How Doomed Are We?

A Philosophical/Theological Consideration of AI/ChatGPT in Relation to Theological Libraries and Theological Education

Brady Alan Beard, Ph.D. (he/him), Reference and Instruction Librarian, Pitts Theology Library, Candler School of Theology
Allison Graham, Reference and Research Librarian, Princeton Theological Seminary
Emily Peterson (she/her), Director of Public Services, John Bulow Campbell Library, Columbia Theological Seminary
David Schmersal, Access and Instruction Librarian, Austin Presbyterian Theological Seminary

Abstract: How should theological librarians think about and respond to the advent of AI in theological education? What role(s) might theological librarians have in this brave new world? How is AI similar to, and different from, other information technology revolutions (writing, printing, the internet)? Should institutions of theological education consider pursuing a kind of new monasticism, leaving the decadent dystopia into which we seem to be heading for the purity of quill and parchment? While we may be unable to provide definite answers, this session offers a forum for raising, considering, and discussing these and related questions.

This panel explores the challenges raised by AI in theological education by considering four key questions, each addressed in turn by a member of the panel:

1. “How is AI similar to, and different from, other information technology revolutions (writing, printing, the internet)?” (David Schmersal)
2. “How should theological librarians think about and respond to the advent of AI in theological education?” (Allison Graham)
3. “What role(s) might theological librarians have in this brave new world?” (Brady Beard)
4. “Should institutions of theological education consider pursuing a kind of new monasticism, leaving the decadent dystopia into which we seem to be heading for the purity of quill and parchment? Why or why not?” (Emily)

What Are We Talking About?

We start by defining a few key terms which are central to the discussion: artificial intelligence and GPT.

Artificial Intelligence (AI)

The Encyclopedia Britannica defines artificial intelligence (AI) as “the ability of a digital computer or computer-controlled robot to perform tasks commonly associated with intelligent beings” (Copeland, 2024). This seems clear enough, yet as Williams asserts, “The very term ‘Artificial Intelligence’ contains assumptions about the nature of intelligence (or consciousness), and who or what can have it—assumptions with no agreed-upon criteria, and a not uncontroversial history of application” (Williams 2023a, 106). As McGilchrist suggests, a more accurate term for AI may be “artificial information processing” (McGilchrist 2024). To illustrate the distinction using an analogy: A calculator can tell you that you owe $7,000 in taxes; an accountant can understand why this might be unsettling.

Moreover, as Hanegan and Rosser helpfully elucidate, AI is not monolithic, but may be classified as Analytical, which leverage incredible amounts of data for dynamic circumstances and real-time decision-making (e.g., recommendation engines like Amazon, auto-complete/auto-correct, Optical Character Recognition), and Generative, which can respond to crafted prompts and create content in various modalities (text, audio, video, images, code, etc.) that did not previously exist (Hanegan and Rosser 2023).

On perhaps somewhat more familiar ground, Herzfeld notes parallels between the three development paradigms of AI and three approaches to defining Imago Dei: substantial, functional, and relational. Just as the Imago Dei can be defined as a property humans have (reason), as a task we perform (exercise of dominion), or on the basis of relationality (grounded in the Trinity), so also the three developmental paradigms of AI may be defined as Substantive (classic or symbolic AI), seeking to replicate human reason; Functional, seeking to augment humans through actions such as accomplishing tasks and solving problems; and Relational, seeking to pass the Turing Test, to be indistinguishable from a human being in “conversation” (Herzfeld 2002, 303–310).

GPT

The GPT in ChatGPT stands for “generative pretrained transformer.” Its name describes a system of interoperable algorithms that weigh, arrange, and create associative distributions of text. They’re built on large language models . . . developed over the past five years or so, with datasets millions, billions, and now even trillions of words in size. LLMs are trained through deep learning—multiple layers of machine learning operating on and refining one another” (Williams 2023b). As Williams goes on to observe, “It is much easier to see through the mystique of ChatGPT and other AI applications once you understand exactly what they are and what they do. The truth about such algorithms is that they’re literally just sets of instructions. You have a set of standardized operations within which particular weights and measures can be adjusted . . . The recipe for any food—just as for anything you have to make—is an algorithm, too. My favorite algorithm is pumpkin pie” (Williams, 2023b).

Why Are We Talking About This Here and Now?

Before turning to the four questions around which our discussion is organized, it is fitting to briefly consider why theological librarians are addressing these questions, and why we are doing so now.

Although few of us have technical expertise in AI technologies, it is appropriate for theological librarians to address this topic because the social, technological and economic developments that have made AI possible have theological roots. As Lynn White observes, “technology is at least partly to be explained as an occidental, voluntarist realization of the Christian dogma of man’s transcendence over nature . . . Since the roots of our trouble are so largely religious, the remedy must also be essentially religious” (White 1967).

And why are we talking about this now? Because the potential ramifications of AI for our libraries, our institutions, and our faith communities are tremendous. As Sigler asserts, the scale of potential social changes instigated by AI will outstrip theologies based on current social conditions, such as liberation theologies (Singler 2017, 221). To paraphrase Space Balls, this is happening now.

Question 1. How is AI similar to, and different from, other information technology revolutions (writing, printing, the internet)?

David Schmersal

While many of the technologies that underlie the latest iterations of AI seem relatively new, the questions they raise echo debates that have been going on for quite some time. At the heart of such debates is the question of whether a “new” technology that seems to promise an external aid to memory or a shortcut to knowledge ultimately proves more beneficial or harmful. As Gay rightly observes: “prudence requires us to at least ask whether our technologies are now enabling us to become more of ourselves; whether they are facilitating deeper and more meaningful human relationships; whether they are enabling us to be more dynamically engaged with the world. One suspects that they are not” (Gay 2018, 25).

While it may be somewhat ironic for librarians to say so, one of the earliest forms of technology at the heart of these debates was writing. For librarians, especially theological librarians serving institutions affiliated with “peoples of the book,” the sanctity of the written word is a fundamental value and unquestioned assumption. Yet, as we will see, this has not always been the case, even for figures and books that are pillars of the Western “canon.”

In his critique of writing in Plato’s Phaedrus, Socrates relates an Egyptian myth in which the god “Theuth (Thoth) claims that writing is a study (μάθημα) that will make the Egyptians wiser and better in memory. Thamus (Ammon) argues that on the contrary, writing will produce forgetfulness in the souls of those who rely on it . . . because those who trust in writing will be relying on (something outside of themselves)” (Werner 2012, 285–286).

Eusebius (HE 3.39.4) offers a similar example from church history, noting Papias’s estimation of the oral testimony of the apostles over the written record. Papias (c. 60–130 CE) perhaps offers a more nuanced assessment of writing as an information technology; he is, after all, known to have written works himself. It is not so much that he considers oral testimony good and written words bad but recognizes the advantage of hearing “the living and surviving voice” (Eusebius 1926, 293). How many New Testament scholars wish they could sit down with Paul and ask what exactly he meant by some of the things he wrote that we find puzzling?

Moving forward in time over a millennium after Papias, the next major development in information technology, the printing press, further illustrates that all such technologies have both benefits and costs, such as the widespread distribution of mass-produced erroneous texts. (For example, consider Francesco Dal Pozzo’s inaccurate editio princeps of Ovid.1) This also offers a clear example of the fact that technology is not neutral, that “it can be directed toward good or evil,” and that all such technologies have unintended consequences (Schuurman 2019, 170). As much as the advent of printing is widely, and rightly, celebrated as a catalyst in the spread of knowledge, one wonders if Luther, presented with foreknowledge of the violence of the peasants’ revolt, the carnage of religious wars, and the consequent spread of skepticism, might have preferred that his ideas for reform were not publicized quite so widely.

The democratization and spread of knowledge that the printing press enabled has been exponentially enhanced—or exacerbated—by the great development in information technology of our own time: networked computers. Again, while the ubiquity of information may be seen as largely positive, it does have profound, unintended, and largely negative consequences, such as creating dependency and consequent loss of skill. (How many of us rely on GPS to find our way anywhere?) One clear example of this in our own context is Bible software. As much as I love and regularly use Accordance and appreciate how easy it is to quickly look up a word or check parsing, it also makes it all too easy to rely on the software without really wrestling with the text—thereby circumventing the struggle that can lead to insight and understanding.

While in some respects AI may be seen as merely the latest iteration of technologies that seem to replicate, or threaten to replace, aspects of the human mind, there does seems to be a fundamental difference. Whereas books may allow us to store our memory externally (so we need not hold all we know inside ourselves), and computers and the internet facilitate access to vast stores of information, AI seems to circumvent the process of thinking itself by processing the information for us and spitting out something useful.

Moreover, as suggested by McGilchrist in his insightful tome The Master and His Emissary, the information processing carried on by AI replicates a left-hemisphere perspective of the world, which seeks to reduce the world to that which can be grasped, mastered, comprehended, and controlled (McGilchrist 2019). This way of perceiving and of being in the world facilitated both the scientific method and the industrial revolution. Just as the industrial revolution reduced man to a machine and the earth to natural resources, so the information revolution that has brought AI equates personhood with intelligence and reduces intelligence to computation. If, as McGilchrist rightly observes, we and our culture have privileged this perspective and this has molded our worldview, then we have not so much created AI in our image as we have created AI based on a partial and distorted vision of ourselves and of reality. This of course has profound implications for theology in particular. AI represents not so much (or at least not merely) a tool that we may adopt and use as it does an expression of a worldview based on incomplete—and therefore erroneous and potentially hazardous—understanding of human beings, of the reality of the world we inhabit, and of being itself.

Question 2. How should theological librarians think about and respond to the advent of AI in theological education?

Allison Graham

This question is undoubtedly one that we all feel the need to address, whether out of our own interest, perplexity, enthusiasm, or worry—or because administrators or faculty at our institutions have told us that we need to have a plan for AI.

First, although it may seem obvious, thinking about AI and its role in theological education is important to avoid more extreme knee-jerk reactions. On the one hand, there is the call to “keep up with the times,” that libraries and theological institutions need to embrace AI or risk falling behind—although it is unclear what the goal of this great rush is. If I sound dismissive of this approach, it is because my tendency is toward the other extreme, of refusing to engage with AI because it seems to pose so many risks and downsides, and surely it would be better if we could just ignore it. But this refusal to engage does a disservice to our students and faculty who are using or encountering AI regularly and need to find ways to approach these tools in light of the goals and values of theological education. Additionally, not exploring AI risks overlooking genuinely useful AI tools. What I think we need instead is a thoughtful middle way that considers how AI tools can or cannot be used in the service of our theological narratives and institutional values.

In preparing for this presentation, I kept thinking of a quote from Wendell Berry about his approach to technology, “Do I wish to keep up with the times? No. My wish simply is to live my life as fully as I can” (Berry 1990, 90). Given that Berry eschews much of modern technology, including computers and modern farming equipment, I have no doubt that he is very much on the side of refusing to use AI. Nevertheless, I think that his prioritization of living a full life—or, on a communal level, human flourishing—is the right approach to take and can serve as a basic measure for evaluating the uses of technology, even for those of us who use AI and numerous other technologies far more than Berry does.

As my mention of human flourishing and a “middle way” alludes to, I take a virtue ethics approach to AI issues. In my context, working at a Christian seminary, this means that the focus is on living out the Christian story and any engagement with AI should be in the service of this larger purpose. For theological librarians and other theological educators, we also need to consider our role in student formation: How can students develop AI habits that they will maintain and share with the communities that they live in and serve, both during their studies and after graduation?

To help develop these habits, there are several methods for thinking about AI (and technology more broadly) that I find useful. I will present four of these frameworks briefly and then conclude by considering how these approaches could be applied in the context of theological libraries.

I begin with the most tech-focused of these frameworks, which was written by the twentieth-century Canadian philosopher and media theorist Marshall McLuhan. His tetrad of media effects proposes four questions to consider about media or technology:

1. What does it enhance?
2. What does it make obsolete?
3. What does it retrieve?
4. What characteristics does it reverse, when pushed to its limits? (McLuhan 1977)

Although these questions seem simple, they require us to think critically about positive and negative effects of technologies. I find the question of what AI retrieves particularly intriguing, though I have not found a satisfying answer to it yet. These questions can help us think about the effects of technology, and librarians could use them to help students and faculty understand what AI is, and how it is similar and different to other technologies. However, these questions are not aimed at addressing how we should respond to those effects and use technology.

Before turning to the question of response, I want to consider the work of Noreen Herzfeld, a contemporary scholar of religion and technology. In The Artifice of Intelligence: Divine and Human Relationship in a Robotic Age, Herzfeld draws on the work of Martin Buber and Karl Barth to discuss the importance of relationality and what makes an authentic I-Thou encounter, which entails looking the other in the eye, speaking and hearing each other, giving mutual assistance, and doing so freely (Herzfeld 2023a). This approach foregrounds values of relational personhood that most of our institutions likely share, and it applies those values to assess uses of AI.

AI and other digital technologies can easily weaken human relationality, so we must make sure that these values remain central and are prioritized over efficiency, cost, novelty, or other factors that are used to sell AI. Librarians should make sure that theological libraries remain spaces for human engagement and connection. I know that students with whom I have a relationship (through library outreach or other involvement on campus) are more likely to come and ask me reference questions. Seeking out opportunities to build those relationships will be increasingly important as students find it easy to use AI research tools on their own. This can be beneficial, but it can also mean that they might not be using these tools effectively or might assume that these tools are more comprehensive than they are.

Having looked at Herzfeld’s relationality-based framework, and McLuhan’s tech-focused one, I turn now to a couple of approaches that combine these methods by encouraging us to think about individual and societal uses and impacts of AI. Michael Sacasas is a Christian thinker and author of The Convivial Society newsletter about technology and society. In one of his newsletters, he poses 41 questions to ask about technology, including questions such as:

• What sort of person will the use of this technology make me?
• What practices will the use of this technology cultivate?
• What feelings does the use of this technology generate in me towards others?
• Could the resources used to acquire and use this technology be better deployed?
• How does the use of this technology shape my vision of a good life?
• What assumptions about the world does this technology tacitly encourage?
• What would the world be like if everyone used this technology exactly as I use it? (Sacasas 2021)

The questions encourage thinking about the impact of a particular technology on oneself, one’s relationship with others, and the world more broadly. (And this need not be exclusive to digital technologies.) Like McLuhan’s tetrad, they prompt consideration of positive and negative effects, but they are more focused on helping us think about how to respond to the impact that technology has.

Although I have not tried this yet, librarians could bring these questions to AI workshops that they lead and give students time to think through and discuss at least some of this list. Doing so could help students see AI issues as relevant to their education and ministries and lead them to develop personal philosophies about AI use. Thinking through these questions could also assist faculty and administrators to develop policies about AI. While the answer to some of these questions might be the same for all types of AI tools, other questions will likely elicit different answers for different products. Thinking through these questions for a variety of AI tools can help ensure that policies reflect the different types of AI and the range of products available.

While Sacasas’s questions address communal and global impacts of technology, the questions themselves are directed at individuals. A more societal perspective for thinking about AI can be found in Catholic Social Teaching. This is the approach that the participants in the AI Research Group at the Dicastery for Culture and Education’s Centre for Digital Culture propose in their new book Encountering Artificial Intelligence: Ethical and Anthropological Investigations. Just as modern Catholic Social Teaching developed in response to the Industrial Revolution, so it should be applied to the AI revolution. The principles of Catholic Social Teaching, which include upholding human dignity, valuing community, justice, solidarity, and care for God’s creation, can provide a strong framework for approaching AI. Librarians at Catholic institutions in particular could host book clubs about this book with faculty and students to encourage campus-wide discussions about AI and its uses.

As the authors of Encountering Artificial Intelligence write, “the most important aspect of an ethical engagement with AI is what the user does while off-line” (AI Research Group 2024, 229–30). People who are striving to live lives of love, service, prayer, contemplation, and learning will find areas in which AI tools can help them without compromising their values and identity. But people who do not work on these habits will find it easy to offload their efforts onto AI tools, thus making it even harder to develop these skills.

In the library context, we need to address AI, but in our excitement or panic about doing so, we should not lose focus on the ongoing importance of helping students develop information literacy, research, and writing skills, especially when the changing demographics of seminary students can mean that many students are admitted who have not developed some of these skills before.

So, how should theological librarians think about and respond to AI? As the work of McLuhan, Herzfeld, Sacasas, and the Vatican’s AI Research Group demonstrate, there are many useful frameworks for thinking through the use and impact of AI. My hope is that as I learn more about different AI tools, I can apply these frameworks to think through using AI in light of my theological commitments and my institution’s values. Doing so alongside faculty and students can help bring a range of perspectives and develop a cohesive understanding of AI and its effects in the context of a community.

Given how quickly AI is developing and how many different products there are, policies about AI are unlikely to address the full range of AI uses and tools. This does not mean that policies about AI should not be written, but they should be viewed as a general framework, not something that can account for every possible use of AI. Rather than focusing on creating a comprehensive policy, theological librarians can address AI by helping students develop research and writing habits that strengthen their skills, even while using some AI tools. There can be cases in which using generative AI is useful, but we should work to ensure that students are using AI intentionally, rather than feeling that they need to outsource their work to generative AI due to a lack of skill, resources, or support. Helping students form these habits and skills will assist them in their studies and as leaders in their communities, which will also have to grapple with the impact of AI.

Question 3. What role(s) might theological librarians have in this brave new world?

Brady Beard

In the brave new world of artificial intelligence, librarians may be tempted to take on many new roles, but they should focus their energy on playing the role they are trained for: librarian. In many of our institutions there is pressure to perform tasks in addition to our responsibilities as knowledge laborers; we often extol the work that librarians do as technology wizzes, pastoral counselors, assignment interpreter, and campus representative. To be sure, many of these roles are in fact important to librarianship, but without critical reflection, they can easily replace what many of us have trained to do: to collect, organize, preserve, navigate, synthesize, understand, and distribute information in many different forms. When it comes to AI and theological education, the temptation to go beyond our professional responsibilities is especially strong. This is due in part to our positionality with respect to information. We can easily see from the cataloging office or reference desk that no one else seems to be working on a topic or issue, or at least they aren’t working on it as quickly as we would like, so we feel that we must mitigate the negative and highlight the positive effects. As librarians, we can respond to artificial intelligence by building on our professional strengths through the four roles of guide, connector, supporter, and revealer.

Before we get to that, however, I’d like to discuss one role that I don’t think most of us are equipped to play. Librarians may feel pressure to engage AI as technical experts. This pressure may be formal or informal at any given institution. Often the library is on the front lines of institutional and educational technology. Additionally, AI tools are being integrated into search products and therefore disrupting information systems and information-seeking behavior. Naturally then, librarians may be expected to guide their institutions through the confounding literature and opinion pieces about AI, machine learning, and everything else. In addition to being on the tech frontline, librarians are often deeply involved in student academic support. Because many in higher education have hyper-fixated on GenAI as it relates to intellectual honesty and academic integrity, it makes sense that libraries would be asked to provide guidance on how/when/if students should be permitted to utilize GenAI.

The focus on technology and academic integrity can pressure librarians to function as their institution’s AI technician. Unless they have some serious skills in computer science, it’s likely not worth engaging on GenAI in a technical aspect beyond what they are already equipped to do. As information workers, librarians regularly deal with data management, computer-human interface, and complex front-end systems. Most librarians don’t have the technical back-end skills to dissect Copilot, ChatGPT, or Gemini. Because of the speed at which technology changes, understanding the technical aspects is likely not a good use of time. Instead, librarians should approach the topic from a place of our professional strengths to help usher our users into the brave new world.

Guide

Librarians often guide others in new systems, tools, and subject areas. In this regard, librarians can act from a point of strength when it comes to AI. AI research seems to change day to day and the speed at which new technologies develop only seems to be increasing. As overwhelming as such advances are to librarians, it must be even more so for faculty and students in theological institutions. Many librarians simply aren’t equipped to stay up to date on developments in AI research, but they do know how to find the most recent and relevant research. Guiding researchers and providing high-level evaluation, providing user support, and engaging critically will be time well spent. To borrow an example from the reference desk, I am often asked to help users navigate a tool like The Perseus Digital Library. My Greek is not as strong as it once was, and my Latin is virtually non-existent. As an expert in those specific areas of evaluation and research, I’m probably not the best resource for students or faculty, but as a guide who can help them navigate the big picture, I can provide a service that they likely will not be able to get from colleagues. I think the same is true for conversations around GenAI: I do not need to know exactly how a neural network functions, but I can understand enough of it to be able to navigate a tool and begin to evaluate the outputs based on what I might know of the general behavior of the product.

Connector

I tell students regularly that my favorite part of being a librarian is getting to do research with them and not have to provide a graded evaluation to them at the end of our time together. And that’s true. When it comes to the role that librarians can play in AI discussions, I think we can also be a point for connection and collaboration that doesn’t necessarily adhere to the strictures of a more formal academic space. To illustrate this point, last year at my institution I convened a small conversation group, “Speaking of the Divine in the Age of Artificial Intelligence: An ALC on Theology and AI,” with a $2,000 internal grant that I was awarded from the AIAI Network. The purpose of the grant was to gather scholars, librarians, students, and staff from Atlanta area theological schools to hold theologically informed discussions about AI (AIAI Network, n.d.). In total, ten or so folks gathered from two institutions. Nearly all the participants were staff who joined for a lunch-time conversation on Zoom, once every three weeks or so. The bulk of the funds went to sponsoring a public lecture on campus. The rest of the funds went to purchasing books for the participants.

When we started, I was surprised that there was as much interest from staff at my school as there was, and then I started listening to their interests. One participant was an assistant in the dean’s office who had a certificate in business analytics and uses AI to help with her daily workflow. Another was our director of career services who wanted to integrate AI into her services to students, and to better understand the space that students would be going into. Each participant brought a wealth of experience, knowledge, and questions that were new to me and were raised outside of library-land. It was a really refreshing way to connect with folks and find common cause in our day-to-day. The type of space that we were able to create was unique and fostered genuine engagement with the possibilities of AI while still addressing the harms that may result from AI.

Supporter

You don’t have to look far in higher education to find blogs, articles, op-eds, etc., about the crisis that AI is causing in academic integrity. A recent study of highschoolers, however, found that the use of AI has had little, if any, impact on how students cheat. In fact, some of the findings simply confirmed the fact that using AI to cheat well (i.e., to cheat without the possibility of getting caught) is difficult and there are lots of other easier ways to cheat (Lee et al. 2023). Instead, students were interested in finding ways to incorporate gen AI into their workflows. This brings us to the third role, that of supporter.

Because we often provide non-evaluative instruction, as librarians we play a special role in the education process of a student’s life. The focus on plagiarism and academic integrity gets in the way of that. GenAI is an opportunity for us to provide information literacy instruction, apply professional and personal ethics, and to think creatively with our students about how the subjects of their study engage the real world.

Revealer

Finally, we can be revealers. Part of our work as librarians is to organize and structure the information around us, but part of our task is to also be open about the very human (and sometimes not-so-arbitrary) reasons that our information landscape looks the way it does. Just as we might teach students how to navigate subject headings and to recognize the implicit biases and ethical dilemmas in our collections, we can also help reveal to students the structures that make up the world of AI. As people who think hard about data and its use, we can reveal the ethical and dubious ways that data of all kinds is gathered, structured, and deployed. With our professional expertise in information systems, we can rely on our strengths at critiquing black-box algorithms, systems that seem natural but aren’t, and show how the data impacts our communities. We can do all of this without having to become technicians who build programs or train models.

AI feels overwhelming, and the need to keep up seems unfair and biased toward the over-resourced, but if we apply our strengths to the question, we can offer our institutions something they might not otherwise get.

Question 4. Should institutions of theological education consider pursuing a kind of new monasticism, leaving the decadent dystopia into which we seem to be heading for the purity of quill and parchment? Why or why not?

Emily Peterson

While AI is not something I have naturally gravitated towards as a user, I have found it to be a fascinating topic of research and conversation because of the ways it draws out other important conversations about mission, values, and practice. That is why I found this question about a new monasticism especially interesting—it begs other essential questions that underpin the lives of our institutions. The appeal of the “new monasticism” response in the face of the risks of AI is also powerful, and it demands a direct treatment.

Each institution must determine its own mission-driven response to AI, yet I would argue that the most productive answer to this question holds across the field: “no, but. . . ” The “no” is not a wholesale embrace of AI, but rather a rejection of a fear-driven wholesale rejection that can have unconsidered consequences. Nuance and context are essential to expanding on that “no,” with particular attention to pedagogy, content, and practical applications.

At the heart of the matter of pedagogy is the question, How do we teach? And how does that affect how we do or do not integrate artificial intelligence into theological education? As much as libraries have often been at the forefront of adopting new technology, academic libraries in particular are often also a focus of academic fantasies that cast them as sacred and hushed intellectual sanctuaries, a vision that is underpinned in many ways by libraries’ historical relationship to monastic communities. The idea of the “purity of the quill and parchment” posed by the above question reflects this vision and comes with assumptions of certain pedagogical practices, particularly handwritten in-class exams and papers. I have heard multiple students and instructors suggest that blue books (a more “modern” iteration of “quill and parchment”) may be the best way to respond to the pedagogical threats of AI.

Yet I think it’s important to ask why we largely do not rely on blue books anymore. Are these “quill and parchment” methods actually “pure,” as our question implies? Were they ever clean, unadulterated assessments of learning? After all, people have been capable of plagiarism and cheating and shortcutting long before we even had typewriters, let alone computers. To set forth any single method as “pure” is fantasy, not reality.

Our understanding of effective and inclusive teaching methods has evolved over time, not only in response to the availability of new technologies, but also as those new technologies have helped to solve identified problems in the methods we were using—not to mention as we learned more about the diversities in how we learn. We have adopted new methods of teaching and assessment (both product- and process-based) because we have confronted convincing reasons for change. Those reasons have not disappeared just because AI has arisen as a new challenge. Simply reverting to older methods of teaching and assessment, then, is not a responsible way of coping with the advent of AI.

It is important to ask ourselves why we romanticize that fantastical academy of the past when confronted with a difficult new challenge—and whether we should. History has much to teach us, but it cannot be our unchanging model for how things should be done now and in the future. Nostalgia for a pre-AI era cannot solve the challenges of the AI age. To practice fearful avoidance that idealizes the past poses the risk of moving us backwards into old problems in the interest of avoiding new ones, cheating our students out of an education that builds on everything we have learned about effective, inclusive, transformative, and accessible teaching.

Therefore, I propose that we adopt this question for thinking about pedagogy in response to AI: How can we use the toolkit of methods we have already developed as a field (including some of those “quill and parchment” methods), combined with new advances and thinking, to creatively address the pedagogical challenge AI poses in theological education? One idea that has emerged in conversations with colleagues at my institution has been leaning more into scaffolded project-based assignments so that instructors can peek “behind the curtain” of students’ processes in a more structured way. Or, if we want to actively integrate AI into pedagogical methods, we can take cues from the creativity of our students; for example, some students at CTS are having generative AI argue against their thesis statements to help them improve their arguments. This is a constructive use of generative AI that can help reinforce learning goals.

In addition to impacting how we teach, technology also impacts what we teach, and that is an essential consideration in relationship to AI, too. Our theological institutions seek to teach content that is applicable to the contexts and realities in which our students might find themselves either during their studies or once they graduate. We want their education to be relevant to the work they do, to help equip and prepare them for it. To that end, then, it’s not just about content knowledge—theology or church history or Bible. Our institutions also seek to teach skills, frameworks, and methods for meeting future challenges in the field. While some theological degrees live mostly in the theoretical realm, most of us are focused on teaching practitioners. These students are not preparing to live and work in a monastic setting after graduation. They are preparing for life and work in the larger world.

It is worthwhile to pause this discussion momentarily to acknowledge that I see value in monastic retreat, including for real-world practitioners. Many of us know firsthand the value to be found in the rest, discipline, and reflection found in such settings. To build our degree programs around such retreats, however, for people who are preparing to do difficult and immersive work in the real world is ultimately a practice of avoidance. If we seek to avoid engaging with AI by adopting a withdraw-and-seclude approach, we are failing to prepare our students for the world in which they will live and work when they graduate. This is not to say that all “new monastic” practices should be discarded in light of AI; rather, that our institutions’ engagement with those practices should be because of their benefits rather than, again, the fearful avoidance of new technologies.

The essential feature here is decision-making. To so strictly control a learning environment for people who are to become leaders of communities saturated in the technology, media, and culture of the modern world is to deprive them of the opportunity to reflect on how to make faithful and ethical decisions. A strong consideration of ethics has always been important in theological education, and it needs to be an area of equal or greater emphasis moving forward, particularly in relationship to AI.

To offer another counter-question, then: How can we help our students to develop more dynamic and critically-reflective theological frameworks for understanding this inevitably changing world and making faithful decisions within it—in this case, choices related to AI? The frameworks Allison presented above are helpful to this end, particularly the questions from Sacasas. This is also a place where Brady’s reflections about the role of theological librarians becomes especially pertinent. Librarians are already in the business of helping students make critically reflective choices about what is reliable, ethical, and scholarly. We can do the same with AI. Whatever our students may individually choose to do with their use of technology, they ought to understand why and be able to reason through the ethics of it themselves. Our libraries can be an essential partner in this process.

Lastly, institutions of theological education must consider what is realistic functionally beyond the curriculum itself, for those of us doing the day-to-day work of our institutions. AI (or at least generative AI) seems so new to us because it really only entered the popular conversation as something real, not hypothetical, with the public release of ChatGPT in November 2022; however, as Brady mentioned above, AI has existed long enough that it is firmly rooted in some of the technologies we already use, and its footprint is growing—not just in standalone products like ChatGPT or Claude or Perplexity, but also as integrated functionalities of other common technologies like Grammarly or various search interfaces. Companies tend to trumpet the addition of AI capabilities as a selling point, but in some cases, AI integration may happen more quietly without us even knowing it is there. If we choose to limit ourselves only to AI-free tools and technologies, we will find ourselves with slimmer and slimmer options over time.

In short, total avoidance of AI is more difficult than it seems and is ultimately unrealistic. Putting aside how issues of user agency and data privacy and usage make it nearly impossible to fully opt out of participating in AI anyway, attempting to do so would likely involve sacrificing many of the useful and untroubling tools we have come to depend on. Some of those AI enhancements can help make our tools more useful and efficient, our data more accurate.

A key final question related to realistic practical applications, especially when it comes to generative AI, is related to terms of enforcement and accountability. Many faculty and administrators want to be able to rely on tools like AI checkers to catch use of AI where it is not permitted, whether that is in applications for admission, in academic assignments, or elsewhere. But we know those tools are notoriously unreliable and will likely always be at least one step behind the exponential advancement of AI technologies. A wholesale ban on AI use is therefore exceedingly hard to enforce, at least with the tools and tech available to us right now.

Rather than attempting a complete opt-out, then, I invite us to consider—much like in our question about what we teach—how to ethically engage with AI. What decisions can we make institutionally to practice, model, and foster responsible use of AI? This is the kind of approach that helps to support critical thinking, effective leadership, and creative ministry.

So, then, should institutions of theological education consider pursuing a kind of new monasticism, leaving the decadent dystopia into which we seem to be heading for the purity of the quill and parchment? It may be an attractive route, but I argue no. We live in a complex world, and we are continuously learning from the realities that we’ve faced in the past and are facing today. Turning back to the past to solve our problems for us disregards the constructive learning that has allowed us to more faithfully and effectively fulfill our educational mission to prepare our students for the world in which they live.

Instead, I invite us to get creative with our pedagogy by combining our toolkit of established methods with new advances and thinking, to help our students develop more dynamic and critically-reflective theological frameworks for making faithful decisions related to AI, and to practice, model, and foster responsible use of AI institution-wide. It will take significant effort, collaboration, and imagination, especially given the pace at which AI is evolving. However, I believe these questions that AI is prompting us to ask can—if we receive them as an opportunity rather than as a threat—help us become more resilient and foster more resilience in our students as we all face the future together.

References

AI Research Group for the Centre for Digital Culture of the Dicastery for Culture and Education of the Holy See. 2024. Encountering Artificial Intelligence: Ethical and Anthropological Investigations. Edited by Matthew Gaudet, Noreen Herzfeld, Paul Scherz, and Jordan Joseph Wales. Eugene: Pickwick Publications.

AIAI Network. n.d. “Speaking of the Divine in an Age of Artificial Intelligence: An ALC on Theology and AI.” AIAI Network. https://www.aiai.network/speaking-of-the-divine-in-an-age-of-artificial-intelligence-an-alc-on-theology-and-ai/.

Al-Kassimi, Khaled. 2023. “A Postmodern (Singularity) Future with a Post-Human Godless Algorithm: Trans-Humanism, Artificial Intelligence, and Dataism.” Religions 14 (8): 1409. https://doi.org/10.3390/rel14081049.

Baytas, Claire and Dylan Ruediger. 2024. “Generative AI in Higher Education: The Product Landscape.” Ithaka S+R. https://doi.org/10.18665/sr.320394

Berry, Wendell. 1990. What Are People For? San Francisco: North Point Press.

Copeland, B. J. 2024. “Artificial Intelligence.” Britannica. Last updated July 2, 2024. https://www.britannica.com/technology/artificial-intelligence.

Eusebius. 1926. Ecclesiastical History, Volume I: Books 1–5. Translated by Kirsopp Lake. Harvard: Loeb Classical Library, 1926.

Gay, Craig M. 2018. Modern Technology and the Human Future: A Christian Appraisal. Downers Grove, IL: IVP.

Hanegan, Michael and Chris Rosser. 2023. “Artificial Intelligence and the Future of Theological Education.” https://bit.ly/theological-education-and-ai.

Herzfeld, Noreen. 2002. “Creating in Our Own Image: Artificial Intelligence and the Image of God.” Zygon 37 (2): 303–316.

Herzfeld, Noreen. 2023. The Artifice of Intelligence: Divine and Human Relationship in a Robotic Age. Minneapolis: Fortress Press.

Lee, Alexander. 2022. “The War Against Printing.” Engelsberg Ideas. Accessed April 26, 2024. https://engelsbergideas.com/essays/the-war-against-printing/.

Lee, Victor R., Denise Pope, Sarah Miles, and Rosalía C. Zárate. 2024. “Cheating in the Age of Generative AI: A High School Survey of Cheating Behaviors Before and After the Release of ChatGPT.” Computers and Education: Artificial Intelligence 7: 100253, https://doi.org/10.1016/j.caeai.2024.100253.

McGilchrist, Iain. 2019. The Master and His Emissary: The Divided Brain and the Making of the Western World. New Haven: Yale University Press.

McGilchrist, Iain. 2024. “Resist the Machine Apocalypse.” First Things. https://www.firstthings.com/article/2024/03/resist-the-machine-apocalypse.

Mikoski, Gordon. 2023. “Will There Be Anything Left for Us?” Theology Today 80 (2): 118–120.

McLuhan, Marshall. 1977. “Laws of the Media.” ETC: A Review of General Semantics 34 (2). (June): 173-79. https://www.jstor.org/stable/42575246.

Morelli, Michael. 2019. “The Athenian Altar and the Amazonian Chatbot: A Pauline Reading of Artificial Intelligence and Apocalyptic Ends.” Zygon 54 (1): 177–190.

Prior, Matthew T. 2020. Confronting Technology: The Theology of Jacques Ellul. Eugene, OR: Pickwick.

Sacasas, L. M. 2021. “The Questions Concerning Technology.” The Convivial Society. Accessed May 13, 2024. https://theconvivialsociety.substack.com/p/the-questions-concerning-technology.

Schuurman, Derek C. 2019. “On Kuyper and Technology, or How a Voice from the Past Can Speak to Our Digital Age.” Christian Scholars Review 49 (2): 161–174.

Singler, Beth. 2017.“An Introduction to Artificial Intelligence and Religion for the Religious Studies Scholar.” Implicit Religion 20 (3): 215–231.

Werner, Daniel S. 2012. Myth and Philosophy in Plato’s Phaedrus. Cambridge: Cambridge University Press.

White, Lynn, Jr. 1967. “The Historical Roots of Our Ecological Crisis,” Science: 1203–1207.

Williams, Damien Patrick. 2023a. “Any Sufficiently Transparent Magic,” American Religion 5 (1): 104–110.

Williams, Damien Patrick. 2023b. “Bias Optimizers: AI Tools Such as ChatGPT Appear to Magnify Some of Humanity’s Worst Qualities, and Fixing Those Tendencies Will Be No Easy Task.” American Scientist 111 (4): 204–207.