; Rose 2017; Wingström et al. 2022). Ihde (1990) argues that technologies mediate the human world: technology enables certain actions and hence humans experience the world with and through it (Wingström et al. 2022.) Posthumanism offers a critical extension to sociocultural perspectives which can be useful in assessing generative AI (Kumar and Tissenbaum 2022). Practice-led Case Study In order to further understand how AI could be integrated in the creative process, a practice- led case study was carried out. Clarke and Cook (2004) state that theorists can successfully Dancecult 17(1)108 use experimentation as a research method, as for example exemplified in Marion Gucks “thought experiments” (1994). Autoethnographic research can make a difference on culture (Smith 2017). Haseman (2006) asserts the significance of practice-led research and argues for it to be understood as a research strategy within an entirely new research paradigm, as separate from traditional qualitative and quantitative research. Practice-led research is situated within the constructivist paradigm, which Maguire (2019) defines as the internal or creative interpretation of the outside world, necessitating differentiation between elements and the fundamentally changeable connections or interrelationships they share. Within this paradigm, researchers learn from experience and practice. The key research method used for this article was a critical reflection on the creative process. Creative reflection is a method of data collection (Fook 2011), with the data consisting of the knowledge or assumptions underpinning practice and the changes made to this knowledge through the critical reflection process. The author acted as both researcher and artist, combining three different AI tools in a novel way in order to produce a music video, reflecting on her creative process while doing so. This meant that three types of research data were created: creative assets and project files within the tools used; the final creative output (the music video); and reflective notes taken throughout the process. As a second step, the researcher structured the reflective notes into conclusions to be presented in this research article. The music video is published on YouTube, as is a brief documentary on the process (figs. 2 and 3 contain QR codes that can be scanned to view both videos). The overall creative process did not directly involve any other human participants. Figure 2. (left) A QR code linking to the music video, viewable . Figure 3. (right) A QR code linking to the documentary, viewable . Image credits: Kirsten Hermes (2023). https://youtu.be/cU8FLW8oZaI https://youtu.be/wdF7sQR5LzI Hermes | Augmented Creativity 109 The researcher is a self-produced electronic music artist under the name Nyokee, creating work that draws on a range of influences including video games, Japanese visual culture, hyperpop, Y2K aesthetics, surrealism and 1990s rave music. Her music incorporates stylistic references to artists such as Bjork, Grimes and Kerli, combining shimmery synths with whimsical ideas (fairies, witches, dolls, cyberpunk, spiritual elements) in an electronic pop format. Nyokee’s song “Ethereal” emerged from a creative process shaped by her experiences of synaesthesia and regular meditative practice. The track’s lyrics (“Touching the inbetween / We are mysterious / We are ethereal”) articulate a kind of pseudo-spiritual sci-fi sensibility, evoking themes of techno-human transcendence and inner energy. These motifs were further translated into visual form through the music video, which features an abstract cartoonish deity, footage of the artist singing and playing the violin, and a glowing, iridescent singing mask, all composited within digitally manipulated skies and science fiction-inspired landscapes (figs. 4 and 5). The aim was to create an immersive audiovisual world that oscillates between utopian and slightly uncanny tones, akin to digital dreamscapes found in video games like “No Man’s Sky”. While the final track is musically situated within electronic pop and progressive house, the creative emphasis was equally placed on the visual and conceptual aspects of the project from the outset. Despite limitations in her own visual design skillset, the artist-researcher planned both the music and visuals in tandem, viewing the project as an integrated multimedia work rather than a song accompanied by illustration. This approach aligns with existing scholarship on music videos that challenges the assumption of the music’s centrality in favour of a more multimodal or audiovisual analysis (Cook 1998; Vernallis 2004; Korsgaard 2017). In this view, music videos are not simply vessels for promoting songs, but complex artworks where image, sound, narrative, gesture and affect all interact. Figure 4. A screenshot from the music video, with the artist and abstract deity dancing in sync. Image credit: Kirsten Hermes (2023). Dancecult 17(1)110 Figure 5. A screenshot from the music video, with the artist playing violin in front of an abstract utopian landscape. Image credit: Kirsten Hermes (2023). Prior to the music video project, the artist had been exploring generative AI tools for approximately six months, producing concept art for a range of speculative projects in a similar envisioned style (an example is shown in fig. 6). This experimentation laid the groundwork for the aesthetic development of “Ethereal”. Figure 6. Examples of concept art generated in Midjourney. Image credit: Kirsten Hermes (2023). Hermes | Augmented Creativity 111 The final audiovisual output was presented in documentary form and shared publicly. In addition, the researcher conducted a semi-structured interview with a media executive (here referred to as ZT) from a company actively developing creative AI workflows. This conversation used the “Ethereal” video as a case study to deepen the understanding of co- creativity between human and machine.1 In the following paragraphs, the creative process will be explained in further detail. It should be noted that this research was conducted during 2022 and 2023, prior to the release of more advanced video-generation AI tools and other recent developments. As a result, some of the technologies now available were not accessible at the time of production, due to the inevitable lag between creative practice and academic publication. Midjourney is a popular generative AI tool built on image diffusion (a type of machine learning) that produces new images according to natural language processing (NLP), based on text entered by the user. For example, by typing “/imagine prompt”, a user can request any image, in any style (fig. 7 shows an example). They can further specify stylistic parameters, image format requirements or use existing images as inputs for new images. Users that are subscribed to the Midjourney platform own the copyright to all images they produce. Figure 7. An example of an image output generated from text in Midjourney. Image credit: Kirsten Hermes (2023). Dancecult 17(1)112 For the music video, the artist created still images of bright skies with rainbows and prism effects, as well as utopian landscapes. To transform these images into a movie, a second platform was used: Runway. Runway has multiple features, one of these being image interpolation. This function can connect still images into a moving picture by generating new image frames to be placed between the input images. Since the input images were not part of a realistic image sequence, the result was a series of unexpected but intriguing “watercolour” effects. At the time, video AI tools were still relatively underdeveloped, with limited control and often surreal or unstable outputs; by contrast, the more recent generation of video AI now allows for the creation of more cohesive, cinematic sequences. Had these tools been available, the artist would likely have created a more complete audiovisual narrative instead. The aforementioned scenes were overlayed with cloud stock footage in Premiere Pro, hence creating a collage of imagery, making the artist feel a greater sense of authorship than AI-only content would have allowed for. A further creative constraint was the importance of including footage of the artist herself within the video, which limited the extent to which fully AI-generated content could be used. Instead, she filmed herself singing, dancing and playing the violin, using an iPhone 11 Pro. In order to overlay this new footage with the sky imagery, the artist needed to remove the background in the videos. Figure 8. AI-powered background removal from raw footage: the highlighted area is retained. Image credit: Kirsten Hermes (2023). Hermes | Augmented Creativity 113 Successful background removal traditionally requires even lighting and a plain (ideally green) background (for greenscreening), neither of which the artist had access to. Instead, she used the AI greenscreen feature within Runway. This works by uploading footage and clicking on parts of the image that should be isolated, and Runway automatically detects outlines based on this (fig. 8). This feature works particularly well for footage with high foreground-background contrast. Since the footage was filmed in a residential house, many small keyframes and adjustments were required, and the result was not quite satisfactory. The artist creatively applied motion effects within Premiere Pro to disguise the errors. This, in turn, led to unintended but effective visual effects. From this, the back-and-forth between the human creator and AI co-creator becomes apparent. The third AI tool was Plask, an automatic motion capture tool. The artist created a new character in the 3D graphics software Blender: an abstract golden deity with gems as eyes (fig. 9). This character was rigged inside of Adobe Mixamo, an online platform that generates a “skeleton” for a character to allow graphics artists to apply movement. In order to animate this character, the artist used Plask to extract movement data from videos of herself dancing (fig. 10). This movement could be retargeted to the character in Blender. Figure 9. An abstract golden deity with gems as eyes, created in Blender 3D. Image credit: Kirsten Hermes (2023). Dancecult 17(1)114 Figure 10. Rigging in Plask. Image credit: Kirsten Hermes (2023). Plask offered 900 free credits per day, which meant that about 30 seconds of footage could be analysed at a time. This was less than the required length and hence the artist applied more movement effects and edits in Premiere Pro to make the video feel fluid. Another tool used was BlendARTrack, a phone application that extracts movement data from faces, which can be used within Blender (figs. 11 and 12). This was used to animate an iridescent mask miming the words of the second verse. Hermes | Augmented Creativity 115 Figure 11. A screenshot of the BlendARTrack interface. Image credit: Kirsten Hermes (2023). Dancecult 17(1)116 Figure 12. 3D mask, controlled via BlendARTrack. Image credit: Kirsten Hermes (2023). Additionally, the artist created landscape elements in Blender and drew further assets on an iPad. All assets were combined and edited in Premiere Pro. Findings and Discussion: The Impact of AI on the Creative Process In this section, the impact of AI on the creative process is analysed through a reflection on the music video workflow, framed within the literature and the discussion with ZT. The analysis will discuss timesaving through task delegation to AI (first subsection), the nature of AI as a creative collaborator (second subsection), challenges and affordances caused by the generative AI tools’ unpredictable outputs (third subsection), the narrow scope of tools (fourth subsection) and copyright concerns (last subsection). Hermes | Augmented Creativity 117 From workflow accelerator to creative actor: what is AI’s place in artistic practice? Traditionally, making art takes time, as does making the required equipment and materials (Becker 1982). The AI technologies significantly accelerated the artist’s creative process, enabling her to produce an entire music video in just five days, which is considerably faster than typical professional production timelines. Previously, the artist created all her work from scratch, whereby an earlier (AI-free) video took five months to complete. ZT explained that, similarly, in the industry context, the purpose of AI is increased efficiency and the enhancement of the creative process, allowing outputs to be created more cheaply and quickly. This unprecedented compression of traditionally human-intensive labour invites a shift in perspective, from viewing AI as a mere efficiency tool to interrogating its status as a potential creative agent. Developers of new AI tools often focus on replicating human skills (McClelland 2009), thereby placing emphasis on the machine becoming human-like (Dautenhahn 2007). This can be contextualised within the framework of computers as creative entities (de Mántaras 2020), a view that has led to the emergence of a subfield of Artificial Intelligence known as Computational Creativity (de Mántaras 2020). Similarly, from a posthuman perspective, as introduced in the second section, AI can be described as a creative actor. Stevens and Zabelina (2020) note that machines can simulate divergent thinking, a key trait of human creatives. AI creativity is typically evaluated based on the output (Ritchie 2007: 69). The Turing test (de Mántaras 2020) asks whether humans can distinguish between human and machine-made artifacts; if not, the machine is said to demonstrate intelligent behaviour. The Lovelace test assesses whether a machine can originate ideas it wasn’t explicitly programmed to produce (Wingström et al. 2022). Many contemporary AI tools pass both tests in principle, suggesting that parallels can indeed be drawn between AI and human creativity, however there are notable limitations to this because passing output-based tests does not guarantee process-based creativity. In the context of the case study, the AI tools functioned as creative assistants (for ideation, green screen removal and background creation). They were capable of producing technically complex and stylistically varied outputs that would previously have required intensive practice. As someone without professional training in visual design, this marked a dramatic shift in what was creatively achievable, giving the artist more creative possibilities especially in a style that would otherwise heavily rely on computer graphics. However, throughout the process, the AI also produced nonsensical results (unusable images and errors in the background removal), lacking the ability to reflect on these shortcomings. These unpredictable outputs are discussed further below. Overall, though AI can simulate artistic techniques, it cannot replicate human reasoning or act autonomously. According to de Mántaras (2020), algorithmic or information-processing systems alone cannot perform creative thinking. Dancecult 17(1)118 Overall, while AI tools may yield novel and unexpected outputs, they cannot transcend their programmed purposes. Jennings (2010) argues that for AI to move from apprentice to creator, it must be able to independently modify the standards it applies. It also cannot integrate diverse skills like humans (Hertzmann 2018) or meaningfully conceptualise goals and inputs (Gioti 2020). “AI lacks identity, feelings, the ability to give meaning to the outcomes it creates” (Wingström et al. 2022: 6) or reflect “the lived experience of the human” (Mazzone and Elgammal 2019: 8). Identifying valuable creative problems requires experience, knowledge and training (Sawyer 2012), which AI currently cannot achieve due to its narrow focus. A human creative director or curator is still needed to interpret and shape meaning from AI-generated content (Wingström et al. 2022). This certainly applied to the case study discussed here, where the artist needed to remain in a directorial role throughout (more on this in the next section). Beyond the assessment of AI as a creative actor, creativity is judged not only by the result, but also by the process (Kasof 1995), and therefore, De Mántaras (2020) raises the question of whether people value computer-generated artifacts as highly as human- made ones. This ties into Magni et al.’s (2024) observation that audiences often ascribe lower creativity to AI-generated works, and that creators often prefer to maintain a visible human imprint in their outputs (Newman et al. 2023). Informal observations on social media suggest that traditional craftsmanship is still appreciated (Susskind 2020), and many are disappointed to learn a work was “just another AI output”. On the other hand, imaginative ideas are sometimes valued more than technical execution. A fitting example is Marcel Duchamp’s L.H.O.O.Q., which gained acclaim for drawing a moustache on the Mona Lisa in 1991. The use of new technologies can itself become part of the artistic concept. Overall, AI may be regarded as a creative actor, but one that fundamentally differs from human creatives. Therefore, it is more useful to assess the complex interaction between AI and human creatives in collaborative settings, rather than studying AI creativity in isolation. AI as a Creative Collaborator AI can complement human work (Susskind 2020), and the concept of “co-creativity” has emerged to describe this (Wingström et al. 2022: 13). “Augmented intelligence” (Carroll 2021) offers an alternative framing that highlights the AI’s assistive role in enhancing human capabilities. The socio-technical assemblage of augmented intelligence brings forth “hybrid actors”, made up of both humans and AI. Through this, access to the domain becomes democratized and available to a greater number of people. As Becker (1982) argues, art is fundamentally social and organisational, not just aesthetic. The second section demonstrated the interconnectedness between creative actors within complex systems. Through AI, humans can brainstorm, prototype, and allow culture and history to “collide in a kaleidoscopic inspirational way” (Rosella 2022). AI supports Hermes | Augmented Creativity 119 this by enabling a cross-pollination of genres and ideas. ZT describes the AI-aided ideation process as “becoming unstuck”, “seeing the idea in context”, or “being taken in a direction you didn’t know was possible”. Hence, the AI becomes a tool for both preparation (as a creative incubator) and elaboration (executing ideas). The artist found it useful to test combinations of concepts and styles through AI, feeling as though she was accessing a large database of stock imagery through Midjourney, for example. In this way, the use of AI created a sense of indirect collaboration with a wider creative ecosystem, but without the logistical and financial complexities of hiring creatives, finding willing collaborators, or negotiating revisions, in turn allowing for practical access to styles and aesthetics that would have otherwise required a much larger team and budget. Redundance, repetitiveness and technical frustration could be reduced, which could also help induce flow states, as characterized by the full immersion in an experience (Csikszentmihalyi 1997). Unlike with human collaborators, there were no limits on the accepted number of revisions. On the other hand, the absence of human film makers meant that the artist needed to employ a greater degree of critical thinking within the domain of visual arts. Human-in-the-loop (HITL) refers to an AI design paradigm in which a human remains actively involved at one or more critical points of the system’s lifecycle, such as data curation, prompt formulation or output selection, so that machine outputs are always guided, corrected or endorsed by human judgment rather than being accepted automatically (Mosqueira-Rey et al. 2023). In this way, the artist’s role shifts from a traditional maker to a kind of creative director or orchestrator of multiple toolsets. Although many parts of the music video creation were outsourced, the artist retained full creative control, refining outputs through iterative prompting. Overall, for the AI collaboration to work effectively, the human must bring domain knowledge, artistic intention and critical judgment. Unlike human collaborators, AI cannot negotiate or reflect, it simply continues to generate. Unpredictable Outputs AI often makes serendipitous mistakes, resulting in near-perfect outputs with curious or striking flaws. These outcomes can resolve false dichotomies or provoke fresh thinking. The surreal, the contradictory, and the impossible become expressible through playful experimentation (fig. 13). Results may be brilliant or nonsensical, and unpredictability becomes part of the appeal. Social media is filled with such outputs: world leaders eating giant food, cartoon characters becoming real, or aliens posing for selfies. This aligns with definitions of creativity as improvisational play (Piplica et al. 2012) or open-ended exploration (Kantosalo and Toivonen 2016), where the AI acts as a playful or thought- provoking collaborator. Dancecult 17(1)120 Figure 13. The surprising result of blending a holiday photo of four hikers with the stock photo of an owl. Image credit: Kirsten Hermes (2023). Hermes | Augmented Creativity 121 Because of its unpredictability, the collaborative process with AI can be studied within the domain of group creativity. Here, emergent properties arise from interactions between group members, one of whom, in this case, is non-human (Glăveanu 2010). When AI misinterprets a prompt in an unexpectedly fascinating way, this misunderstanding can unlock new directions. However, this unpredictability can also make it very difficult to achieve the desired results. The most significant challenge in the creation of the music video was the difficulty for the artist to express her nuanced visual ideas through text prompts. She found that her imagination often outpaced what she could clearly verbalise, and this mismatch led to frustration. While this encouraged her to develop a greater artistic clarity and to be more deliberate and self-reflective, prompting still felt like a search rather than a more direct expression. The absence of social context or shared understanding added friction and introduced complexity, sometimes shifting the focus away from the creative vision toward technical problem-solving. The artist often asked: Why does this look like that? How do I change this detail? Should I try another tool entirely? These micro-decisions contributed to choice paralysis and a sense of creative overload. Prompting, or “AI whispering”, is now seen as a distinct creative skill (Susskind 2020), with tutorials and courses teaching users how to collaborate effectively with AI. This demonstrates that AI tools require entirely new skills to be used effectively. Interestingly, looking back, the artist notes that despite the challenges, earlier iterations of the AI tools sometimes felt more creatively inspiring than the newer ones. They were more unpredictable and idiosyncratic; qualities that prompted more experimentation. By contrast, current AI platforms feel more polished but also more homogenised, unless the user invests significant effort into deep prompting or fine-tuning. The novelty has worn off, and with it, some of the initial sense of play. Therefore, the unpredictable nature of generative AI has both advantages and disadvantages in the creative process. Narrow Scope of Tools AI tools often come with a narrow operational focus and are designed for highly specific tasks, which limits the creative possibilities unless the user experiments with multiple tools or platforms (especially for a multidisciplinary output such as a music video). AI, in its current form, functions as narrow AI rather than general AI (Gioti 2020), and is unable to generalise or cross domains in the way humans do. The narrow, style-specific training in many AI tools also reinforces the risk of homogenisation across creative outputs. As researchers have noted, generative AI models trained on similar datasets tend to reproduce dominant styles, which can limit diversity and innovation (Karpouzis 2024). Therefore, the artist needed to combine multiple tools, each serving a different function and operating within different interfaces. This approach can quickly become costly, given that AI platforms tend to be subscription-based. The artists’ workflow required continual exporting, reformatting and problem-solving between stages. Many tools (in this case Plask) additionally impose Dancecult 17(1)122 restrictions on how much can be generated, which creates stress and pressure when it is unclear how many attempts it will take to achieve the desired result. Due to all this, the final video did not fully match the vision expressed in the earlier mood board. The resulting aesthetic had a slightly jarring, collage-like feel that was not entirely satisfying to the artist. In that sense, the process revealed the gap between possibility and practicality: AI did not instantly “solve” creative challenges but rather reshaped them. Copyright Concerns Copyright remains a further complication. Many generative AI tools have been trained on artworks without the artists’ consent. According to ZT, in industry, the impetus is to train new tools solely using data that the company own, or that a specific individual has created, to ensure compliance with copyright law and to easily create new outputs in a very specific style. This approach was not available to the artist, and therefore, she was concerned about the ethical implications of AI training data and copyright. As a result, fully outsourcing the visuals to AI would have felt wrong. Instead, the process became a hybrid collage of AI-generated and self-produced material. This choice also helped to avoid an overtly “AI- generated” look. At times, the artist deliberately opted for more generic imagery to avoid the risk of unintentionally replicating another artist’s distinctive style through AI. While this felt like the safer ethical choice, it also meant the final video lacked some of the visual boldness and originality it might otherwise have had. Despite copyright concerns, borrowing and remixing are foundational to artistic practice. Jim Jarmusch famously claimed that “nothing is original” (Piazza 2015), while Ferguson (2021) observed that “everything is a remix”. Csikszentmihalyi noted that studying creativity in isolation from social and cultural context is futile (McIntyre 2011: 6). In this light, AI offers new mechanisms to access, filter, and recombine the domain. Bourdieu (1993) also argued that art must be understood as a manifestation of its field. Barthes (1977) questioned authorship entirely, shifting the focus to the receiver. This view supports Susskind’s (2020) suggestion that AI enables a more interactive reception of cultural material. In a 2022 interview, experimental composer Holly Herndon remarked: “These are just human systems that come out of human culture… Our biases are really baked into that stuff… It’s really a way for us to understand our shared human archive of music and read it in interesting new ways” (Hermes 2022: 185–92). These considerations add further complexity to the discussion around ethics and copyright, which are outside the scope of this article to resolve. Overall, AI is not a passive helper in the creative process, but an active and complex contributor. Reflecting on this, the artist noted that her creative process has gradually shifted. She now finds that the clarity of her vision must come before engaging with AI tools; otherwise, the tools begin to dictate the work’s direction. Hermes | Augmented Creativity 123 The artist shared her work on YouTube and Reddit and since the purpose of the work was to test AI tools in the creative process, this in itself generated interest and positive feedback. The socio-political implications of using AI spark debates and artist identities around technophobia, cyborgs and futuristic aesthetics. This also applies to the artist, whose work is inspired by human perceptions of technology. The collaboration with a non-sentient, but arguably intelligent force becomes a cultural phenomenon, which this article demonstrates. Conclusion This practice-based case study demonstrates that AI opens up new creative possibilities for music video production, particularly in achieving stylistic effects that would otherwise require advanced skills in computer graphics. For the artist (primarily a musician with no formal training in visual design), AI acted as a bridge between disciplines, enabling her to produce a visually rich video in a fraction of the time and with fewer traditional resources. However, while the AI tools accelerated ideation and execution, they also introduced a new set of creative, ethical and technical challenges. The use of generative AI raised concerns around authorship, copyright and artistic identity. Wary of being perceived as over-reliant on AI, or unintentionally reproducing copyrighted material, the artist deliberately opted for a more generic visual style and reintroduced her own, less-polished visual work to preserve a sense of authorship. This compromise, while ethically motivated, limited the creative boldness of the final output and reflected the unease many artists feel when navigating AI tools in their creative process. Practically, the process revealed how most AI tools remain narrow in scope, requiring the use of multiple subscription-based platforms, each with its own limitations (particularly for a complex multimedia project such as a music video). Issues such as background removal errors, limits on how much can be generated, high subscription costs and inefficient workflows highlighted the friction of working within a fragmented AI ecosystem, requiring constant exporting, reformatting, re-prompting and troubleshooting. Text prompting emerged as a critical creative skill. The artist often struggled to translate her nuanced visual ideas into effective prompts, leading to decision fatigue and creative overload. Though AI outputs sometimes sparked playful or unexpected insights, they just as often introduced noise, confusion, or disappointment. What began as creative freedom frequently became a form of technical problem-solving. Importantly, while AI meets some of the criteria of a creative actor (that is, the ability to produce novel and valuable outputs) it remains fundamentally different from human creators. AI can mimic stylistic techniques and draw from vast domain knowledge, but it lacks reasoning, critical thought and the capacity to form intentional meaning. As such, it cannot operate autonomously within the creative process. Human creators must still interpret, curate and develop the significance of what AI generates. This supports emerging ideas around co-creativity and hybrid actors, in which humans and machines form collaborative systems, each contributing in distinct ways. Dancecult 17(1)124 The unpredictability of generative AI outputs has both advantages and disadvantages in the creative process. AI functions well during the stages of preparation (as a creative incubator) and elaboration (as a technical executor), but it cannot substitute for the intentional, critical and reflective capacities of a human artist. Drawing on the systems view of creativity (Csikszentmihalyi 1997), this case illustrates how creative work arises not in isolation, but through complex, interdependent sociocultural dynamics, many of which are encoded into the data used to train AI models. Posthuman theory offers a valuable lens to explore these evolving relationships between human and machine, where agency is distributed and creativity is no longer confined to the individual. As traditional technical skills become increasingly replicable by machines, the value of creative ideas and conceptual originality becomes more important than ever. In this context, prompt craft, critical thinking and the ability to explore conceptual spaces are likely to become the defining skills of future creatives. As this project has shown, AI does not solve creative problems but rather reshapes them, requiring artists to adapt, direct and question more than ever before. Notes 1 ZT provided their informed written consent, and all research practices adhered to ethical approval conventions at the University of Westminster (London). References Amabile, Teresa M. 1996. Creativity in Context: Update to the Social Psychology of Creativity. Boulder, CO: Westview Press. Barthes, Roland. 1977. “The Death of the Author”. In Image, Music, Text, translated by Stephen Heath, 142–48. New York: Hill and Wang. Becker, Howard S. 1982. Art Worlds. Berkeley: University of California Press. Boden, Margaret A. 2004. The Creative Mind: Myths and Mechanisms. 2nd ed. London: Routledge. Bostrom, Nick. 2014. Superintelligence: Paths, Dangers, Strategies. Oxford: Oxford University Press. Bourdieu, Pierre. 1993. The Field of Cultural Production: Essays on Art and Literature. New York: Columbia University Press. Browne, Kieran. 2022. “Who (or What) Is an AI Artist?” Leonardo 55(2): 130–34. . Carroll, Noel. 2021. “Augmented Intelligence: An Actor-Network Theory Perspective”. Proceedings of the European Conference on Information Systems (ECIS) 2021, Paper 37. , (Accessed 4 August 2025). Clarke, Eric, and Nicholas Cook, eds. 2004. Empirical Musicology: Aims, Methods, Prospects. Oxford: Oxford University Press. . https://doi.org/10.1162/leon_a_02092 https://doi.org/10.1162/leon_a_02092 https://aisel.aisnet.org/ecis2021_rp/37 https://doi.org/10.1093/acprof:oso/9780195167498.001.0001 https://doi.org/10.1093/acprof:oso/9780195167498.001.0001 Hermes | Augmented Creativity 125 Cook, Nicholas. 1998. Analysing Musical Multimedia. Oxford: Oxford University Press. Costa, Paul T., Jr., and Robert R. McCrae. 2008. “The Revised NEO Personality Inventory (NEO-PI-R)”. In The SAGE Handbook of Personality Theory and Assessment, vol. 2, eds. Gregory J. Boyle, Gerald Matthews and Donald H. Saklofske, 179–98. Thousand Oaks, CA: SAGE. Csikszentmihalyi, Mihaly. 1988. “Society, Culture and Person: A Systems View of Creativity”. In The Nature of Creativity: Contemporary Psychological Perspectives, ed. Robert J. Sternberg, 325–29. Cambridge: Cambridge University Press. ———. 1997. Creativity: Flow and the Psychology of Discovery and Invention. New York: HarperPerennial. ———. 1999. “Implications of a Systems Perspective for the Study of Creativity”. In Handbook of Creativity, ed. Robert J. Sternberg, 313–35. Cambridge: Cambridge University Press. ———. 2013. Flow: The Psychology of Happiness. London: Penguin. Dautenhahn, Kerstin. 2007. “A Paradigm Shift in Artificial Intelligence: Why Social Intelligence Matters in the Design and Development of Robots with Human-Like Intelligence”. In 50 Years of Artificial Intelligence, eds. Max Lungarella, Fumiya Iida, Josh Bongard and Rolf Pfeifer, 288–302. Berlin: Springer. . de Mántaras, Ramón López. 2020. “Artificial Intelligence and the Arts: Toward Computational Creativity”. In The Next Step: Exponential Life, eds. Aubrey de Grey and John Rossiter. Madrid: BBVA Open Mind. Eindhoven, Jan E., and W. Edgar Vinacke. 1952. “Creative Processes in Painting”. Journal of General Psychology 47: 139–164. European Innovation Council and SMEs Executive Agency. 2024. “Artificial Intelligence and Copyright: Use of Generative AI Tools to Develop New Content”. European IP Helpdesk, 16 July. , (Accessed 8 August 2025). Ferguson, Kirby. 2021. Everything Is a Remix. , (Accessed 4 August 2025). Fook, Jan. 2011. “Developing Critical Reflection as a Research Method”. In Creative Spaces for Qualitative Researching: Living Research, eds. Joy Higgs, Angie Titchen, Debbie Horsfall and Donna Bridges, 55–64. Rotterdam: Sense Publishers. Furnham, Adrian. 2008. Personality and Intelligence at Work. London: Routledge. Gioti, Artemi-Maria 2020. “From Artificial to Extended Intelligence in Music Composition”. Organised Sound 25(1): 25–32. . Glăveanu, Vlad-Petre. 2010. “Principles for a Cultural Psychology of Creativity”. Culture & Psychology 16(2): 147–63. Guilford, Joy Paul 1950. “Creativity”. American Psychologist 5(9): 444–54. Guck, Marion A. 1994. “Rehabilitating the Incorrigible”. In Theory, Analysis and Meaning in Music, eds. Anthony Pople, 57–73. Cambridge: Cambridge University Press. Harris, Daniel, and Stacy Holman Jones. 2022. “A Manifesto for Posthuman Creativity Studies”. Qualitative Inquiry 28(5): 522–30. . https://doi.org/10.1007/978-3-540-77296-5_26 https://intellectual-property-helpdesk.ec.europa.eu/news-events/news/artificial-intelligence-and-copyright-use-generative-ai-tools-develop-new-content-2024-07-16-0_en https://intellectual-property-helpdesk.ec.europa.eu/news-events/news/artificial-intelligence-and-copyright-use-generative-ai-tools-develop-new-content-2024-07-16-0_en https://www.everythingisaremix.info/ https://doi.org/10.1017/S1355771819000438 https://doi.org/10.1177/10778004211066632 Dancecult 17(1)126 Hermes, Kirsten. 2022. Performing Electronic Music Live. London: Routledge. Hertzmann, Aaron. 2018. “Can Computers Create Art?” Arts 7(2): 18. . Ihde, Don. 1990. Technology and the Lifeworld: From Garden to Earth. Bloomington, IN: Indiana University Press. Jennings, Kyle E. 2010. “Developing Creativity: Artificial Barriers in Artificial Intelligence”. Minds and Machines 20(4): 489–501. . Kantosalo, Anna, and Hannu Toivonen. 2016. “Modes for Creative Human-Computer Collaboration: Alternating and Task-Divided Co-Creativity”. In Proceedings of the Seventh International Conference on Computational Creativity, eds. Anna Cardoso, Véronique Corruble and Francesco Ghedini, 77–84. Paris. Karpouzis, Kostas. 2024. “Plato’s Shadows in the Digital Cave: Controlling Cultural Bias in Generative AI”. Electronics 13(8): 1457. . Kasof, Joseph. 1995. “Explaining Creativity: The Attributional Perspective”. Creativity Research Journal 8(4): 311–66. Kaufman, James C. 2016. Creativity 101. New York: Springer. Kerrigan, Susan. 2014. “Accommodating Creative Documentary Practice within a Revised Systems Model of Creativity”. Journal of Media Practice 14(2): 111–27. . Korsgaard, Mathias. 2017. Music Video After MTV: Audiovisual Studies, New Media, and Popular Music. London: Routledge. Kumar, Vishesh, and Mike Tissenbaum. 2022. “Supporting Collaborative Classroom Networks through Technology: An Actor-Network Theory Approach to Understanding Social Behaviours and Design”. British Journal of Educational Technology 53: 1549–70. . Latour, Bruno. 1996. “On Actor-Network Theory: A Few Clarifications”. Soziale Welt 47(4): 369–81. ———. 2013. “Reassembling the Social: An Introduction to Actor-Network-Theory”. Journal of Economic Sociology 14(2): 73–87. . Lubart, Todd I. 2010. “Models of the Creative Process: Past, Present and Future”. Creativity Research Journal 13(3–4): 295–308. . Magni, Federico, Jiyoung Park and Melody M. Chao. 2024. “Humans as Creativity Gatekeepers: Are We Biased Against AI Creativity?” Journal of Business Psychology 39: 643–56. . Maguire, Kate. 2019. “Methodology as Personal and Professional Integrity: Research Designing for Practitioner Doctorates”. In Methodology as Personal and Professional Integrity: Research Designing for Practitioner Doctorates, eds. Carol Costley and John Fulton, 95–116. London: SAGE. . Mazzone, Marian, and Ahmed Elgammal. 2019. “Art, Creativity, and the Potential of Artificial Intelligence”. Arts 8(1): —. . https://doi.org/10.4324/9780367340759 https://doi.org/10.4324/9780367340759 https://doi.org/10.3390/arts7020018 https://doi.org/10.3390/arts7020018 https://doi.org/10.1007/s11023-010-9206-y https://doi.org/10.3390/electronics13081457 https://doi.org/10.1386/jmpr.14.2.111_1 https://doi.org/10.1386/jmpr.14.2.111_1 https://doi.org/10.1111/bjet.13274 https://doi.org/10.1111/bjet.13274 https://doi.org/10.17323/1726-3247-2013-2-73-87 https://doi.org/10.1207/S15326934CRJ1334_07 https://doi.org/10.1007/s10869-023-09910-x https://doi.org/10.1007/s10869-023-09910-x https://doi.org/10.4135/9781526453327.n6 https://doi.org/10.3390/arts8010026 Hermes | Augmented Creativity 127 McClelland, James L. 2009. “Is a Machine Realization of Truly Human-Like Intelligence Achievable?” Cognitive Computation 1: 17–21. . McIntyre, Phillip. 2011. Creativity and Cultural Production: Issues for Media Practice. London: Palgrave Macmillan. . Mosqueira-Rey, Eduardo, Elena Hernández-Pereira, David Alonso-Ríos, José Bobes-Bascarán and Ángel Fernández-Leal. 2023. “Human-in-the-Loop Machine Learning: A State of the Art”. Artificial Intelligence Review 56(4): 3005–54. . Mumford, Michael D., Michele I. Mobley, Roni Reiter‐Palmon, Charles E. Uhlman and Lesli M. Doares. 1991. “Process Analytic Models of Creative Capacities”. Creativity Research Journal 4: 91–122. . Newman, Michelle, Lidia J. Morris and Jin H. Lee. 2023. “Human-AI Music Creation: Understanding the Perceptions and Experiences of Music Creators for Ethical and Productive Collaboration”. In Proceedings of the 24th International Society for Music Information Retrieval Conference (ISMIR 2023), Milan: 80–88. Piazza, Sara. 2015. Jim Jarmusch: Music, Words and Noise. London: Reaktion Books. Piplica, Andreya, Christopher DeLeon and Brian Magerko. 2012. “Full-Body Gesture Interaction with Improvisational Narrative Agents”. In Intelligent Virtual Agents 2012, eds. Yukiko Nakano, Michael Neff, Ana Paiva and Marilyn Walker, 514–16. Lecture Notes in Computer Science 7502. Berlin: Springer. . Rhodes, Mel. 1961. “An Analysis of Creativity”. Phi Delta Kappan 42(7): 305–10. Ritchie, Graeme. 2007. “Some Empirical Criteria for Attributing Creativity to a Computer Program”. Minds and Machines 17(1): 67–99. . Rose, Gillian. 2017. “Posthuman Agency in the Digitally Mediated City: Exteriorization, Individuation, Reinvention”. Annals of the American Association of Geographers 107(4): 779–93. . Rosella, Francesca. 2022. “Video Killed the Radio Stars. Will the AI Kill the Designer?” Medium, 25 September. (Accessed 4 August 2025). Runco, Mark A., ed. 1994. Problem Finding, Problem Solving and Creativity. Norwood, NJ: Ablex. Runco, Mark A., and Selcuk Acar. 2012. “Divergent Thinking as an Indicator of Creative Potential”. Creativity Research Journal 24(1): 66–75. . Sarmiento, Johann W., and Gerry Stahl. 2008. “Group Creativity in Interaction: Collaborative Referencing, Remembering and Bridging”. International Journal of Human-Computer Interaction 24(5): 492–504. Sawyer, R. Keith. 2012. Explaining Creativity: The Science of Human Innovation. 2nd ed. Oxford: Oxford University Press. Shneiderman, Ben. 2022. Human-Centered AI. Oxford: Oxford University Press. https://doi.org/10.1007/s12559-009-9015-x https://doi.org/10.1007/s12559-009-9015-x https://doi.org/10.1057/9780230358614 https://doi.org/10.1007/s10462-022-10246-w https://doi.org/10.1007/s10462-022-10246-w https://doi.org/10.1080/10400419109534380 https://doi.org/10.1007/978-3-642-33197-8_63 https://doi.org/10.1007/s11023-007-9066-2 https://doi.org/10.1007/s11023-007-9066-2 https://doi.org/10.1080/24694452.2016.1270195 https://medium.com/@francesca_rosella/video-killed-the-radio-stars-6c8fe00e3015 https://medium.com/@francesca_rosella/video-killed-the-radio-stars-6c8fe00e3015 https://doi.org/10.1080/10400419.2012.652929 https://doi.org/10.1080/10400419.2012.652929 Dancecult 17(1)128 Simonton, Dean K. 2013. “What Is a Creative Idea? Little-c versus Big-C Creativity”. In Handbook of Research on Creativity, eds. Kerry Thomas and Janet Chan, 71–81. Cheltenham: Edward Elgar. . Smith, Brett. 2017. “Narrative Inquiry and Autoethnography”. In Routledge Handbook of Physical Cultural Studies, eds. Michael Silk, David L. Andrews and Holly Thorpe, 505–14. London: Routledge. Stein, Morris I. 1987. “Creativity Research at the Crossroads: A 1985 Perspective”. In Frontiers of Creativity Research: Beyond the Basics, ed. Scott G. Isaksen, 417–27. Buffalo, NY: Bearly Limited. Stevens, Carl E. Jr., and Darya L. Zabelina. 2020. “Classifying Creativity: Applying Machine- Learning Techniques to Divergent-Thinking EEG Data”. NeuroImage 219: 116990. . Susskind, Daniel. 2020. A World Without Work: Technology, Automation and How We Should Respond. London: Allen Lane. Thompson, Paul. 2015. Creativity and Collaboration in the Recording Studio: An Empirical Study. PhD dissertation, University of Liverpool. , (Accessed 4 August 2025). Vernallis, Carol. 2004. Experiencing Music Video: Aesthetics and Cultural Context. New York: Columbia University Press. Wingström, Rosa, Johanna Hautala and Riina Lundman. 2022. “Redefining Creativity in the Era of AI? Perspectives of Computer Scientists and New Media Artists”. Creativity Research Journal, 177–93. . Zhou, Mi, Vibhanshu Abhishek, Timothy Derdenger, Jaymo Kim, Kannan Srinivasan. 2024. “Bias in Generative AI”. arXiv preprint arXiv:2403.02726. , (Accessed 8 August 2025). https://doi.org/10.4337/9780857939814.00015 https://doi.org/10.1016/j.neuroimage.2020.116990 https://doi.org/10.1016/j.neuroimage.2020.116990 https://livrepository.liverpool.ac.uk/2033799/1/ThompsonPau_Oct2015_2033799.pdf https://livrepository.liverpool.ac.uk/2033799/1/ThompsonPau_Oct2015_2033799.pdf https://doi.org/10.1080/10400419.2022.2107850 https://arxiv.org/abs/2403.02726 https://arxiv.org/abs/2403.02726