Expert Eyes, Machine Hands

Abstract

Introduction Generative artificial intelligence (GenAI) has become one of the defining cultural technologies of the 2020s, transforming not only how content is produced but also how creativity itself is imagined. From text generation to image synthesis, GenAI blurs distinctions between professional and amateur practice, art and automation, originality and imitation. These shifts reach far beyond the studio or agency: they unsettle cultural hierarchies, redistribute expertise, and reconfigure what it means to author or evaluate creative work with scholars and practitioners highlighting both GenAI’s disruptive and enabling potential (Amankwah-Amoah et al.; Coffin; Hartmann et al.; Matthews et al.). As such, GenAI sits squarely at the intersection of technology and culture, making it vital to examine how creative practitioners experience and negotiate its affordances. These questions mirror broader cultural discussions about automation, expertise, aesthetics, and creativity in digital media, situating our research within contemporary debates about creativity in an algorithmic age. This article explores those negotiations through a collaborative autoethnography (Chang) in which creative professionals who are also educators worked with GenAI tools to develop an advertising campaign from concept to final artefact. By grounding our analysis in lived experience, we examine not only what GenAI can produce, but how it alters the creative process. Drawing on Csikszentmihalyi’s systems model of creativity and Bourdieu’s account of cultural capital, we situate our reflections within a wider theoretical conversation about expertise, taste, and technological mediation. These frameworks allow us to interrogate how GenAI participation affects the balance between human judgment and machine generation or between creative flow and algorithmic patterning. In doing so, we contribute to broader debates on the cultural implications of GenAI by identifying how expert practitioners sustain, redefine, and sometimes resist aesthetic standards in an era of algorithmic co-creation. The Positive Case for GenAI in Creative Practice The rapid rise of GenAI has intensified debates about its role in creative practice, particularly in creative industries where efficiency, cost-saving, and innovation are central (Cui et al.; Hartmann et al.; Matthews et al.). The issue is no longer simply whether GenAI is useful but how it reshapes the very conditions of creativity. By automating routine tasks, the technology frees practitioners to focus on more imaginative and rewarding work. As Amankwah-Amoah et al. highlight, GenAI’s transformative potential can produce “compelling” content without the burden of extensive “manual research and editing” (2), saving creative professionals time and energy that can be repurposed to more invigorating and empowering activities. Yet, as Csikszentmihalyi’s model suggests, creativity emerges from the dynamic interaction between domain (cultural knowledge), field (gatekeepers and experts), and individual (the creator). When GenAI intervenes in this system, it can rebalance these forces, enabling rapid idea generation but also introducing algorithmic biases that affect the domain of possible outcomes (Atkinson and Baker), or expectations that one can demand more for less (Matthews et al.). This theoretical perspective stresses that what feels like creative freedom can also be a reconfiguration of control. In the debates supporting GenAI, Hettrich et al. suggest that the technology can challenge human creators to think beyond the obvious by offering alternative directions that spark the creative imagination in ways they might not reach alone. Its adoption, then, is not merely a matter of convenience but a means of expanding creative possibility. What debates advocating for GenAI use emphasise is the fact that its use does not sideline the human creator. Instead, GenAI reshapes their role: Osadchaya et al. underline its function as collaborator and catalyst, enabling creatives to overcome the ‘tyranny of the blank page’, while Cui et al. highlight how it sparks novel ideas and accelerates responses to creative briefs. These perspectives make clear that GenAI should not be viewed as a competitor to human creativity but as a partner that expands its possibilities, challenging practitioners to reimagine both the process and the product of creative work. From a systems perspective, these collaborations demonstrate how GenAI alters what Csikszentmihalyi describes as a flow state: the absorption and momentum characteristic of deep creative engagement by accelerating iteration cycles and reframing what counts as exploration. Furthermore, audiences are also becoming less apprehensive about AI-generated content, with some studies showing they find AI images equally persuasive, especially in areas such as advertising and brand communication (Hartmann et al.). This suggests that combining GenAI with human expertise can enhance both the creative potential of the work and its acceptance by audiences. The Negative Case for GenAI in Creative Practice Despite GenAI’s growing value in creative processes, anxieties remain that it could displace professionals and dilute the aesthetic and quality standards traditionally associated with creative work that is exclusively human-made (Amankwah-Amoah et al.; Coffin). Scholars such as Runco contend that GenAI produces only “pseudo-creativity” (1), outputs that may appear effective but lack the contextual grounding and cultural depth of human innovation. Creative professionals echo this scepticism, observing that while AI can generate polished drafts, the raw outputs often feel “off” (Matthews et al. 15) or are “formulaic” outputs that undermine the intrinsic drive for experimentation and play that defines creative practice (Runco 1). These critiques highlight a deeper concern that, as GenAI becomes more widely adopted, the benchmarks by which creativity is judged may be lowered or distorted. Furthermore, from a Bourdieusian lens, the tension between technology and expertise reflects a struggle over cultural capital and taste: if expertise is what legitimises creativity, then the automation of creative labour threatens to redistribute symbolic power toward technological systems. The democratisation of generative tools, then, adds another layer of complexity. With platforms like DALL·E and Midjourney enabling anyone to produce seemingly professional content, the boundaries of what counts as valuable creativity risk becoming blurred (Ahmed and Ali; Chamakiotis and Panteli; Matthews et al.). Studies point to both the promise and the peril of this accessibility. Namely, GenAI lowers barriers to participation, but in doing so, it may also flood cultural spaces with superficial, poorly developed, distorted work (Ahmed and Ali), referred to openly as ‘AI slop’ (Smith and Southerton). The degradation of creative content could be inevitable because, as Chamakiotis and Panteli argue, GenAI privileges speed and quantity over originality and finesse. In professional settings, this tension becomes evident as GenAI enables less experienced practitioners to produce more polished work, but excessive dependence on the technology may hinder the growth of critical thinking and the refinement of creative skills (Hettrich et al.). Taken together, these dynamics raise important questions about how to preserve aesthetic standards and creative depth in an era where content creation is increasingly democratised. Accordingly, expertise remains crucial. Creativity has long been judged not only by novelty but also by peer recognition, audience reception, or what Csikszentmihalyi refers to as gatekeepers. Experts are well-placed to uphold aesthetic and quality standards because they understand the rules and practices that govern their discipline. As Bourdieu argued, those with cultural competence are deemed the most versed in encoding value in creative work and by extension maintain “explicit or implicit schemes of perception and appreciation” (2). Previous studies highlight this point, showing that while professionals view GenAI as helpful for generating ideas or producing early drafts, they stress that only a specialist can refine those outputs into work that is truly original and meaningful (DiStefano et al.; Matthews et al.). In this sense, the value of expertise lies not in resisting technology but in knowing how to use it effectively, drawing out its benefits while guarding against formulaic or tone-deaf outcomes, especially given the copyright infringement concerns that accompany GenAI-produced work. As scholars have contended, copyright conflict in creative industries stems from GenAI models using vast amounts of protected works as training data without permission (Lucchi; Torrance and Tomlinson). Supporters claim this ‘fair training’ is transformative, non-expressive, and socially beneficial (Lucchi; Torrance and Tomlinson), while critics see it as misappropriation that enables derivative, market-harming outputs (Lucchi; Torrance and Tomlinson). Given the loose legal restrictions that currently exist, it falls, in part, to creative experts to operate GenAI technology cautiously to prevent the misuse and exploitation of the creative work of others.   Approaching the Research Our research set out to investigate how we, as experts in visual communication, experience GenAI. We are not only creative professionals with years of experience in our respective industries, but also academics responsible for teaching undergraduates the skills required for effective visual communication, positioning us as both practitioners and gatekeepers of disciplinary standards. As part of our research, we aimed to understand not only the functional opportunities and challenges presented by these technologies, but also the evaluative judgments that can be made about AI-generated content. We operationalised this inquiry through a collaborative