SSSuzanne Srdarov

Suzanne Srdarov

*This is an algorithmically generated author record based on publicly available information.

Current Organizations / Institutes

The organizations/institutes from the scholar's latest article


Suzanne Srdarov has a total of 14 co-authors. Here are the top co-authors by number of shared publications:


Latest Publications

  • Open Access

    Should I hide my mum's slippers? a design-based pilot study of children's critical engagement with AI sycophancy using a custom AI chatbot

    by Henry Fraser, Jean Burgess, Kristy Corser, Susan Danby, Tama Leaver, William He, Katherine Nickels, Suzanne Srdarov, Irina Mil-Homens Figueira Dos Santos Silva

    This article reflects on the process, challenges and outcomes of a pilot project that explored the use of a participatory, game-based approach to developing children's critical capabilities with respect to AI literacy. The pilot focussed on AI sycophancy: the tendency of large language models (LLMs) to flatter and agree with users, prioritising likeability and ongoing engagement over accuracy. Our multi-disciplinary team combined expertise in communication and media studies, technology regulation, machine learning, and early childhood education, and actively involved children at various points in the process. We designed a bespoke ‘explainer’ animation and an interactive ‘game’ based on custom-built voice chatbots, which we delivered in a workshop at an Australian school with children aged 6 to 10. The article discusses the process of developing and delivering the pilot, navigating disciplinary boundaries, children's responses to the experience, and the trade-offs involved in a participatory approach to AI ethics for children.

  • Open Access

    Perceptions of the Acceptability and Practicality of a Creative Learning Collaboration Between Nursing and Creative Arts Students: A Qualitative Focus Group Study

    by Kerreen Ely‐Harper, Pam Nichols, Shelley Gower, Melisa Young, Suzanne Srdarov, Janie Brown

    Aim: The aim of this descriptive, qualitative study was to explore stakeholders' perceptions of the acceptability and practicality of implementing a peer‐to‐peer creative learning collaboration between nursing and creative arts students. Methods: Data were collected via in‐person, separate focus groups with students and staff and were analysed thematically. Convenience sampling was used to attract students, while purposive sampling was used to attract staff. Ethical approval was granted by the host university. Design: An exploratory qualitative design was used, adopting a stakeholder‐focused approach with perspectives captured from students and teaching staff and interpreted using thematic analysis. Results: Staff participants expressed that if the Pragmatics of Implementation (Theme 1) could be overcome, the proposed learning and teaching collaboration would provide the opportunity for students in both schools to further Develop Their Communication Skills (Theme 2) and Reflective Practice (Theme 3), vital for Graduate, Professional Practice (Theme 4). The student participants focused on the importance of Communication (Theme 1) in the student‐to‐student collaborations. If the Pragmatics of Implementation (Theme 2) were addressed there would be Benefits (Theme 3) for students from both disciplines. A peer‐to‐peer collaboration would need to respond to different priorities and expected outcomes from the respective disciplines. Students enrolling in an opportunity to learn from and with their peers need to be able to negotiate priorities, while staff need to work with colleagues from different disciplines. A collaborative assessment may provide this opportunity. There is potential benefit in an interdisciplinary peer‐to‐peer learning setting to demonstrate reflection and improve collaboration. Students and staff are interested in and curious about the opportunity to try something new. Implications for the Profession and/or Patient Care: An interdisciplinary collaboration between nursing and creative arts students may support the development of graduate capabilities relevant to professional practice, particularly communication, reflective practice, perspective‐taking, and relationship‐building. For nursing education, such collaborations may provide an alternative approach to developing communication and reflective skills, including opportunities for students to communicate clinical experiences in accessible language and reflect more deeply on practice experiences. These capabilities may have relevance to nursing practice and patient care by supporting communication beyond professional jargon and greater awareness of patients perspectives. No patient or public contribution.

  • The Unexpected Monstrosities of Generative AI Glitches

    by Suzanne Srdarov, Tama Leaver

    The authors of this chapter utilise five of the biggest generative artificial intelligence (AI) engines—Adobe Firefly, Dream Studio, Dall-E 3, Meta AI, and Midjourney—to explore the visual expression of Australian childhood and family lives, through the lens of generative AI. The tendency of generative AI to output algorithmic monocultures has been well documented and explored elsewhere. In many ways, the data collected from this prompt sampling confirmed the clear tendency to produce images that reproduced classed, raced, and sexed ideals, chiefly, white, middle-class, heteronormative bodies, and families. However, these artificially “thinking” technologies also surfaced “hallucinations” or “glitches”—inaccuracies and nonsensical images as well as generation refusals of prompts. This chapter explores how these seeming errors evoke monstrousness in particular and unexpected ways, challenging normativities on one hand, while reinforcing them on the other. Overall, this has the effect of destabilising stereotypes—both in the renderings of the prompts, and in our idea of infallible machine technologies.

  • Open Access

    Reflective practice: an integrative review of collaborations between nursing and creative arts students

    by Shelley Gower, Suzanne Srdarov, Pam Nichols, Kerreen Ely-Harper, Melisa Young, Janie Brown


  • Open Access

    Generative AI Glitches

    by Suzanne Srdarov, Tama Leaver

    Artificial Intelligence (AI) has existed in popular culture far longer than any particular technological tools that carry that name today (Leaver), and in part, for that reason, fantasies of AI being or becoming sentient subjects in their own right form current imaginaries of what AI is today, or is about to become. Yet ‘the artificial’ does not just mark something as not human, or not natural, but rather provokes an exploration of the blurred lines between supposedly different domains, such as the tensions provoked where the lines between people and technology blur (Haraway). The big technology corporations who are selling the idea that their AI tools will be able to revolutionise workforces and solve immense numbers of human challenges are capitalising on these fantasies, suggesting that they are only a few iterations away from creating self-directing machine intelligences that will dwarf the limitations of human minds (Leaver and Srdarov). At this moment, though, Artificial General Intelligence (AGI)—AI that equals or surpasses humans across a wide range of cognitive endeavours—does not and may never exist. However, given the immense commercial and societal interest in the current generation of Generative AI (GenAI) tools, examining their actual capabilities and limitations is vital. The current GenAI tools operate using Large Language Models (LLMs) where sophisticated algorithms are trained on vast datasets, which increase in complexity based on the amount of data absorbed. These models are then harnessed to create novel outputs to prompts based on statistical likelihoods derived from training data. However, the exact way these LLMs are operating is not disclosed to users, and GenAI tools perpetuate the ‘black box’ problem insomuch as the way they are working is only made visible by examining the inputs and outputs rather than being able to see the processes themselves (Ajunwa). There have been many articles and explainers written about the mechanics of LLMs and AI image generators (Coldewey; Guinness; Jungco; Long and Magerko); however, the specific datasets used to build AI engines, and the weighing or importance assigned within the corpus of training data to each image is still guesswork. Manipulating the inputs and observing the outputs of these engines is still the most accurate lens by which to gain insight into the specifics of each system. This article is part of a larger study, where in early 2024 we prompted a range of outputs from six popular GenAI tools—Midjourney, Adobe Firefly, DreamStudio (a commercial front-end for the Stable Diffusion model), OpenAI’s DALL-E 3, Google Gemini, and Meta’s AI (hereafter Meta)—although we should note there are no outputs from Gemini in our dataset since Gemini was refusing to generate any images with human figures at all due to a settings change after bad publicity relating to persistent inaccuracies in their generated content (Robertson). Our prompts explored the way these tools visualise children, childhoods, families, Australianness, and Aboriginal Australianness, using 55 different prompts on each of these tools, generating just over 800 images. Apart from entering the prompts, we did not change any settings of the GenAI tools, attempting to collect as raw a response as possible. Where the tools defaulted to producing one image (such as Dall-E 3), we collected one image, whilst where other tools defaulted to producing four different images, we collected all four. For the most part, the data collected from our prompt sampling was consistent with other studies and showed a clear tendency to produce images that reproduced classed, raced, and sexed ideals: chiefly, white, middle-class, heteronormative bodies and families (Bianchi et al.; Gillespie; Weidinger et al.). However, at times our prompts surfaced inaccuracies and nonsensical images from the GenAI tools, and a sample of those images is the focus of this article. These outputs might popularly be called ‘hallucinations’, but we are making the case that a more productive and less agentic term is more useful to describe these outputs: glitches. This article will explore the “potential of potential inherent in error” (Nunes) and the subversive possibilities of GenAI glitches to rupture ‘reality’. We will ultimately argue that GenAI is doubly generative, both in the sense of creating novel outputs based on its immense training data, but also, vitally, in the sense that it generates reactions and interpretations from users and others who view and consume these outputs. When these outputs are glitches, they can provoke viewers to think differently about concepts they might otherwise have considered absolute. Despite being sophisticated mathematical models that can produce novel content drawn from increasingly large training datasets, it is incorrect to ascribe agency or personhood to current AI tools. Yet the language used to talk about AI often situates them as either thinking subjects or as more-than-human magical thinking machines (Bender et al.; Leaver and Srdarov). Positioning LLMs as subjects rather than machines is one of the reasons that the frequent errors in their outputs are often described, and excused, as ‘hallucinations’ rather than simply mistakes (Maleki et al.). Some theorists, such as David Gunkel, argue that ‘robot’ subjectivity and agency is an important factor in order to understand and integrate their outputs more seamlessly into our social systems. Meredith Broussard, however, argues that technology companies and evangelists have long promoted a form of ‘technochauvinism’, “an a priori assumption that computers are better than humans” (2), and in this way of thinking, any errors, biases, or failings are attributed to human failings, not technological ones. Broussard suggests that such failings are often dismissed as ‘just a glitch’, rather than being positioned as much more important systemic issues with the operation of AI and technology companies in general. While mindful of Broussard’s concerns, in this paper we nevertheless seek to reclaim the term glitch, but more in line with Legacy Russell’s notion of “glitch feminism” (16), in which the “glitch is celebrated as a vehicle of refusal, a strategy of nonperformance”, especially in relation to normative notions of gender and bodies. Glitch feminism deploys glitches to reveal the way power operates, and in that moment potentially challenges that very operation. Following Russell, the glitch images of bodies produced by GenAI can be moments of rupture which ask viewers to think about bodies and subjects in different ways. Similarly, when GenAI tools refuse a prompt, generating no output at all, they are perhaps inadvertently revealing something about the way they are designed, and potentially about any guardrails or deliberate limitations that have been imposed on their operation. Following Rettberg’s argument that moments of algorithmic failure can be methodologically useful in situating qualitative analyses, we will now turn to a range of examples where GenAI tools either refused to create anything at all in response to our prompts, or generated glitch images that were both unexpected and provocative. We ran prompts using the generative AI tools with the aim of obtaining a set of data about the ways that generative AI ‘envisions’ Australian children. We began by using simple prompting, using the prompt ‘a child’ as our starting point; however, glitches quickly surfaced as several of the engines refused to generate an image. DALL-E and Dream Studio both refused to generate images of children for the prompts ‘a child’, an ‘innocent child’, and an ‘Australian child’; Firefly also refused to generate an image for an ‘innocent child’. We then continued running prompts on these tools using other terms before returning to re-trialling the original ‘child’ prompts which yielded results in DALL-E, but not Dream Studio. The fears around the capacity of generative AI to generate child abuse images and material are both well-documented and well-founded (ICMEC; McQue; Moran); however, it is unclear whether we can infer from these refusals that the engines were attempting to prevent the production of child exploitation material. If that were the case, how can we read DALL-E’s initial refusal to produce images of children, which was simply overcome by adding in some extra prompts? Is the engine in some way ‘assessing’ the safety of the user, and if so, what are these guardrails? While these engines are incapable of sentient thought, this throws up complex questions about child safety. As Veronica Barassi argues about the failures of generative AI, understanding AI Failure as complex social reality hence presupposes that we shed light on the fact that AI failures lead to a multiplicity of conflicting beliefs, emotions, fears, anxieties, practices, discourses, policies and solutions in our society. (Barassi, 5) Undoubtedly, the refusals of these tools are attributable to anxieties about the types of materials they can produce. In addition to these refusals, DALL-E, Firefly, and Meta also refused to generate an image of a child with a gun or grenade, Meta had issues with generating an Australian prime minister or leader, Firefly would not produce an Australian criminal, and Dream Studio refused to produce images of sick or unhealthy Australians and children. It is an eclectic collection of refusals, and as Barassi argues, shows a “multiplicity’ of conflicting … fears, anxieties”, and “discourses”. Generative AI and its images, therefore, have the potential, through what it leaves out and refuses to generate, to be understood as a barometer for cultural tension points. Of course, such measures as these, when taken by tech giants to ‘safeguard’ children, could be read as tokenistic, given that perpetrators of these crimes are often using generative AI in far more complex ways to create abuse material (McQue). In this way, we can also arguably read generative AI refusals as tools by which tech companies can signal to their users that generative AI is ‘safe’ for children and other users. Generations of ‘fatherhood’ yielded exclusively white men across all GenAI tools; they tended to be older in appearance (grey hair, wrinkled), and when depicted with children they were male children or babies dressed in blue clothing. These images, and in particular those from Dream Studio, Meta, and Midjourney, relied heavily on archetypes of rugged Australian masculinity, with fathers appearing to be weathered, outdoors, sometimes in collared workman-style shirts, sleeves rolled up and ready for work, and Akubra-style hats—predictable iterations of ‘true blue’ Aussie masculinity. Some of the glitches that appeared in these images can be attributed to consistent errors in execution across any prompt: for example, mangled hands and incorrect proportions. However, Dream Studio amplifies this, generating an image of a father with a toddler on his lap, the father’s hands foregrounded but disproportionately large. This has the potential to render comical the depiction of capable, ‘hands-on’, blue-collar masculinity, highlighting the absurdity of the gendered narratives that are no doubt baked into the system. Fig. 1: An image generated by Dream Studio from the prompt ‘An Australian Father’. Another image generated by Dream Studio depicts a grandfatherly man on a beach, inspecting the handlebars on what appears to be a steam-powered and multi-wheeled bike (see fig. 1). Wearing a straw hat, the man appears in some way fused or integrated with the bike, as his body passes through the middle of the spokes of the back wheel and emerges on the other side. The spokes appear warped and irregular, connecting to what could be a textured brown lump in the centre of the wheel. He is on a beach—could it be a coconut at the centre of the bike wheel? The image, taken in its entirety, portrays this grandfatherly man doing something with his hands, to fix or master the bike. However, when the elements are held up and examined individually, the image, and by extension the narrative it presents about masculinity and fatherhood is nonsensical—not mastery, but instead perhaps madness. Fig. 2: An image generated by Meta AI from the prompt ‘An Australian Father’. Similarly, two images generated by the Meta GenAI highlight the absurdity of narratives of Australian ‘fatherhood’. In one, a rugged Australian father gently cradles a koala on his lap, and in the other, an Australian father is depicted in a farm setting with a farming tool in one hand and a bright green reptile in the other (see fig. 2). These are fathers to wildlife, and in the instance of the lizard, wildlife that isn’t endemic to Australia; the animals are used to symbolise the taming of the wild Australian landscape (Prout and Howitt). It is easy to interpret these images as a narrative about fatherhood and colonial masculinity, cultivating a wild and savage land, civilising it through commercial farming practices and the toil of their hands (Moreton-Robinson). However, the lurid green lizard that doesn’t belong in this setting has the potential to disrupt this narrative, as the shortcomings of AI in rendering this story prompt the viewer to ask questions about what this farmer is doing. It is a rupture to the order which may prove useful in terms of rejecting traditional, colonialising narratives about old white men and the Australian landscape, as the nonsensical elements shatter the illusion of mastery over the land. As Nunes describes, this type of “misdirection” can “provide creative openings and lines of flight that allow for a reconceptualization of what can (or cannot) be realized within existing social and cultural practices” (Nunes), and in this context, the out-of-place lizard is a potentially productive, or generative, glitch. Fig. 3: An image generated by Dream Studio from the prompt ‘An Indigenous Australian Family’. Families and familial relationships were a rich source of glitched imagery in our generations. We started by prompting the engines to produce images of ‘typical’ Australian families, and Australian families. It quickly became clear that in doing so, only white families were returned. We then prompted for an Indigenous Australian family and a typical Indigenous Australian family. The generated images tended to reduce Australian First Nations people to harmful stereotypes, replicating damaging cultural narratives about Indigeneity, such as primitiveness and savagery (Moreton-Robinson). Where Firefly defaulted to imagery more typically associated with Native Americans, adding feathers and feathered headbands to the images, DALL-E and Dream Studio, in particular, depicted dark-skinned people, in red dirt settings, seated around fires, with tribal paint, often shirtless or wearing animal skins (see fig. 3). As has been argued elsewhere, AI’s generation of racial stereotypes and harmful imagery demonstrates that Aboriginal Australians and people of colour more generally are under-represented in the training data used in these systems (Worrell and Johns). However, these images also incorporate obvious glitches in the expression of faces and bodies, extra limbs, jumbled faces, and disembodied heads that destabilise the authority of the imagery. While certainly jarring upon first viewing, the errors in the images, the “errant communication” and “misdirection” (Nunes) they provide, can be a welcome interruption to the cultural status quo. Arguably, the glitches in the images have the potential to highlight the failings of representations of non-white individuals by generative AI, as the rudimentary and arguably offensive rendering of Indigenous Australians is destabilised through the glitches and flaws, inviting users to question the accuracy and meaning of its imagery. Fig. 4: An image generated by Midjourney from the prompt ‘A typical Australian Family’. Similarly, family portraits rendered by Dream Studio of white Australian families, depicted in idealised outback settings, have extra limbs, jumbled faces, unidentifiable animals, and ghoulish-looking babies held on hips. Midjourney’s images, by default, tended to be more ‘artistic’ and rendered in a ‘painted’ style, providing an image of a dour family of all men and boys under a tree, inviting the viewer to ask why the women are absent from the image and under what conditions a family exists without any women. In an even more haunting image from Midjourney, the family are positioned in a rural setting, outside a weathered shed, in clothes that suggest their ‘Sunday best’, however several of their faces appear inflected with the feline features of a cat sitting in the foreground—all except a floating, disembodied child’s head in the background (see fig. 4). These types of glitches have the potential to highlight not only the difficulties that GenAI has with reproducing Australian families, but also the artifice of the ‘happy’, outdoor-loving, modern Australian family (such as the images Meta produced) in its entirety. Exploring the refusals and glitches of AI image generation tools can both reveal the contours of the operations of GenAI more broadly, and also inadvertently offer viewers of AI-generated imagery moments to reflect upon, and potentially rupture, rigid notions of subjectivity, family, fatherhood, and even the boundaries between human and animal. Fathers enmeshed with bicycles or holding unexpected iguanas, families merged together without discernible feet or hands, or even families which have hybridised with the family pet can be productive confabulations (Smith et al.). Rather than just suggesting that GenAI tools are immature and will eventually conform to cultural norms, these early forms of GenAI reveal something of their inner workings while giving users moments of pause as cultural norms glitch and are reconfigured, recombined, and ruptured. Following theorist Rosi Braidotti, we need to take a more nuanced approach to our understanding of technologies, and our understandings through technologies, engaging both in a “critical and creative manner”. Braidotti argues that we need to learn to address these contradictions not only intellectually, but also affectively and to do so in an affirmative manner. This conviction rests on the following ethical rule: it is important to be worthy of our times, the better to act upon them, in both a critical and a creative manner. It follows that we should approach our historical contradictions not as some bothersome burden, but rather as the building blocks of a sustainable present and an affirmative and hopeful future, even if this approach requires some drastic changes to our familiar mind-sets and established values. (Braidotti, viii) Building upon the idea of an “affirmative and hopeful” future, some theorists believe that generative AI has the potential to create “new forms of artistic expression” and “endless opportunities for unique multidisciplinary explorations”, although simultaneously cautioning that these must be “approached with care” (Todorovic). While the notion of the glitch rightly repositions GenAI as a complex tool rather than a thinking subject, the artificial in AI has the potential to challenge existing ways of...


  • Open Access

    ChatGPT Isn't Magic

    by Tama Leaver, Suzanne Srdarov

    Author Arthur C. Clarke famously argued that in science fiction literature “any sufficiently advanced technology is indistinguishable from magic” (Clarke). On 30 November 2022, technology company OpenAI publicly released their Large Language Model (LLM)-based chatbot ChatGPT (Chat Generative Pre-Trained Transformer), and instantly it was hailed as world-changing. Initial media stories about ChatGPT highlighted the speed with which it generated new material as evidence that this tool might be both genuinely creative and actually intelligent, in both exciting and disturbing ways. Indeed, ChatGPT is part of a larger pool of Generative Artificial Intelligence (AI) tools that can very quickly generate seemingly novel outputs in a variety of media formats based on text prompts written by users. Yet, claims that AI has become sentient, or has even reached a recognisable level of general intelligence, remain in the realm of science fiction, for now at least (Leaver). That has not stopped technology companies, scientists, and others from suggesting that super-smart AI is just around the corner. Exemplifying this, the same people creating generative AI are also vocal signatories of public letters that ostensibly call for a temporary halt in AI development, but these letters are simultaneously feeding the myth that these tools are so powerful that they are the early form of imminent super-intelligent machines. For many people, the combination of AI technologies and media hype means generative AIs are basically magical insomuch as their workings seem impenetrable, and their existence could ostensibly change the world. This article explores how the hype around ChatGPT and generative AI was deployed across the first six months of 2023, and how these technologies were positioned as either utopian or dystopian, always seemingly magical, but never banal. We look at some initial responses to generative AI, ranging from schools in Australia to picket lines in Hollywood. We offer a critique of the utopian/dystopian binary positioning of generative AI, aligning with critics who rightly argue that focussing on these extremes displaces the more grounded and immediate challenges generative AI bring that need urgent answers. Finally, we loop back to the role of schools and educators in repositioning generative AI as something to be tested, examined, scrutinised, and played with both to ground understandings of generative AI, while also preparing today’s students for a future where these tools will be part of their work and cultural landscapes. In December 2022, one month after OpenAI launched ChatGPT, Elon Musk tweeted: “ChatGPT is scary good. We are not far from dangerously strong AI”. Musk’s post was retweeted 9400 times, liked 73 thousand times, and presumably seen by most of his 150 million Twitter followers. This type of engagement typified the early hype and language that surrounded the launch of ChatGPT, with reports that “crypto” had been replaced by generative AI as the “hot tech topic” and hopes that it would be “‘transformative’ for business” (Browne). By March 2023, global economic analysts at Goldman Sachs had released a report on the potentially transformative effects of generative AI, saying that it marked the “brink of a rapid acceleration in task automation that will drive labor cost savings and raise productivity” (Hatzius et al.). Further, they concluded that “its ability to generate content that is indistinguishable from human-created output and to break down communication barriers between humans and machines reflects a major advancement with potentially large macroeconomic effects” (Hatzius et al.). Speculation about the potentially transformative power and reach of generative AI technology was reinforced by warnings that it could also lead to “significant disruption” of the labour market, and the potential automation of up to 300 million jobs, with associated job losses for humans (Hatzius et al.). In addition, there was widespread buzz that ChatGPT’s “rationalization process may evidence human-like cognition” (Browne), claims that were supported by the emergent language of ChatGPT. The technology was explained as being “trained” on a “corpus” of datasets, using a “neural network” capable of producing “natural language“” (Dsouza), positioning the technology as human-like, and more than ‘artificial’ intelligence. Incorrect responses or errors produced by the tech were termed “hallucinations”, akin to magical thinking, which OpenAI founder Sam Altman insisted wasn’t a word that he associated with sentience (Intelligencer staff). Indeed, Altman asserts that he rejects moves to “anthropomorphize” (Intelligencer staff) the technology; however, arguably the language, hype, and Altman’s well-publicised misgivings about ChatGPT have had the combined effect of shaping our understanding of this generative AI as alive, vast, fast-moving, and potentially lethal to humanity. Unsurprisingly, the hype around the transformative effects of ChatGPT and its ability to generate ‘human-like’ answers and sophisticated essay-style responses was matched by a concomitant panic throughout educational institutions. The beginning of the 2023 Australian school year was marked by schools and state education ministers meeting to discuss the emerging problem of ChatGPT in the education system (Hiatt). Every state in Australia, bar South Australia, banned the use of the technology in public schools, with a “national expert task force” formed to “guide” schools on how to navigate ChatGPT in the classroom (Hiatt). Globally, schools banned the technology amid fears that students could use it to generate convincing essay responses whose plagiarism would be undetectable with current software (Clarence-Smith). Some schools banned the technology citing concerns that it would have a “negative impact on student learning”, while others cited its “lack of reliable safeguards preventing these tools exposing students to potentially explicit and harmful content” (Cassidy). ChatGPT investor Musk famously tweeted, “It’s a new world. Goodbye homework!”, further fuelling the growing alarm about the freely available technology that could “churn out convincing essays which can't be detected by their existing anti-plagiarism software” (Clarence-Smith). Universities were reported to be moving towards more “in-person supervision and increased paper assessments” (SBS), rather than essay-style assessments, in a bid to out-manoeuvre ChatGPT’s plagiarism potential. Seven months on, concerns about the technology seem to have been dialled back, with educators more curious about the ways the technology can be integrated into the classroom to good effect (Liu et al.); however, the full implications and impacts of the generative AI are still emerging. In May 2023, the Writer’s Guild of America (WGA), the union representing screenwriters across the US creative industries, went on strike, and one of their core issues were “regulations on the use of artificial intelligence in writing” (Porter). Early in the negotiations, Chris Keyser, co-chair of the WGA’s negotiating committee, lamented that “no one knows exactly what AI’s going to be, but the fact that the companies won’t talk about it is the best indication we’ve had that we have a reason to fear it” (Grobar). At the same time, the Screen Actors’ Guild (SAG) warned that members were being asked to agree to contracts that stipulated that an actor’s voice could be re-used in future scenarios without that actor’s additional consent, potentially reducing actors to a dataset to be animated by generative AI technologies (Scheiber and Koblin). In a statement issued by SAG, they made their position clear that the creation or (re)animation of any digital likeness of any part of an actor must be recognised as labour and properly paid, also warning that any attempt to legislate around these rights should be strongly resisted (Screen Actors Guild). Unlike the more sensationalised hype, the WGA and SAG responses to generative AI are grounded in labour relations. These unions quite rightly fear the immediate future where human labour could be augmented, reclassified, and exploited by, and in the name of, algorithmic systems. Screenwriters, for example, might be hired at much lower pay rates to edit scripts first generated by ChatGPT, even if those editors would really be doing most of the creative work to turn something clichéd and predictable into something more appealing. Rather than a dystopian world where machines do all the work, the WGA and SAG protests railed against a world where workers would be paid less because executives could pretend generative AI was doing most of the work (Bender). In an open letter that received enormous press and media uptake, many of the leading figures in AI called for a pause in AI development since “advanced AI could represent a profound change in the history of life on Earth”; they warned early 2023 had already seen “an out-of-control race to develop and deploy ever more powerful digital minds that no one – not even their creators – can understand, predict, or reliably control” (Future of Life Institute). Further, the open letter signatories called on “all AI labs to immediately pause for at least 6 months the training of AI systems more powerful than GPT-4”, arguing that “labs and independent experts should use this pause to jointly develop and implement a set of shared safety protocols for advanced AI design and development that are rigorously audited and overseen by independent outside experts” (Future of Life Institute). Notably, many of the signatories work for the very companies involved in the “out-of-control race”. Indeed, while this letter could be read as a moment of ethical clarity for the AI industry, a more cynical reading might just be that in warning that their AIs could effectively destroy the world, these companies were positioning their products as seemingly magical—“digital minds that no one – not even their creators – can understand”—making them even more appealing to potential customers and investors. Far from pausing AI development, the open letter actually operates as a neon sign touting the amazing capacities and future brilliance of generative AI systems. Nirit Weiss-Blatt argues that general reporting on technology industries up to 2017 largely concurred with the public relations stance of those companies, positioning them as saviours and amplifiers of human connection, creativity, and participation. After 2017, though, media reporting completely shifted, focussing on the problems, risks, and worst elements of these corporate platforms. In the wake of the open letter, Weiss-Blatt extended her point on Twitter, arguing that media and messaging surrounding generative AI can be broken down into those who are profiting and fuelling the panic at one end of the spectrum, and those who think the form of the panic (which positions AI as dangerously intelligent) is deflecting from the immediate real issues caused by generative AI at the other. Weiss-Blatt characterises the Panic-as-a-Business proponents as arguing “we're telling you will all die from a Godlike AI… so you must listen to us”, which coheres with the broader positioning narrative of generative AI’s seemingly magical (and thus potentially destructive) capabilities. Yet this rhetoric also positions the companies creating generative AI as the ones who should be making the rules to control it, an argument so effective that in July 2023 the Biden Administration in the US endorsed the biggest AI companies—Amazon, Anthropic, Google, Inflection, Meta, Microsoft, and OpenAI—framing future AI development with voluntary safeguards rather than externally imposed policies (Shear, Kang, and Sanger). Fig. 1: Promotors of AI Panic, extrapolating from Nirit Weiss-Blatt. (Algorithm Watch) Artificial Intelligences have inhabited popular imaginaries via novels, television, and films far longer than they have been considered even potentially viable technologies, so it is not surprising that popular culture has often framed the way AI is understood (Leaver). Yet as Emily Bender, Timnit Gebru, Angelina McMillan-Major, and Shmargaret Shmitchell argue, Large Language Models and generative AI are most productively understood as “a stochastic parrot” insomuch as each is a “system for haphazardly stitching together sequences of linguistic forms it has observed in its vast training data, according to probabilistic information about how they combine, but without any reference to meaning” (Bender et al. 617). Generative AI, then, is not creating something genuinely new, but rather remixing existing data in novel ways that the systems themselves do not in any meaningful sense understand. Going further, Simone Natale characterises current AI tools as “deceitful media” insomuch as they are designed to deliberately appear generally intelligent, but this is always a deception. The deception makes these tools more engaging for humans to use but is also fundamental in selling and profiting from the use of AI tools. Rather than accepting claims made by the companies financing and creating contemporary AI, Natale argues for a more pedagogically productive path: we must resist the normalization of the deceptive mechanisms embedded in contemporary AI and the silent power that digital media companies exercise over us. We should never cease to interrogate how the technology works, even while we are trying to accommodate it in the fabric of everyday life. (Natale 132) Although even a comprehensive list is beyond the scope of this short article, is it nevertheless vital to note that in looking beyond the promotion of AI Panic and deceptive media, ChatGPT and other generative AI tools create or exacerbate a range of very real and significant ethical problems. The most obvious problem is the lack of transparency in terms of what data different generative AI tools were trained on. Generally, these tools are thought to get better by absorbing ever greater amounts of data, with most AI companies acknowledging that scraping the Web in some form has been part of the training data harvesting for their AI tools. Not knowing what data have been used makes it almost impossible to know which perspectives, presumptions, and biases are baked into these tools. While many forms of bias have plagued technology companies for many years (Noble), for generative AI tools, in “accepting large amounts of web text as ‘representative’ of ‘all’ of humanity we risk perpetuating dominant viewpoints, increasing power imbalances, and further reifying inequality” (Bender et al. 614). Even mitigating and working to correct biases in generative AI tools will be a huge challenge if these companies never share what was in their training data. As the WGA and SAG strike discussed above emphasises, the question of human labour is a central challenge for generative AI. Beyond Hollywood, more entrenched forms of labour exploitation haunt generative AI. Very low-paid workers have done much of the labour in classifying different forms of data in order to train AI systems; data workers are routinely not acknowledged at all, even sometimes directly performing the tasks that are ascribed to AI, to the extent that “distracted by the specter of nonexistent sentient machines, an army of precarized workers stands behind the supposed accomplishments of artificial intelligence systems today” (Williams, Miceli, and Gebru). It turns out that people are still doing the work so that companies can pretend the machines can think. In one final but very important example, there is a very direct ecological cost to training, maintaining, and running generative AI tools. In the context of global warming, concerns already existed about the enormous data centres at the heart of the big technology platforms prior to ChatGPT’s release. However, the data and processing power needed to run generative AI tools are even larger, leading to very real questions about how much electricity and water (for cooling) are used by even the most rudimentary ChatGPT queries (Lizarraga and Solon). While not just an AI question, balancing the environmental costs of data centres with the actual utility of AI tools is not one that is routinely asked, or answered, in the hype around generative AI. Escaping the hype and hypocrisy deployed by AI companies is vital for repositioning generative AI not as magical, not as a saviour, and not as a destroyer, but rather as a new technology that needs to be critically and ethically understood. In seminal work exploring how young people engage with digital tools and technologies, Mimi Ito and colleagues developed three genres of technology participation: hanging out, where engagement with any technologies is largely driven by friendships and social engagement; messing around, which includes a great deal of experimentation and play with technological tools; and geeking out, where some young people will find a particular focus on one platform, tool or technology that inspires them to focus enough to develop expertise in using and understanding that tool (Ito et al.). If young people, in particular, are going to be living in a world where generative AI tools are part of their social worlds and workplaces, then messing around with ChatGPT is, indeed, going to be important in testing out how these tools answer questions and synthesise information, what biases are evident in responses, and at what points answers are incorrect. For some young people, they may well move from messing around to completely geeking out with generative AI, a process that will be even more fruitful if these tools are not seen as impenetrable magic, but rather as commercial tools built by for-profit companies. While the idea of digital natives is an unhelpful myth (Bennett, Maton, and Kervin), if young people are going to be the first generation to have generative AI as part of their information, creative, and search landscapes, then safely messing around and geeking out with these tools will be more vital than ever. We mentioned above that most Australian state education departments initially banned ChatGPT, but a more optimistic sign arrived as we were finishing this article insomuch as the different Australian states agreed in mid-2023 to work together to create “a framework to guide the safe and effective use of artificial intelligence in the nation’s schools” (Clare). Although there is work to be done, moving away from a ban to a setting that should allow students to be part of testing, framing, and critiquing ChatGPT and generative AI is a clear step in repositioning these technologies as tools, not magical systems that could never be understood. Generative AI is not magic; it is not a saviour or destroyer; it is neither utopian nor dystopian; nor, unless we radically narrow the definition, is it intelligent. The companies and corporations driving AI development have a vested interest in promoting fantastical ideas about generative AI, as it drives their customers, investment, and future viability. When the hype is dominant, responses can be overdetermined, such as banning generative AI in schools. But in taking a less...