The Mirror University 4 (part 1): Does Generative AI Dream of Digital Sheep?

This special two-part post is the fourth in a series of blogs that reflect on the educational and organisational challenges facing universities as they navigate intersecting existential, epistemic crossroads. I have called the series ‘The Mirror University’ for several reasons. The Mirror Universe is an alternate reality from the cultural phenomena that is Star Trek, in which people who exist in the prime universe have ‘evil’ alternates in the mirror universe. These alternates maintain the structures, relationships, roles and even identities of the ‘prime’ characters but exhibit traits, morals and behaviours that are the anthesis of the higher moral ground taken by our heroes. The mirror university maintains the structures, organisation, practices, and mythologies of what we understand to be a higher education institution but behaves in ways that are the antithesis of the ways of being they aspire to.
 
The mirror university is a construct.
It doesn’t represent any single institution.
It is not an allegory for the overarching influence of ‘management’.
 
The aspirational contradictions create a sense of emotional and idealistic liminality for those whose personal, intellectual, or professional identity is deeply rooted in the altruistic conceptions of university as a site of transformational social good for the community, our students, and the academy itself.

 
TLDR version: This article explores the impact of generative AI on human creativity and higher education. It discusses the complex relationship between technology and creative practices, highlighting debates around AI’s capabilities and limitations. It critiques the often crisis-driven responses of higher education institutions to AI, drawing parallels to past technological disruptions. The post emphasises the importance of human experience, storytelling, and serendipitous connections in learning and creativity, arguing that these elements risk being lost in an over-reliance on AI. The article concludes by advocating for a more nuanced approach that values human creativity and agency in education.

 
The capabilities, resistances and potentials of AI have informed the aesthetics, development and attitudes of a variety of forms of human creativity, creating new modes of participation and transforming others (see Lee, 2023 and their work on AI and creative industries and Henriksen et al., (2022) and their philosophical inquiry into AI as a space for posthuman creativity). The ‘insidious’ influences of technology has prompted some writers to observe a nostalgic return to the practices and aspirations of the handmade and a collectivist activism aesthetic (Rauch, 2015). The state of making and human creativity that exists between digital technology and social media practices determining and defining how something is made and the rejection of its influences for a more analogue approach has been labelled by some writers as post-digital, a term which has taken on several wider meanings over the ensuing decade (Cramer, 2015; Taffel, 2015).

Cramer argues that a post-digital media world is messy; a state arising from making practices being subject to hacking, argumentative communications, strident defences of positions and sometimes incorrectly held binary and pejorative descriptions of practice. Generative AI has made these states even messier, imbuing the technology with many of the characteristics and independent actions that were previously attributed to the human creator and user. We assert that the LLM ‘learns’ and is being ‘trained’. We ask an image generating AI to ‘imagine’ an image. It is ‘creating’ something new. It is ‘replacing’ the work that humans don’t want or need to do. McNamara (2023) puts forward the notion of the co-creativism where the ‘machine’ transcends their instrumental role as a tool and becomes an active participant in the artistic process, enabling democratisation, pattern making and connectivity that was not possible before AI. This is articulated in the context of reflexivity and activism by da Veiga (2024) who argues that:

The contemporary plane of discourse within post-digital cultures may thus be characterized by the observation that singular media entities have dissolved into a techno-social ecosystem, inhabited by smart things that intend to engage and cooperate with human users. This socio-technological environment is translated into a hyper-nervous system of existences, which addresses, affects, and appropriates the human on a preconscious level

Without descending into the rabbit hole of the debates about the nature of machine or digital sentience, there is a complex, challenged and often fragmented and under-evidenced debate about what constitutes creativity, imagination and human capacity in GAI. Swedish philosopher Nick Bostrom (see Jackson, 2023) theorises that ‘it’s not doing them [large language models] justice to say they’re simply regurgitating text… (they) exhibit glimpses of creativity, insight and understanding that are quite impressive and may show the rudiments of reasoning”. Computational architect Makoto Watanabe (2017) answers the question posed by author Philip K Dick in his 1968 dystopian science-fiction novel ‘Do Androids Dream of Electric Sheep?’ asserting that ‘machines are better than people at solving complex problems with many intertwined conditions. In that realm, people are no match for machines. But people are the only ones who can create an image that does not yet exist. Machines do not have dreams.’ Roibu (2023) argues that in the realms of GAI enabled modernity, the machine (which has the ‘… semblance of sentience at the very least imitating intuition and emotion) cannot exist without the human and vice versa and it is in that dialogue between the machine assembling the past that future creativity and scientific discovery will emerge, noting that the ‘expectations  of  the  human  interlocutor  may  transform,  birthing  new  areas  of interest, questions, and new vistas of knowledge to explore’ whilst the machine may ‘learn to create not just for the eye but also for the heart, mind, and the myriad of individual experiences that colour our perception of beauty.’ In a chapter entitled ‘The Shepherds of Electric Sheep’ Andrew Klobucar (2024) attempts to find the middle ground in this quasi-religious hankering for unrequited faith in the transformative potential of GAI, positing:

‘As artists and writers already understand, AI cannot replace human creativity and storytelling as an epistemologically driven tool; yet it can, with skill, be controlled and used to enhance aesthetic aims to question, if not disrupt knowledge systems.’

In a deeply provocative and quite mind-bending article, Carl Öhman (2024) posits that AI models are gods, arguing that ‘They are products of vast volumes of data, literally traces of past human (speech) acts, synthesized into a single agency that is (falsely) experienced by users as extra-human.’ He goes onto to argue using critiques of religion as an analytical tool that AI binds our decisions and future entirely to the lessons of the past warning that ‘the danger of LLMs is thus the temptation to place one’s faith in the past over one’s faith in oneself and one’s peers’. Taking this analogy to the nth degree, the eulogizing of the future potential of AI (tempered by fear-allaying mediation that any superintelligent ‘sentient’ AI will cause its designers no harm) excludes the possibility of human ingenuity and endeavour to do anything other than create this super-human intelligence. And it will only be through that inevitable progression towards AI super-intelligence that the capability to (defend or wage) future wars, to defy age and disease and to live forever will be within our grasp (see Kurzweil’s predictions from 2024). Hall (2024) argues a similar though less deified positionality for generative AI where it offers the possibility to remember (or not forget) our past synthesising, repurposing and integrating vast forgotten lakes of human endeavour, experience and knowing, asserting that ‘In remembering, AI might serve a generative purpose, similar to that of the elder, able to bear witness to what-was and to help us to (de)liberate what might be’.

Crisis? What crisis?

In the higher education sectors crisis-induced (or perhaps crisis-created) responses to generative AI, there is an uncritical idolising of generative AI and its capability (or wished-for capacity) to disrupt and transform the nature of the whole sector. It is labelled an existential threat to the sector (Schroeder, 2024) and the harbinger of the end of assessment (Risquez & O’Hara, 2024), the end of humanities study (Snook, 2023) and the end of academic work and the world of graduate employment (Wong, 2024) amongst other, sometime breathless, predictions. Institutional responses have veered from zealotism and decidedly uncritical decision-making around its use for the emancipation of academic practice from policy and rigorous evidence, to the superficial intertwining of output, productivity and creativity (Watermeyer, Phipps, Lanclos & Knight 2023).

Caught blind in our vendor promulgated rush towards the arrival of the superhuman possibilities of AI, we have reignited the internecine identity wars that have plagued innovation, agility and adaptability in higher education and our practices for centuries. Generative AI is just the latest technological crisis to draw the fault lines in the same techno-sage versus techno-resistant battles that have disrupted our history. Whether it be the introduction of computer-mediated learning and calculators (Deutsch, 1951), the use of video in teaching (Deds, 1983), the Internet (Baldwin, 1998), Web 2.0 (Steinhauer, 2021), MOOCs (Craig, 2015 amongst many others), Pokémon Go (van Apeldorn, 2021) or Second Life (Flavin, 2012). You name the disruption, I will show you the volumes of debate, hand wringing and conscientious objection that ensued. It is true that these resistances and acceptances are not experienced, enabled or exposed equally (Bryant, Coombs & Pazio (2014) and they vary from context to context, and arguably from era to era. What is true is that higher education institutions react to these disruptions in increasingly predictable and frequently neoliberal ways.

When compromised and threatened by financial, regulatory or reputational actions, institutions can defy the slow-moving behemoth in them to enable the kind of change unimaginable for business as usual activities (Alhazzanni, 2020). Critical infrastructure areas, leadership teams and under threat functions turn to the vendors, to consultants and often to everyone but their own experts to seek competitive insight (Gupta, et al., 2005 and Gunter & Mills, 2017). This external focus enables cultures of fear, change management, adversity and innovation to galvanise staff and students to lift productivity, react, respond and ideate and create content and delivery strategies (Farquharson et., 2018, Andrew, 2024 and Watermeyer, et al., 2023). Noted management scholar and theorist Peter Drucker (writing with Harold Ernarson) observed in 1960 that ‘Though the university community is a major force of innovation in our society, it is curiously resistant-even hostile-to innovations attempted within the university. The universities send out skilled specialists who spearhead purposeful change. But though specialists in the strategy and techniques of innovation crowd the campus, rarely are their skills brought to bear on the university itself.’

This is a chord. This is another. This is a third. Now form a band

This very famous call to action comes a 1977 handmade and photocopied zine for punk fans. It demystifies the art of playing guitar and energises those wanting to make and create, enabling a sense of DIY and expression. The ethos this zine cover represented seeded and catalysed a whole culture of music, art, fashion and political ideology and emancipation that continues to this day. It made music accessible, possible and inventive, because the rules that bound its production were sublimated and the historical precedent that privileged wealth and status over accessibility (especially in the excesses of the seventies). Punk ignited creativity that was informed by the stories and sounds of the past but was something entirely new. It was created. It was dreamt. It was not the technology that made the music, it was the experience of learning how to make it, and then doing it. In garages, in pubs, in bedrooms, onto tape machines and sung in rooms with mates, with chords drawn on the neck of your guitar with sharpie or a handmade keyboard you built from a kit. It was experiencing the exhilaration of experimenting and hearing those three chords come together in harmony with other instruments and players, even momentarily.

Prompt to ChatGPT: I don’t know how to play any instruments and I cannot sing. How do I start a punk band?

ChatGPT: Starting a punk band is not limited to playing instruments or singing! Here are some steps you can take to start a punk band without playing instruments or singing:

  1. Find musicians: Look for people who can play instruments or sing and share your passion for punk music. You can try finding bandmates through local music forums, social media, or music events.
  2. Manage the band: Even if you don’t play an instrument, you can take on a managerial or organisational role within the band. This could involve booking gigs, managing social media accounts, or handling band finances.
  3. Write lyrics: If you enjoy writing, you can contribute to the band by writing lyrics for songs. Collaborate with your bandmates to create music that reflects your punk ethos and style.
  4. Design visuals: Punk bands often have distinctive visual aesthetics. You can contribute by designing band logos, album artwork, flyers for gigs, or merchandise.
  5. Be the frontperson: Consider taking on the role of a frontperson or hype person for the band. You can engage with the audience during performances, introduce songs, and bring energy to the stage.

Remember, the punk ethos is all about creativity, individuality, and a DIY attitude. Feel free to think outside the box and find unique ways to contribute to your punk band!

What gets lost in our crisis-informed reaction to GAI is the criticality of the connections between knowledge and learning and how they evolve, grow and develop, get passed down and applied to problems to ‘dream’ something entirely new and unique to each successive learner. The capability of GAI to deliver the ‘most likely answer’ to any prompt or question skews our actions and our understanding towards the superficial and expedient. We move from problem to solution in decidedly linear ways, the beginning leads to the end of the book in an entirely predictable but economic way. Generative AI can’t tell stories, and it can’t tell you its own story, or what it has learnt from the experiences it has had. It can aggregate the past and send it back to us in a series of plausible and likely outputs. It may even (in the truest tradition of a thousand monkeys in the room writing Shakespeare) randomly make noise that comes together with harmony and serendipitous resonance. It still requires a human interpretation imbue its randomness and conflations with meaning. GAI models cannot articulate how the experiences of human relationships, discovery, pattern making, and luck inspire ideation, innovation and inspiration in often unexpected ways. When we construct a literature review or generate feedback on a student’s assignment using GAI, we risk losing all the uniqueness that comes from our creativity and humanity. Any LLM can construct a competent review of the literature on creativity and education (although it may make some of it up to fill the gaps an all-knowing God should not have!).

Prompt to Claude: Can you show me the references you cited?

Certainly, I can provide the references I cited in the literature review. However, I should clarify that as an AI language model, I don’t have direct access to a database of academic sources. The references I provided are based on my training data, which includes information about common publications in this field. As such, while these references are likely to exist and be relevant to the topic, you should verify them independently for accuracy. Here are the references cited

The human capability to take risks and seize the opportunity to smash two completely different ingredients together allows us to discover a unique new taste. An LLM can write a plausible summary of the literature directly responding to the prompts given to it. It is that human capability to imagine, dream, invent and discover connections between ideas and create something new that makes research and teaching so transformative. Any person (or bot) can repeat back quotes from a study and put them in an order to make an argument or write disconnected phrases to offer a justification of assessment outcomes. True impact, novel interpretations and inspiration discovery comes from our ability to be creative and share and evidence the narratives we are telling. Living, sharing and making up stories locates learning in the personal, exposing it across and throughout the non-linear experiences of how we live, work, play and lifelong learn. Stories and the experiences they represent provides students and staff with a sense of agency over what they share and how they represent their experiences and their learning. Storytelling as a part of experiential and connected learning sustains communities, validates experiences, nurtures relationships and serves as a critical form of cultural and disciplinary continuation. The joys and fears of uncertainty and the transformative, visceral and disconcerting effects of liminality are what it means to be human. In the second part of this blog, I will propose a manifesto for the creative university in a post-digital age and make my own assertions as to what role humans play in the engine of higher education.

Prompt to ChatGPT: Write me two verses and chorus of a punk song about nihilism and hating sitting in my lectures at university in 1977

Robo Nasty (or ChatGPT):

Verse 1: Sitting in these lectures, feeling so numb
1977, but my brain’s going numb
Lecturer drones on, words don’t mean a thing
Nihilism’s creeping in, can’t hear myself sing

Verse 2:Books and papers piling high, drowning in despair
Rebellion stirring, I’m gasping for fresh air
System wants to box me in, but I won’t comply
Nihilism fuels my fire, ready to defy

Chorus: Nihilism burning bright, in this academic fight
University’s a cage, but I won’t be staged
Punk spirit in my veins, tearing down the chains
1977’s the year, when my rebellion’s clear

Leave a Reply

Your email address will not be published. Required fields are marked *