Scholarslop: The Dangers of an Algorithmic Idea of the University
David M. Berry
For centuries, academics have maintained a form of discursive advantage within universities. When university managers proposed reforms, restructures, or change, academics could mobilise disciplinary knowledge, theoretical frameworks, and rhetorical sophistication to contest these moves (Shore and Wright 2000). The administrator might possess institutional authority, but the academic wielded disciplinary understanding and the pragmatic experience of teaching students and undertaking research. This asymmetry created a certain space for resistance by scholars, however limited. The administrator could make demands, but often struggled to engage on the terrain of ideas themselves, giving way to the expertise of the academic faculty. Throughout the 20th century, academics ceded broader organisational control to an emerging managerial class, usually made up of former academics who had moved into administrative roles. Nonetheless, despite what Halsey (1982, 1995) termed a "decline of donnish dominion," academics have retained authority over academic matters, most notably curriculum design, teaching methods, and scholarly standards.
This balance may be shifting again. Large language models now provide administrative structures with something approaching a quasi-academic fluency. When an academic challenges a new policy on pedagogical grounds, citing educational theory or empirical research, the augmented administrator can increasingly generate counter-arguments that mimic scholarly discourse. Indeed, the LLM can produce responses that reference relevant literature, use the appropriate terminology, and construct arguments that seem to possess surface-level coherence. These AI systems can therefore cite studies (whether accurately or not), reference theoretical frameworks, and engage in what appears to be reasoned debate.[2] They can also create a quantity of discourse, which elsewhere has been called “AI Slop,” or “workslop” in corporate contexts (see Niederhoffer et al 2025), but in this case it is a quasi-academic slop (Berry 2025). This we might call "scholarslop" as AI-generated content that mimics academic discourse whilst lacking genuine scholarly insight, or more specifically "adminslop" when used in administrative communications, produced at volumes that exhaust the faculty’s capacity to respond.[3]
What matters here is not whether these AI-generated responses demonstrate genuine understanding or intellectual insight. Rather, what is important is their potential to undermine academic resistance through using LLM generation of discursive responses and counter-arguments. The scholar who could once rely on the command of disciplinary knowledge now faces responses that sound sufficiently credible to non-expert audiences, including crucially, senior managers, governing boards, and the lay oversight bodies who must resolve these disputes. The augmented bureaucracy does not need to think, it needs only to generate plausible-sounding arguments and proposals that can be used as institutional tools. It may be enough to "flood the zone" with adminslop (Montgomery 2025).[4]
 |
| Generated with Google Gemini |
The technical characteristics of LLMs exacerbate this threat. These systems operate by identifying statistical patterns in training data, generating outputs that are probabilistic averages of existing content. They tend towards standard responses that are adequate or competent but rarely excellent (Berry 2025). For example, when such systems generate curricula, module descriptions, or teaching materials, they tend to produce averaged content, materials stripped of the specificity and disciplinary expertise that characterise the best academic teaching – literally scholarslop. Applied systematically across institutions, this can create what we might term regression to mediocrity, resulting in a flattening of educational quality towards an algorithmically-determined average. The result is a bland, barely differentiated teaching environment that is cheap to produce, cheap to deliver and which students will be in a poor position to contest.
This might also represent a profound shift in university power (Ball 2012). It is likely that the aim of creating AI-generated administrative discourse is not truth-seeking, not understanding, and not education. It is control. It is the neutralisation of academic voice through the production of a quasi-academic text that can occupy, and fill, the same discursive space whilst serving managerial imperatives. Indeed, we might be on the cusp of what might be termed a managerial revolution in universities, one that deploys computational power not to enhance thinking but to suppress it, not to generate knowledge but to manufacture the appearance of intellectual legitimacy for administrative decisions (Deem et al 2007).[6]
Where previously the university administrator required academics to translate complex ideas into manageable metrics and reports, the LLM mediates this relationship differently (Burrows 2012). It generates the discourse of scholarship without the labour of thinking, creating what I call zombie knowledge through textual performances that simulate intellectual work whilst remaining empty of the judgement, interpretation, and critique that characterise real scholarship. This scholarslop seems to me to be a more serious threat to university scholarship and the internal coherence of the university, not to mention disciplinary boundaries, than historical attempts to impose managerial controls, such as through public management techniques.
Addressing this threat requires academics to abandon certain comfortable assumptions about institutional permanence and discursive authority. The university will not protect academic values simply because those values once animated it (Barnett 2000; Docherty 2018; McGettigan 2013; Newfield 2011; Wellmon 2016). The capacity to produce scholarly-sounding discourse might no longer suffice as a form of professional defence when bureaucracy can generate equivalent-seeming scholarslop at scale. In response, academics will need to find new means to defend the university and contest the use of scholarslop before it is too late.[7]
** Headline image generated using DALL-E 3 in October 2025. The prompt used was: "An oil painting inspired by Hieronymus Bosch’s The Garden of Earthly Delights, depicting three diverse scholars in a 19th-century Oxford college tutorial setting, wearing black academic gowns and seated around a wooden table in a richly detailed library. Around them swarm bizarre, surreal creatures made from computational and robotic elements — part machine, part organic. These beings resemble laptops, computers, webcams, code fragments, circuit boards, and robot-like figures, rendered in Bosch’s intricate and whimsical style. The walls and bookshelves are alive with strange mechanical flora and techno-creatures. Logos of AI companies such as OpenAI, Anthropic, and NVIDIA appear subtly emblazoned on some of these beings or their devices, blending into the fantasy environment. The lighting is warm and painterly, evoking a traditional oil technique, while the overall tone combines reverence for scholarship with a sense of the surreal, symbolizing the computerisation and AI transformation of the university." The second was using Google Gemini using the prompt "draw an oil painting. use the The Garden of Earthly Delights as an inspiration with the scholars surrounded by bizarre computational creatures but which are identifiable as laptops, computers, webcams, software and code, or robot style creatures. Use logos of AI companies, like OpenAI, Anthropic, Nvidia, and so on on things. have the three diverse academics sitting in a library. show Oxford college tutorial style of education (19th century in black academic gowns) and the coming computerisation of the University. use tech companies logos." Due to the probabilistic way in which these images are generated, future images generated using this prompt are unlikely to be the same as this version.
Notes
[1] The Quarrel of the Ancients and the Moderns (Querelle des Anciens et des Modernes) was a literary and intellectual dispute in late 17th century France, subsequently spreading across European learned culture, concerning whether contemporary (modern) literature, art, and thought could equal or surpass the achievements of classical antiquity. The analogy here is interesting because just as the Moderns challenged the Ancients' monopoly on cultural authority by asserting the validity of new methods and contemporary literature, so might an augmented bureaucracy attempt to challenge academics' monopoly on legitimate scholarly discourse. However, the parallel operates ironically in that the original Moderns sought to advance knowledge through reason and innovation, the contemporary "modern" bureaucracy deploys AI not to advance understanding but to simulate it, not to create new knowledge but to neutralise those who do. This is the notion of the inversion, whereby the genuine gets supplanted by the fake or the generative (see Berry 2025 for a complete description).
[2] Perhaps most worryingly, bureaucrats and administrators can now generate "instant" academic modules, lecture descriptions, and week-by-week teaching plans via LLMs. Whilst these AI-generated curricula may be limited, basic, and pedestrian by academic standards they nonetheless offer administrators a potential means of challenging academics' power over their own teaching. This represents a direct threat to the concepts of Lehrfreiheit and Lernfreiheit, terms denoting respectively the freedom to teach and the freedom to learn. Lehrfreiheit describes the academic's right to determine what and how to teach based on scholarly expertise, free from external interference which are principles enshrined in the Humboldtian university model (see Cole 2012; Jaspers 1965). Lernfreiheit refers to students' freedom to shape their own educational paths, to question, to explore ideas critically without predetermined outcomes. Both concepts relate to Bildung, the ideal of education as self-cultivation and the development of critical personhood. What is worrying is the potential for non-expert non-academic control over the curriculum, and thereby the reduction of academic autonomy, through the transformation of teaching into a scripted performance of administratively-determined content. This serves the interests of managerial control, not the good of students or education.
[3] The notion of scholarslop might also be usefully applied to the current deluge of student work generated via LLMs, or to popular commentators deploying LLMs to create the affectation or register of academese in their blogs and podcasts, generating the appearance of scholarly authority without its rigour. This is very evident in the "platform scholarslop" found on Medium, Substack, or LinkedIn, for example.
[4] I am grateful to Michael Dieter for making the link to Steve Bannon's notion of "flooding the zone" (Broadwater 2025).
[5] The "idea of a university" has been contested throughout its history, with different conceptions emphasising different purposes and values. Newman's The Idea of a University (1996) articulated the university as concerned with liberal education through the cultivation of intellect through exposure to universal knowledge, pursued for its own sake rather than immediate utility. Humboldt's vision emphasised the unity of teaching and research, viewing the university as a site where knowledge is actively produced through Wissenschaft rather than merely transmitted (see Cole 2012). Jaspers (1965) developed this into a conception of the university as dedicated to the "search for truth" through free enquiry and Oakeshott (2001) described the university as as a place of learning rather than instruction, where students encounter intimations of different modes of understanding through conversation with scholars. More recently, Collini (2012) has defended the university's role in providing extended reflection and criticism of society's beliefs and practices. What unites these varying conceptions is an emphasis on intellectual freedom, critical enquiry, and education as intellectual transformation rather than training (see Docherty 2018; McGettigan 2013; Newfield 2011; Wellmon 2016).
[6] We might also note that the excitement over AI and its perceived solution to productivity issues, whether in the university or in the larger economy, is an effect of the current AI boom. This growth in value in the AI industries represents a potential for a dangerous financial shock if, and more likely when, the AI bubble bursts. As Sharma (2025) notes, "AI companies have accounted for 80 per cent of the gains in US stocks so far in 2025."
[7] Resisting scholarslop will require new institutional policies and academic practices. Some suggestions might include, (1) governance reforms: requiring named authorship of management documents, making individuals accountable for the content they present as their own work and where AI is used it is clearly marked; (2) collective academic action: a refusal to engage with anonymous policy documents that exhibit characteristic LLM patterns of production, generic work, or suspiciously rapid production of bland academic content; (3) alliance-building with students: increasingly students should be taught to recognise AI-generated content in their own work and may be effective allies in identifying and contesting scholarslop in institutional documents; (4) alliance-building with administrative staff: working with professional services colleagues to ensure that there is a shared understanding about how these tools should be used and where they should not; (5) development of professional standards or codes of practice: within universities these should address the use of AI-generated content in university governance, policies and documents, establishing clear expectations about authorship in institutional discourse and how AI generated materials should be treated (e.g. clearly marked as such, read and signed off by a responsible human administrator on behalf of the university). This could perhaps be mandated through the Office for Students in the UK, or through professional associations or bodies.
Bibliography
Ball, S.J. (2012) ‘Performativity, Commodification and Commitment: An I-Spy Guide to the Neoliberal University’, British Journal of Educational Studies, 60(1), pp. 17–28. Available at: https://doi.org/10.1080/00071005.2011.650940.
Barnett, R. (2000) ‘University knowledge in an age of supercomplexity’, Higher Education, 40(4), pp. 409–422. Available at: https://doi.org/10.1023/A:1004159513741.
Becher, T. and Trowler, P.R. (2001) Academic tribes and territories, Philadelphia, PA: Open University Press.
Berry, D. M. (2011) The Computational Turn: Thinking About the Digital Humanities, Culture Machine 12 (January). https://culturemachine.net/the-digital-humanities-beyond-computing/.
Berry, D.M. (2025) ‘Synthetic media and computational capitalism: towards a critical theory of artificial intelligence’, AI & SOCIETY. Available at: https://doi.org/10.1007/s00146-025-02265-2.
Broadwater, L. (2025) ‘Trump’s “Flood the Zone” Strategy Leaves Opponents Gasping in Outrage’, The New York Times, 28 January. Available at: https://www.nytimes.com/2025/01/28/us/politics/trump-policy-blitz.html (Accessed: 6 October 2025).
Burrows, R. (2012) ‘Living with the H-Index? Metric Assemblages in the Contemporary Academy’, The Sociological Review, 60(2), pp. 355–372. Available at: https://doi.org/10.1111/j.1467-954X.2012.02077.x.
Cole, J.R. (2012) Great American University: Its Rise to Preeminence, Its Indispensable National Role, Why It Must Be Protected. Public Affairs.
Collini, S. (2012) What are Universities For?, Penguin.
Deem, R., Hillyard, S. and Reed, M. (2007) Knowledge, Higher Education, and the New Managerialism: The Changing Management of UK Universities, Oxford University Press.
Docherty, T. (2018) The new treason of the intellectuals: Can the University survive? 1st edition. Manchester University Press.
Halsey, A.H. (1982) ‘The Decline of Donnish Dominion?’, Oxford Review of Education, 8(3), pp. 215–229.
Halsey, A.H. (1995) Decline Of Donnish Dominion: The British Academic Professions in the Twentieth Century. Oxford: Oxford University Press, U.S.A.
Jaspers, K. (1965) The Idea of the University. Peter Owen.
McGettigan, A. (2013) The Great University Gamble: Money, Markets and the Future of Higher Education. Illustrated edition. London: Pluto Press.
Montgomery, B. (2025) ‘How Musk and Trump are flooding the zone’, The Guardian, 11 February. Available at: https://www.theguardian.com/technology/2025/feb/11/musk-trump-doge-ice-google-seo-spotify (Accessed: 6 October 2025).
Newfield, C. (2011) Unmaking the Public University: The Forty-Year Assault on the Middle Class. Reprint edition. Cambridge, Massachusetts London, England: Harvard University Press.
Newman, J.H. (1996) The Idea of a University, Yale University Press.
Niederhoffer, K., Kellerman, G.R, Lee, A., Liebscher, A., Rapuano, K., and Hancock, J. T. (2025) AI-Generated “Workslop” Is Destroying Productivity, Harvard Business Review. https://hbr.org/2025/09/ai-generated-workslop-is-destroying-productivity, https://hbr.org/2025/09/ai-generated-workslop-is-destroying-productivity.
Oakeshott, M. (2001) Voice of Liberal Learning, Liberty Fund Inc.
Readings, B. (1996) The University in Ruins, Harvard University Press.
Comments
Post a Comment