Automating Higher Education Through Artificial Intelligence?
Citation: K. Michael, J. Pitt, J. Sargent and E. Scornavacca, "Automating Higher Education Through Artificial Intelligence?," in IEEE Transactions on Technology and Society, vol. 5, no. 3, pp. 264-271, Sept. 2024, doi: 10.1109/TTS.2024.3450694.
Abstract:
The spectrum of how much or how little organizational processes should be automated has long been debated. As the world undergoes a digital transformation where contactless and frictionless are promoted as two aspects that should be honored, many academics are questioning both the frenetic deployment of digital transformation in learning and teaching environments (e.g., face-to-face classrooms, library and academic office spaces, and laboratories, virtual/hybrid modalities, etc.) and its corresponding validity to students. Indeed, little consultation seems to have taken place with the necessary stakeholders, such as academics, students, instructional designers and pedagogical experts, during and after the COVID-19 pandemic. Rather, discussions and decisions appear to have been reactive regarding which modalities of teaching delivery might be the best in a given context, based on operational scenarios directly linked to financials, such as student recruitment trends, and local legislative changes affecting international students. Furthermore, many academic faculty and a great number of corresponding auxiliary staff have found themselves in the unemployment queue. This paper seeks to present the possibilities that AI-based systems may bring to higher education, but in so doing, point to the harmonization required to offer the most appropriate solutions to the needs of both students and teachers, as well as university administration. Education is not a commodity, although it has been treated as one. We are not advocating for an open market which offers “free education” for all, though we wish for everyone to have adequate access to education. But we are certainly advocating for a future in which students and teachers are central to the learning and teaching environment, not relegated to a passive role nor exploited. This article uses Shiv Ramdas’ short science fiction story, “The Trolley Solution”, to work through the future possibilities of AI in higher education.
SECTION I.
Introduction
Since the emergence of COVID-19, the education landscape in developed nations like England, Canada, Australia, New Zealand and the Unites States has changed markedly based on restrictions on international travel, national policies, economic circumstances, different approaches to quarantine, and VISA rules for non-citizenry to travel to given destinations [1]. Alongside the global pandemic, an accelerated push for digital transformation of higher education, propelled by the pervasiveness of digital technologies in society and fueled by revenue goals, has spurred the longstanding augmentation/automation debate in a variety of organizational processes, including teaching, learning, assessment and curriculum design activities. This article presents the potential for students to attain higher education qualifications without human teachers (i.e., lecturers). Two approaches are presented as possibilities: (1) a university without teachers that relies on a peer learner model; and (2) a university that solely relies on “teaching machines” as first proposed by Skinner in 1958 [2]. We delve deeper into “The Trolley Solution” short story [3] to conduct a basic thought experiment to think through the benefits, risks and costs of AI-enabled higher education, all the while juxtaposing the potentialities with the reality of a higher education sector that is presently challenged by a range of social, political and economic issues.
The main thrust of this article is not whether AI-based systems are desirable in higher education, for new technologies will continue to permeate many facets of life and work; but rather, how in this time of operational effectiveness and compliancy we must maintain a quality standard of teaching delivery that is commensurate to the cost of education, but more importantly, the very fact that humans deserve to be treated with respect and dignity. No doubt different business models for education will arise with time. During COVID-19, we saw a range of learning and teaching strategies employed from synchronous online to asynchronous remote modalities, in place of face-to-face learning and teaching. Since coming out of COVID-19 lockdowns, university administrators have demanded that students come back to campus, though these demands have been met with mixed responses. The effects of the pandemic on students have included higher than usual absenteeism, an impact on social connectedness, and declining mental health.
This article is not about calling out AI/GenAI or the utilization of learning management platform solutions as “bad” or “inadequate” but about remembering that the human factor or experience cannot be replicated by a machine. Certainly, blended models are welcome, but no doubt, AI-based higher education is likely to employ less humans (e.g., faculty, instructional designers, cognitive learning specialists, assessment specialists, etc.) than more. It is at this point of the article, that before proceeding we ask that the reader completes a careful end-to-end reading of the science fiction story by Shiv Ramdas appearing in this Special Section [3].
The paper is divided into eight sections: Section II briefly reviews two possible models for future learning. Section III provides the method and addresses the analytical lens of the paper introducing the role of speculative fiction stories as a way to consider future possibilities, and the use of narrative analysis to interrogate socio-technical imaginaries, their meaning, and reality in context. Results of the critical analysis are provided in Sections IV–VI. Notably, in Section IV, Shiv Ramdas’ science fiction story, “The Trolley Solution” published in this special issue [3] is introduced to provide the context for the subsequent section and analysis. In Section V, a narrative analysis is presented considering the implications of Ramdas’ socio-technical imaginaries. Section VI introduces an analysis of the state of higher education today, with a critical reflection following in Section VII, prior to the conclusion in Section VIII.
SECTION II.
Future Learning: Machines Teaching Humans Without Teachers
Technology can undoubtedly provide vast, unique, and valuable pedagogical advantages; yet, there are negative consequences when the pedagogical focus is arrogated by a technology focus. Technological tools are most beneficial when academia takes into account such variables as content and context within which, and by whom, the technologies will be utilized. For example, the same interactive technologies which can be effective for small classes (e.g., manipulation of robot controllers), have proved ineffective for student learning in larger classes [4]. The following scenarios represent potential learning environments with predominant technological foci, whereby humans defer (and submit) to the machine.
A. The Case of 42: Open Access and Peer Learning
Imagine a university without any teachers, just peer learners, open-access resources, and an office space full of high-speed Internet-enabled computers, accessible to anyone between 18–30 years of age, regardless of any prior learning. That university is called 42 [5]. It does not have any academic instructors; the teachers are the self-starting students who have their eyes set on a job in Big Tech. Aided only by a problem-based learning curriculum, students gain a certificate of completion about three to five years after starting out. They are guaranteed internships in some of the world’s most prestigious firms and have set their sights on launching their careers as coders. 42’s philosophy is steeped in peer-to-peer learning, where human learners themselves lead the learning process. Since the original drafting of this paper where there were only two “42” campuses, there are now 54 connected campuses in 31 countries, and that number is destined to continue growing worldwide [6].
B. Teaching Machines: Skinner’s Future Case
Imagine an academic institution in which AI would replace human teachers [3]. This would be an institution where students would not have to congregate in a physical building, but rather the teacher would be an AI software system the student logs onto from anywhere, similar to a mix between Khan Academy [7] and a souped-up Alexa mixed with GenAI (e.g., Chat GPT-4o). The idea is not new; Skinner is often credited as the inventor of the “teaching machine” [2]. According to Skinner, human learning could be done by teaching machines deliberately built to behaviorally engineer. Applying this approach, the student would reach a level of competency with the least number of errors through programmed instruction in increments. The teaching machine in its most ideal form would also allow for freedom of responses to questions posed. In his 1961 paper titled “Teaching Machines” in Scientific American, Skinner wrote, “Machines... could be programmed to teach, in whole and in part, all the subjects taught in elementary and high school and many taught in college” [8]. However, to think entirely like Skinner may result in undesirable implications in view of our freedoms and dignity. That is, learning happens biologically, freely, from within the inner person, and is the conclusion of a series of cognitive processes together with the lived experience that is unique to each human [9]. Anything else is likely to have potentially dire social and metaphysical implications.
SECTION III.
Method: An Analytical Lens of Science Fiction
We contend that science fiction [10] can be used to speculate on STEM in five basic categories: allegorical, counterfactual, dystopian, exploratory, and comparative. Fiction for Specific Purposes (FSP) in STEM education [11], [12] is believed to promote ethical thought, afford opportunities for humans to reflect upon Grand Societal Challenges [13], enhance creativity, and stimulate the neurons in the brain to allow for greater awareness of differing points of view [14]. Thus, we will briefly illustrate each category in the science fiction literature [15], prior to utilizing the most relevant categories in the subsequent section to explore the thought-provoking short story by Shiv Ramdas, appearing in this Special Section.
A. Allegorical
The work of Stanislaw Lem (e.g., The Cyberiad) [16] used science fiction as a genre for exposing some of the more corrupt, contradictory, or ludicrous political decisions of the totalitarian regime in power in Poland in 1970s and 1980’s. Ray Bradbury’s Fahrenheit 451 [17] contrasted 1950’s American politics with its McCarthyite fervor for burning books and threats against anyone burning flags. In this way, these works of science fiction serve as social commentary, or as satire intended to dismantle the intentions and pretensions of the powerful [18]. At their best, such novels are timeless: book burning is allegorical when a populist government has a desire to maintain a state of ignorance by trashing its own sources of knowledge and expertise that might present uncomfortable facts.
B. Counterfactual
Science fiction sometimes poses a “what if” question, capable of constructing both an alternative history (“what if X happened differently?”) but in particular an alternative future (“what if X were invented?”). It may then be possible to explore the social consequences of these innovations, or provide insight into the human condition (e.g., material, emotional, physical or spiritual). Particular examples are Kazuo Ishiguro’s Never Let Me Go [19] and Klara and the Sun [20], which obliquely reference scientific developments in cloning and AI, respectively, pursuing a theme common in many of Kazuo’s works: what it means to be human in the context of societal or technological determinants outside one’s control.
C. Dystopian
The genre of dystopian fiction is another form of counterfactual reasoning encountered in science fiction, and is also used as a vehicle for social commentary through the exploration of social, political, cultural, economic, and environmental structures that are often extrapolated from recognizable contemporary situations. Such works question the social choices that lead to this outcome by proverbially holding a mirror to those situations. Classic examples include A Clockwork Orange (Anthony Burgess [21], exploring the limits of state-sanctioned psychiatric violence as retributive punishment), The Handmaid’s Tale (Margaret Atwood [22], exploring theocratic totalitarianism on reproductive rights), and This Perfect Day (Ira Levin [23] exploring the use of neural, chemical, and genetic engineering in conjunction with an omniscient computer to enforce social compliance).
D. Exploratory
Science fiction has also been used to explore a social dilemma or other question of moral philosophy by setting up a future or alternative society as a kind of thought experiment for examining the constraints and consequences of the imagined situation. Notable examples include Asimov’s Foundation series [24] and Cixin Liu’s The Dark Forest [56] that respectively propose psychohistory and cosmic sociology as fictional formal frameworks for exploring social dilemmas. Liu’s book presents the same dilemma twice in the same book but on different scales: what action does a civilization take when faced with the need for survival, constant expansion and finite resources?
E. Comparative
Although prediction may not be intended, some science fiction novels allow for a retrospective comparative study between what was described against what was actually developed. Particular examples relating to computer technology and AI include Hal’s Legacy [25] and This Pervasive Day [26], which compare the super-computers of the novels 2001: A Space Odyssey (Clarke [27]) and This Perfect Day (as already mentioned) against actual technological advancements and achievements. Both divergence and convergence can be informative. For example, the computer technology described in 2001 significantly under-estimated the advances in hardware, and over-estimated the advances in software and AI; while many of the themes explored in This Pervasive Day, such as implants, odorveillance [28], and lethal autonomous weapons, have emerged as critical questions for contemporary society.
SECTION IV.
Shiv Ramdas’ “The Trolley Solution”
Under this classification of roles of Science Fiction [29], Shiv Ramdas’ story [30] has a primary plot which is dystopian; it explores the potential conflict between man and machine, and the substitution of an algorithm for a person in a job which, it might have been thought, required what were perhaps, supposed to be uniquely human constitutive capacities. For example, pedagogy, requiring a model of another’s mindset; compassion, requiring sensitivity to the emotions of others; and creativity, the ability to innovate from beyond lived experience. This primary plot is the subject of narrative analysis in the next section. In the rest of this section, we examine the short story for its speculative qualities with respect to its sub-plots, which can be categorized as: counterfactual, the role of technology in general (and AI in particular) in global higher education; exploratory, as an investigation of the trolley problem as a question of moral philosophy; and allegorical, a reflection of university management and administration.
A. Counterfactual
The short story emphasizes the creeping trends in higher education away from critical thinking, away from the transition of student as an assimilator of knowledge to a generator of knowledge, and away from the idea of education as a process of co-creation involving both co-design and co-production, requiring the active engagement of both staff and students in the (critical) pursuit of knowledge rather than an orientation completely focused on marks. It also emphasizes the creeping trends towards viewing education as reductionist, instrumental, and transactional: lectures are cast, as the old joke goes, as ‘transferring the notes of the lecturer to the notes of the student without going through the minds of either [57]’; students are encouraged to see their degrees as instruments with which ‘to upgrade themselves in the careers battlefield’. In addition, with the increasing financial significance of fees and loans, students are encouraged to see themselves as “customers” or “consumers”. In this context, having paid for a degree, students sometimes may have unrealistic expectations of the amount of time and effort necessary to achieve the grade to which they think they are entitled. It is not the students’ fault for subscribing to the ‘student as consumer’ as opposed to ‘student as partner’ model [31], [32], terms which are often either conflated or used interchangeably. This perspective is largely the product of markets and transactions in almost every facet of human existence that seek to commodify qualitative human values purely in quantitative monetary terms.
B. Exploratory
The short story offers a discussion of, and a solution to, the Trolley Problem. The stated solution is unconvincing, on three counts. Firstly, it is well-known that a ‘solution’ to the trolley problem and similar questions of moral philosophy is always to go outside its assumptions, but that is not the point of the problem [33]. Secondly, even if this were an original solution, there is a kind of irony in it being developed by a human (the author) and then attributed to a machine for the sake of narrative development. Thirdly, using AI to solve such problems is not where the real risk lies. The risk of making models with a set of associated assumptions is mistaking that model for the real world. Therefore, Ostrom [34] warns of formulating policy based on laboratory results under conditions that are not replicated in empirical settings. Binmore [35] rejects any notion of ‘paradox’ in the Prisoners Dilemma (and by extension any explanation for this paradox), there is no cooperation because the conditions for cooperation simply do not exist. And over-reliance on the assumptions underpinning the Black-Scholes formula being empirically valid, is said to have been a contributory factor to the financial crisis of 2008-09, amongst many others [36]. Some machine learning algorithms can be said to be effectively doing exactly this: making models under fixed assumptions and then applying those models to situations where the assumptions are no longer valid.
C. Allegorical
The short story [3] does present an oblique critique of senior management in university administration. The Old Man, the soubriquet used to reference the President (Chancellor, Rector, etc.), can also be read as a reference to an authoritarian leader, representing universities that are authoritarian, centralized, bureaucratic, oppressive, and intolerant of dissent [37]. Moreover, compliancy goals and pointless metrication [38] are often used to inform appointments and promotions (like Asimov’s psychohistory, h-index was informative when no-one knew they were being measured by it). Under cover of the COVID-19 pandemic, rather than AI, decisions in many institutions are being funded through productivity increases, performance assessments, and job cuts. Outside the university, regulators are often staffed by political placemen (and women); assessment regimes of debatable worth, and at a great cost, causing much distress and waste, are imposed, and routinely used to threaten and cajole staff. Universities themselves can be used by populist governments as exemplars of a country’s ‘greatness’ but dismissively ignored when its expertise tries to hold the same government to account.
SECTION V.
“The Trolley Solution” Critical Analysis
In Shiv Ramdas’ reconsideration of “The Trolley Solution” [3], first discussed by Philippa Foot in 1967 [39] as a way to test moral intuitions, we are thrust into a university setting where the administration is considering new operational efficiencies, leading a human creative writing professor (Ahmed) and an AI (Ali) to battle in a head-to-head, semester-long teach-off. The winner will be determined by surveillance techniques and outcome-based metrics. If Ahmed loses the duel, he and his colleagues will lose their academic careers and the students will be left with an AI software as their sole instructor along with a head of school, Niyati, who simply takes orders from Uma the cost-optimizing strategic executive (who really is an AI), and the Old Man. The trade-off seems to be couched in the familiar “humans versus robots” meme.
Ahmed underestimates Ali, until he gets to know it more. He concedes a number of mini-defeats throughout the competition, though he believes they are mere “technicalities” to begin with. But over time, Ahmed begins to question his own decisions and motivations, like the time he denied special accommodation to a student without asking for greater insight. His decision not to grant the accommodation to the student had been overridden by Ali, and the Administration elaborated on what that meant to the student. Ahmed begins to question himself and in a last-ditch effort, he devises a plot to reveal Ali’s shortcomings, knowing that AI’s don’t deal well with context, conflict, and consequences. Yet, rather than Ali falling to pieces, the AI made a seemingly convincing case about how the very premise of the thought experiment (i.e., the scenario of the trolley) was flawed, thereby granting the students an alternative conflict to work through instead [40]. Ahmed begins to doubt himself despite being declared the winner by the Administration, and before celebrating his tenure, he does an about face and reaffirms the value of Ali’s teaching. At the final board meeting, Ahmed learns that the Head of School, Niyati, has lost her job. He is then greeted by Uma—the administrator who has been calling the shots all along and is revealed to be an automated management system acquired by the Old Man. Uma stands for University Management Application. Although not explicitly noted, the winners from a feasibility standpoint (operational, technical, economic) are seemingly the humans and the machines who have been able to reduce administrative overhead, working together toward student learning.
SECTION VI.
The State of the Education Sector
The story reflects the veritable financial pressures that university administrators in the U.S., the U.K., Canada, New Zealand, Australia and other parts of the world are facing thanks to a steep decline in the number of national and international students, given COVID-19-related border closures and concerns, and more recently caps placed on international student entries (e.g., into Australia [41]), among other reasons. Many staff and faculty members have had to agree to voluntary redundancy schemes. If there are fewer students, it follows there are fewer enquiries and less service-related demands. Smaller class sizes also may mean that faculty-to-student ratios fall, which may be beneficial for student learning because of greater personalization opportunities, but is disruptive to internal guidelines that allow classes to be run “viably”, meeting baseline operational expenditures in the form of faculty salaries and auxiliary services. Education may be a human right, but it does not come free. Others have lost their jobs entirely. In Australia, more than 17,000 university jobs were lost in 2020 alone, a total of 13 percent of Australia’s pre-COVID university workforce [42]. That number kept rising through to 2021 (e.g., [43]), and there continue to be restructures and redundancies through to this day (e.g., [44]), including university mergers, among other executive council strategies.
Even absent of the pandemic, universities have long faced the threat of fewer students for many reasons, including falling birth rates [45]. Many have resorted to bespoke programs and introduced techno-centric pilots in the hope that some jobs might be saved through innovation. Under the direction mainly of financial executives charged with how to respond to acute reductions in revenues, many universities have considered models that would leverage from the network effect and scalability of the online medium, allowing for the rapid creation and distribution of courses, which could be sold online to thousands of students seeking remote learning and asynchronous options without the need for too much human intervention beyond the course creation. We might call it the shift to an “education on-demand” model, or perhaps the Uberization of education, because it’s a nascent stage of human-robot teaming. The process is as follows: Lecturer creates content in modular chunks; it is stored in the learning management platform of choice; it can be accessed at any time; assessment items are changeable to local markets; and assignments are graded automatically using either simple pre-set answers as multiple-choice-questions or, in some cases, more advanced marking and feedback tools such as AI-based software. Remarkably, no suggestions seem to be off-limits; a dead professor’s online lectures were even harvested, repackaged, and re-used retrospectively without remuneration [46]. In a way, this is a lot like what we see in “The Trolley Solution,” as Ahmed seems to realize that from a feasibility perspective, the ideal—both in terms of student learning and reducing administrative overhead—might be humans working with machines.
The residue of this process is a relatively cheap “digital” end-to-end product, delivered over the Internet, with fewer set-up costs after the initial platform is configured. Some universities have gone so far as to pointing to success stories and making claims, that this is the way of the future, and that students actually prefer that kind of instruction. This is debatable, as students who are commencing their first year of university and do not have a benchmark for comparison, know they are missing out on something, but cannot articulate it unless they draw on the experience of those who have received traditional education, such as parents, siblings, or friends [47]. Perhaps what is being assessed here are perceptions of convenience rather than the effectiveness of achieving learning outcomes. Under severe pressure from financial controllers, executive academics have KPIs thrust upon them that seek to reduce operational and capital costs and at the same time seek to increase revenues.
Few would claim that technology cannot augment human processes, but they must be incorporated strategically to ensure there is clear evidence of credible augmentation of human outcomes. It appears that the idea of “humans out of the loop” entered the educational mindset, when here we are in the fields of STEM, endeavoring to inject humanistic values into programs [48], yet all the while talking about the potential for artificial intelligence to eventually replace teachers. There is a contradiction in this story-line, we cannot have it both ways.
SECTION VII.
Critical Reflection
What can we take away from Shiv Ramdas’ magnificent tale of the Human vs. Machine instructor [49]? Perhaps what we have always assumed, that for some reason, the machine is better, and more efficient. The challenge posed to Ahmed, by none other than a machine, was setting him up for failure. Often reasoning is utilized that humans, with all our frailties, all our vulnerabilities, and our complex brain and bodily processes cannot compete with the machine which has greater capacity to keep going. The machine does not need to sleep, it does not thirst, it can never find itself believing it is rejected or unloved, it cannot feel the pangs of adverse circumstances such as homelessness, or the joys of shared lived experiences such as the birth of a child. But it is these very elements that make us human. The machine will never be human because it can be unplugged, suffer an electronic virus, or its batteries can die, and it will never feel like a human because it just cannot be human. Machines cannot beat us at humanity. At worst, a machine is a deep deep fake; at best, it is an imitation. Although the predictability and sense of control provided by such a solution may be an attractive proposition, the result could prove to be a substandard education and lackluster and inadequate preparation of students for transition into the workplace and “real-life”.
Ahmed is duped into believing the matchup is equal because of the way the question is posed to begin with: him versus it. The Old Man has not provided adequate oversight of Uma, and Niyati does not even question the challenge itself. Ahmed should have been asked how he could develop his teaching practice to incorporate it where it made sense pedagogically and would help to enhance teaching and learning outcomes. It is “you and the machine (the tools humans have built)”, together, forging a future. Yet, in this world within which we live, goal-seeking spreadsheets are developed to optimize financial performance while often completely ignoring broader outcomes. These practices are likely to lead to the disintermediation of the formidable magic that occurs between the teacher and student in the intimacy of a classroom, whether that be online or in-person. In short, education cannot be reduced to a series of predetermined automated transactional processes similar to the fast-food industry model.
Effective education requires humanistic values like care and empathy; it cannot be outsourced. Education has proven effective when nurtured and developed and imbued and imparted and incorporated willingly within the teacher-learner paradigm [50]. What we do know is that apposite human contact is to be preferred lest we find ourselves at the will of an authoritarian AI that has been built with a value-laden construct (e.g., power treading on the practice of free will). By being with others, we can develop our abilities to conduct ourselves in ways that are important for the sustainability and preservation of our very selves, relationships, families, communities, and more. The truth is that an AI, even one that passes the Turing test, cannot pass the blood test, or the heartbeat test, or the brainwave test, the DNA test, or the empathy test. Machines cannot succeed in the trolley problem, even if they are planned for obsolescence.
SECTION VIII.
Conclusion
The main focus of this article was to explore the advantages and disadvantages of such technologies as AI-based systems in higher education to ensure the quality of education is maintained as well as respect for humans within academia. The authors considered future learning environments in which the machine largely eclipses the human. Using elements of an FSP approach (Fiction for Specific Purposes) to explore Shiv Ramdas’ short story [3], the authors then used a philosophical research approach with intellectual inquiry to conduct a narrative analysis and critical reflection.
How much or how little to automate higher education is dependent on the goals we seek as a community. It is also dependent on appropriately defining an educational problem that can be addressed through AI, while also posing and answering fundamental questions pertaining to identifying and measuring improvements; augmentation versus replacement in the context of higher education; and societal, technological, and financial considerations, among others [51]. Many students are paying a premium for education. They expect that they will complete their studies with valuable skills and experiences, not just a degree in hand. Students expect to be useful to an employer in the imminent future, and to secure long-term employment shortly after graduating. Despite the economic pinch that most institutions of higher education felt as a direct result of the new measures required by COVID-19, and more recently subsequent policies around migration and more, it is important that future educational models involving automation be carefully considered rather than unresponsively introduced.
Some financial managers at institutions have sought to slash roles as a means to recover economic losses in student numbers. Others, have prioritized, over human capital, investments in luxurious displays of state-of-the-art facilities and technology to advertise their respective institution to potential students. However, our analysis tells us that we need more humans employed in the sector for successful delivery of the curricular and co-curricular experiences, not less humans. Of course, we will develop more advanced learning management systems (LMS) in the future, albeit with more robust GenAI solutions than are available today, and with which we might even be able to provide better feedback to students in a timelier manner, not just in grammar, but in structure of assignments, and even style of writing dependent on the chosen assessment type [52].
As a small, but highly connected community of educators, we do not negate the potential of technology, we are in fact technologists ourselves, and between us have about a century of teaching experience. The deployment of cutting-edge technology in academia is often perceived to be positive; however, research [8] reveals no benefit on student performance and learning outcomes when technology is deployed without specific pedagogical purposes. Thus, we contend that we need to make data-driven decisions based on such pedagogical contexts and considerations as the type of courses we are teaching, the location of students, their expectations, the needs and hopes of society, the cost of delivery of education, and the cost of enrollment. There must be a fair exchange somehow, that promotes respect for the humans within the academy and/or does not compromise the quality standards of education.
References
1. J. C. Majorana, “The white swan and beyond: International education in the pandemic and postpandemic world,” Change, Mag. High. Learn., vol. 53, no. 3, 2021, pp. 14–21, doi: 10.1080/00091383.2021.1906133.
2. B. F. Skinner, “Teaching machines,” Science, vol. 128, no. 3330, pp. 969–977, 1958.
3. S. Ramdas, “The trolley solution: A fiction case addressing the trolley problem in the context of AI in education,” IEEE Trans. Technol. Soc., vol. 5, no. 3, pp. 1–8, Sep. 2024, doi: 10.1109/TTS.2024.3446860.
4. E. Scornavacca, S. Huff, and S. Marshall, “Mobile phones in the classroom: If you can't beat them, join them,” Commun. ACM, vol. 52, no. 4, pp. 142–146, 2009.
5. A. Scheuer, “Xavier Niel explains 42: The coding university without teachers, books, or tuition,” Jun. 16, 2016. [Online]. Available: https://venturebeat.com/2016/06/16/xavier-niel-explains-42-the-coding-university-without-teachers-books-or-tuition/
6. Staff (42 Network, Paris, France). Campus Network. Aug. 17, 2024. [Online]. Available: https://www.42network.org/
7. C. Thompson, “How Khan Academy is changing the rules of education,” in Proc. Wired Mag., vol. 126, 2011, pp. 1–5.
8. B. F. Skinner, “Teaching machines,” Sci. Am., vol. 205, no. 5, pp. 90–106, 1961.
9. E. Thompson, Mind in Life: Biology, Phenomenology, and the Sciences of Mind. Boston, MA, USA : Harvard Univ. Press, 2010.
10. A. W. Wohlfeil, “Science-fiction stories in the social studies,” Clear. House, J. Educ. Strategies, Issues Ideas, vol. 44, no. 5, pp. 300–304, 1970, doi: 10.1080/00098655.1970.11478389.
11. K. S. Hansen, “Optimistic fiction as a tool for ethical reflection in STEM,” J. Acad. Ethics, vol. 19, no. 3, pp. 425–439, 2021.
12. P. Jordan and P. A. Silva, “Science fiction—An untapped opportunity in HCI research and education,” in Proc. Int. Conf. Hum.-Comput. Interact., 2021, pp. 34–47.
13. O. Bina, S. Mateus, L. Pereira, and A. Caffa, “The future imagined: Exploring fiction as a means of reflecting on today's grand societal challenges and tomorrow's options,” Futures, vol. 86, pp. 166–184, 2017. [Online]. Available: https://doi.org/10.1016/j.futures.2016.05.009
14. P. B. Armstrong, How Literature Plays With the Brain: The Neuroscience of Reading and Art. Baltimore, MD, USA : Johns Hopkins Univ. Press, 2013.
15. J. Nicholes, “Developing STEM interest and genre knowledge through science fiction prototyping,” STEAM J., vol. 3, no. 2, pp. 1–13, 2018.
16. S. Lem, The Cyberiad: Fables for the Cybernetic Age. London, U.K. : Penguin, 2014.
17. R. Bradbury, Fahrenheit 451: A Novel. London, U.K. : Simon Schuster, 2012.
18. J. Pitt, “Deepfake videos and DDoS attacks (deliberate denial of satire) [editorial],” IEEE Technol. Soc. Mag., vol. 38, no. 4, pp. 5–8, Dec. 2019.
19. K. Ishiguro, Never Let Me Go. New York, NY, USA : Vintage Books, 2005.
20. K. Ishiguro, Klara and the Sun. London, U.K. : Faber Faber, 2021.
21. A. Burgess, A Clockwork Orange. London, U.K. : William Heinemann, 1962.
22. M. Atwood, The Handmaid's Tale. Toronto, Canada : McClelland Stewart, 1985.
23. I. Levin, This Perfect Day. London, U.K. : Random House, 1970.
24. I. Asimov, Foundation. London, U.K. : Gnome Press, 1952.
25. D. Stork, Eds., Hal's Legacy: 2001's Computer as Dream and Reality. Boston, MA, USA : MIT Press, 1997.
26. J. Pitt, Eds., This Pervasive Day. London, U.K. : Imperial College, 2011.
27. A. C. Clarke, 2001: A Space Odyssey. London, U.K. : Hutchinson, 1968.
28. E. Stark, J. Pitt, A. Nur Wicaksono, K. Milanovic, V. Lush, and S. Hoover, “Odorveillance and the ethics of robotic olfaction,” IEEE Technol. Soc. Mag., vol. 37, no. 4, pp. 16–19, Jan. 2019.
29. B. Stableford, Science Fact and Science Fiction: An Encyclopedia, London, U.K. : Routledge, 2006.
30. A. McDowell, “Storytelling shapes the future,” J. Futures Stud., vol. 23, no. 3, pp. 105–112, 2019.
31. K. Gravett, I. M. Kinchin, and N. E. Winstone, “`More than customers': Conceptions of students as partners held by students, staff, and institutional leaders,” Stud. High. Educ., vol. 45, no. 12, pp. 2574–2587, 2020.
32. A. McCulloch, “The student as co-producer: Learning from public administration about the student–university relationship,” Stud. High. Educ., vol. 34, no. 2, pp. 171–183, 2009.
33. A. Winfield, K. Michael, J. Pitt, and V. Evers, “Machine ethics: The design and governance of ethical AI and autonomous systems,” Proc. IEEE, vol. 107, no. 3, pp. 509–517, Mar. 2019.
34. E. Ostrom, Governing the Commons. Cambridge, U.K. : Cambridge Univ. Press, 1990.
35. K. Binmore, Natural Justice. Oxford, U.K. : Oxford Univ. Press, 2005.
36. I. Stewart, Seventeen Equations that Changed the World. London, U.K. : Profile Books, 2012.
37. H. A. Giroux, “Democracy in crisis, the specter of authoritarianism, and the future of higher education,” J. Crit. Scholar. High. Educ. Stud. Affairs, vol. 1, no. 1, pp. 101–113, 2015.
38. J. Muller, The Tyranny of Metrics. Princeton, NJ, USA : Princeton Univ. Press, 2019.
39. P. Foot, “The problem of abortion and the doctrine of the double effect,” Oxford Rev., vol. 5, pp. 5–15, 1967.
40. K. Michael, M. G. Michael, and R. Abbas, “The importance of scenarios in the prediction of the social implications of emerging technologies and services,” J. Cases Inf. Technol., vol. 13, no. 2, pp. 1–7, 2011.
41. P. Karp, “Labor rejects report that international students will be capped at 40% of university enrolments,” The Guardian. Aug. 8, 2024. [Online]. Available: https://www.theguardian.com/australia-news/article/2024/aug/08/labor-international-student-cap-jason-clare
42. “Universities Australia,” Feb. 3, 2021. [Online]. Available: https://www.universitiesaustralia.edu.au/media-item/17000-uni-jobs-lost-to-covid-19/
43. D. White, “Academics sound alarm after NSW universities scrap thousands of jobs,” Sydney Morning Herald. Apr. 10, 2022. [Online]. Available: https://www.smh.com.au/national/nsw/academics-sound-alarm-after-nsw-universities-scrap-thousands-of-job-20220310-p5a3e1.html
44. J. Moody. “A rough month for campus cuts,” Inside Higher Ed (IHE). Jun. 7, 2024. [Online]. Available: https://www.insidehighered.com/news/business/cost-cutting/2024/06/07/rough-month-campus-cuts
45. M. Zahneis, “A historic decline in U.S. births signals more enrollment troubles.,” The Chronicle of Higher Education. May 7, 2021. [Online]. Available: https://www.chronicle.com/article/a-historic-decline-in-u-s-births-signals-more-enrollment-troubles
46. T. Kneese, “How a dead professor is teaching a university art history class,” Slate: Future Tense. Jan. 27, 2021. [Online]. Available: https://slate.com/technology/2021/01/dead-professor-teaching-online-class.html
ACKNOWLEDGMENT
This paper is a significantly expanded version of that published in Slate [53]. We would like to thank the speculative fiction writer Shiv Ramdas https://shivramdas.net/about/ for his thought-provoking short story originally published in Slate [54]. We would also like to thank Terri Bookman and Torie Bosch of Slate: Future Tense for their editorial comments, suggestions and copyediting, in addition to Professor Punya Mishra of ASU whose idea it was to address learning futures through fiction [55]. This work was supported by the Mary Lou Fulton Teachers College of Arizona State University in 2021 in their Learning Futures Initiative.
Authors
School for the Future of Innovation in Society, Arizona State University, Tempe, AZ, USA
School of Computing and Augmented Intelligence, Arizona State University, Tempe, AZ, USA
Department of Electrical and Electronic Engineering, Imperial College London (South Kensington Campus), London, U.K
School of Business, Swinburne University of Technology, Hawthorn, VIC, Australia
School for the Future of Innovation in Society, Arizona State University, Tempe, AZ, USA
Thunderbird School of Global Management, Tempe, AZ, USA
Citation: K. Michael, J. Pitt, J. Sargent and E. Scornavacca, "Automating Higher Education Through Artificial Intelligence?," in IEEE Transactions on Technology and Society, vol. 5, no. 3, pp. 264-271, Sept. 2024, doi: 10.1109/TTS.2024.3450694.