Introduction

This book provides practical guidance for educators who are panicking about AI. Spoiler alert: There is no need to panic. But adaptation is needed, and this book provides concrete suggestions for assignments that eliminate the possibility of students using AI or get them to use it in meaningful, critical ways. Later in this introduction, you’ll find a guide to those assignments that will allow you to come back to this book and use it as a reference. It will also help you if you’re picking up this book the day before the semester starts and don’t have time for anything other than specific, concrete solutions.

If any of those scenarios apply to you, please feel free to skip ahead. We hope you’ll come back at some later point, however, because before offering assignments and pedagogical strategies, we talk you through in layperson’s terms what generative AI is and how it works.^[1] Understanding this technology—both its capabilities and its shortcomings—is crucially important. It is essential not only for effectively teaching students in the present and foreseeable future, but also for your well-being. Many educators have had feelings of anxiety and despair concerning AI. Some have already given up and quit teaching, while others are thinking about doing so. Reading the whole book will allow you to not only have hope for the future of humanity and the humanities but also understand why this technology is not a threat. Human beings are still essential in the era of AI. The humanities are still essential. And most excitingly, strategies that we recommend in this book are not just workarounds related to AI. They are pedagogically meaningful and will help your students learn more effectively. They are things that it would be worth implementing even if modern AI technology did not exist.

The book is primarily written for humanities professors at universities and colleges. That said, a large proportion of what is in this book will be readily applicable in teaching high school. For those teaching earlier grades, there will still be principles that are transferrable. We also think that everyone, even those outside of education altogether, will benefit from reading our explanation of today’s AI tools and technologies. There are widespread misunderstandings about the technology itself and its projected role in society’s evolution. The authors nonetheless know that if we tried to write for everyone, no one would find this to be the precise book they are looking for. We’ve written as university colleagues and are speaking first and foremost to our peers. We trust that we will not be understood thereby to be excluding anyone else. If you are trying to make sense of the current technology and what it means for where society is headed, at least parts of this book are for you. If you are an educator outside the humanities or teach students who are younger than college age, you will also find most of what we share here readily applicable and adaptable to your own situation and needs.

Why Us, and Why This Book?

Behind the book you are reading is a story of two colleagues who connected and became friends well before the appearance of ChatGPT. James McGrath is a humanities professor, now the chair of his university’s Department of Philosophy and Religious Studies. He has taught First Year Seminar, Global and Historical Studies, Texts and Ideas, and Perspectives in the Creative Arts courses in his university’s core curriculum, which he also directed for a few years. Being in Religious Studies, you may not be surprised that McGrath has taught in just about every part of his university’s core that connects with the humanities. Religion is a field, and all humanistic tools of inquiry are applied to it (as well as few others borrowed from the social sciences). Although absolutely a nerd interested in science, he is not a humanities professor who is also secretly a computer programmer. He dabbled a bit but never got beyond the basics (and, for those old enough to remember, it literally was the BASICs).

McGrath’s interest in technology led him to connect with a colleague in computer science, Ankur Gupta, who is strong in all the places that McGrath himself is not. As the chair of his university’s Department of Computer Science and Software Engineering, Gupta is almost a mirror image of McGrath. He also shares interests in the humanities but hasn’t had much direct connection with them. Before meeting McGrath, Gupta had begun exploring the intersection of computation with the humanities through Templeton Foundation–funded research on “Defining Wisdom”—the question of whether computers can be not only intelligent (e.g., able to play chess, understand human language, autonomously operate vehicles, and so on) but also ethical (i.e. able to decide the best course of action when the outcomes have impacts on living things). Discovering a shared interest that intersected from opposite directions, McGrath and Gupta started a conversation with a view to perhaps writing a book on AI and the humanities. The plan was to tackle topics such as autonomous vehicles, bias in search results, and beyond.

Then, while McGrath was on sabbatical, teaching a class as a visiting scholar at another institution, the generative AI chatbot ChatGPT became available.^[2] He quickly got his students to experiment with it and together figure out the potentials and shortcomings of this new technology. He and Gupta exchanged a lot of messages during this time. There were moments when McGrath thought for sure that this technology must be genuinely smart, truly understanding and reasoning in ways that went beyond what he had thought possible. Gupta patiently helped him understand what the technology was doing and how. Through conversations across areas of expertise, they came to a richer understanding of generative AI, what it means for the humanities, and, conversely, what the humanities mean for generative AI.

In this book, readers are in good hands. Genuine expertise in teaching across the humanities is coupled with genuine expertise in computer science and AI. Gupta has been actively involved in the field that developed this new and disruptive technology. McGrath has been experimenting with that technology in relation to his teaching, ensuring students continue to do the things they must do themselves but also exploring how to get them to use AI in meaningful ways. Both professors have watched this technology evolve. The reassurances in this book about the humanities being more essential than ever, with no possibility that that will change, are grounded in a firm understanding of what generative AI is and how it works, and of what it takes to learn in the humanities.

McGrath—a humanities professor, remember—figured this stuff out and found solutions. You can, too—or, at the very least, you can feel free to borrow his. That is the takeaway message that both authors want to drive home before moving on to any other details. What humanities professors do will always matter, both inside and outside the classroom. In fact, students learning what humanities professors teach matters now more than ever before.^[3]

We should also clarify that most of the book uses “I” rather than “we.” You may think that you can safely assume that McGrath wrote the humanities sections and Gupta the tech ones. In fact, we both worked on both aspects of the book. It seems convoluted to add an individual author’s name each time “I” is appropriate, and using “we” everywhere will not work since not every story reflects both of our experiences. If your curiosity needs to be satisfied as to whose “I” it is at any given point, please write to the authors and they (I mean we) will tell you.

Something Not Entirely Unlike a Flow Chart

The second half of the book offers narrative interspersed with suggested assignments. That narrative is ideal for those reading the book straight through. However, one of the authors (yes, confirming your stereotype at this point, it was indeed the one in the highly structured and mathematical field) worried that this way of presenting the material will make it harder to return to this book as a reference and find a particular assignment. It might also make it difficult for those readers who need to jump to the concrete suggestions because they’ve just picked up the book the day before classes begin.^[4] There will also be readers who already understand the technology and are ready to skip ahead. We designed the book to be used and useful in all of these ways. In this section, we provide a high-level outline that will allow you to use the book as a handy reference that you can return to over and over again.^[5] There will also be a brief appendix at the end of the book showing where to find assignments. (The combination of exploratory narrative and bullet point lists is probably just what you would expect when a computer scientist and a humanities professor write a book together.^[6])

• Chapter 1, “Understanding Generative AI,” explains what large language models (LLMs) are, what they do, and, most importantly, what they are not and do not do. We promise to make this exposition as accessible as possible.
• Chapter 2, “Non-Solutions,” addresses problems and limitations with things people are doing in response to AI. Some of these purported solutions may be worth implementing for other reasons, but in and of themselves they do not adequately address inappropriate AI use by students, and we explain why.
• Chapter 3, “The Process Is the Point,” introduces ways to explain to students the importance of doing work themselves even if they believe AI can do it for them. We suggest methods for having students write essays and offer a grading system to help address the use of AI. Both of these approaches also have wider pedagogical benefits that make them worth implementing, regardless of concerns about AI.
• Chapter 4, “Starting with AI,” offers several assignments that present students with LLM-generated output to use as a starting point for doing their own work. One option is to have students evaluate and annotate AI-generated text. Another is to require them to offer something better. There are also activities that ask students to interact with an LLM and you grade the human side of their interaction.
• In Chapter 5, “Getting More Detailed,” we explain how being detailed and highly specific with the questions students must address when writing about assigned reading, asking for reflection on the process of doing research, and having students interact with an LLM as a study partner can prevent misuse of AI as well as give you access to evidence of their learning process that you would not otherwise have.
• Chapter 6, “Getting More Creative,” introduces why games and game-like activities, presentations of research and conclusions through visual representation, and other creative forms of work are less likely to be done satisfactorily by an LLM.
• Chapter 7, “Students Using and Outpacing AI,” focuses on activities and assignments that show what humans can do that AI cannot, as well as some examples of how AI can assist humans in doing cutting-edge research. These include using local archives, working with less-curated materials, tackling problems for which widely accepted, satisfactory solutions have been identified, using AI to interact with primary texts and scholarship in languages the students do not know and are not studying, and more.
• Finally, in the epilogue, we offer a few parting thoughts and words of encouragement. This is followed by an appendix that directs you to specific assignments for specific purposes, which can serve as a reference guide for ongoing use of this book.