Writing Across the Curriculum in the Age of AI

Writing is a powerful tool for processing information, integrating new learning with background knowledge, and developing new perspectives and ideas.  These educational goals are obviously relevant to all disciplines, and they might even be considered the ultimate goals of any educational encounter.  For this reason, Writing Across the Curriculum initiatives have fostered the integration of writing assignments into Math, Science, and even Physical Education curricula at every instructional level.  In all these contexts, writing helps students use the power of the written word to crystallize meaning out of the “total flow” of the learning environment.

The dramatic emergence of AI chatbots capable of generating commonly assigned styles of academic writing, however, has required teachers who use writing in their classroom to rethink how they do so.  In English classrooms, where writing assignments are the curriculum, the emergence of this kind of chatbot functionality has obviously been uniquely disruptive.  English departments in secondary and postsecondary environments are in the process of adapting to the new reality, and, in the best cases, this adaptation provides a rich opportunity to revisit the basic questions of what writing is and why it is worth teaching and learning.  

If a chatbot can write a five-paragraph essay about the pros or cons of gun control, for example, then why do we need to teach humans to do it?  The answer, obvious in retrospect, is that the piece of writing itself was never the point.  As process-oriented compositionists have been teaching us for many decades now, writing is a verb; it’s something you do with ideas. 

The traditional writing classroom can be thought of as a place where students are trained to be chatbots – to generate text in response to a prompt – in a way that perhaps mimics the sense in which Math students in the days before the pocket calculator were trained to emulate pocket calculators.  The chatbots have freed English teachers and their students to use writing to do the human work of reflecting on their ideas, groping toward new thoughts, and situating themselves within an ever-evolving intellectual landscape. 

The open-endedness of the English curriculum allows writing classrooms to experiment with strategies for adapting to the emergence of chatbot-generated text, and the same strategies can apply fruitfully to WAC contexts.  In a pre-AI educational environment, it was pedagogically sound for a Biology teacher to assign for homework a short essay explaining the stages of cell division.  Writing such an essay would encourage students to synthesize what they know into their own words and sentences in ways that help them to internalize this knowledge. 

In a post-AI landscape, however, this same assignment is doomed, not only because it is all too tempting for students to get a chatbot to write it for them, but also because of a background awareness that this kind of question falls under the domain of chatbot chores, in the same way that most people, when faced with a slightly complicated math problem, open their calculator app instead of working it out for themselves.  The same observation applies to other typical WAC writing prompts, such as, in Math, Explain how to calculate a percentage; in Engineering, Describe how an internal combustion engine works; in Social Studies, Write a research paper about a historical figure; etc.  In the Age of AI, these sound more like chatbot prompts than writing assignments.    

That doesn’t mean that Writing Across the Curriculum has become an obsolete notion.  Writing still plays a critical role in the assimilation and manipulation of knowledge in any discipline.  But it does mean that WAC assignments need to adapt to this new reality in ways that both avoid the banal pitfalls of chatbot litigation (according to this website, the student’s essay may be up to 75% AI-generated, etc.) and also take advantage of the exciting opportunity to rethink how writing can enhance intellectual engagement across the curriculum.

Considered as a verb, writing can be thought of as a kind of pump.  In the same way that the heart pumps blood to the body, writing circulates individual perspectives into the world.  Like a pump pulling water up from a well, writing slurps up unconscious insights and channels them into a visible gush of language.  Like a muscle getting pumped up by repeated flexing, writing builds neural tissue and stimulates synapse-formation.  Fortuitously, the most salient strategies for designing relevant writing assignments in the Age of AI – in English classes or across the curriculum – happen to spell out the mnemonic acrostic: PUMP.

Process-oriented.  In any age, the most effective writing assignments focus on the development of ideas, guiding students through a process of gathering information, conducting research, formulating a rhetorical strategy, drafting, revising, publishing, and self-assessment.  While it does require the investment of instructional time, process-writing encourages students (and teachers) to use writing as a tool for playing with ideas, something that emerges from the writers themselves and that then develops a life of its own through a kind of evolutionary process.  In STEM classes, process-based assignments that encourage students to keep a research notebook and to develop their observations into experiments, proposals, and publications mimic and indeed replicate the circumstances of professional scientists, who regularly use writing in this way.  The kind of writing that emerges from this process is also chatbot-resistant, since it lowers some of the barriers that tempt students to use chatbots and also because each assignment is part of a continuous series of refinements on the students’ unique styles of expressing themselves.      

Upfront.  Upfront writing is writing that students do in “real time” in the presence of a community.  It may be handwritten or typed, it may be through an LMS or in pairs or small groups, it may be in response to a prompt or it may be completely unstructured, but in any of these scenarios, it challenges students to respond, to create, and to inquire “on their feet.”  Unlike a class discussion, which can sometimes marginalize certain students, an active, real-time writing assignment engages every student, and the writing that they do can become the basis for whole-class or small-group discussion.  Many people who work with ideas acknowledge the value of such “flash” writing exercises to assist in the generation and clarification of ideas, especially in the presence of a community of supportive colleagues.  For students in any discipline, upfront writing activities encourage students to activate their knowledge about a topic, to feel around the edges of what they don’t know, and to articulate their own position in relation to the topics they write about.  Also, of course, the real-time condition of upfront writing makes it difficult to outsource this writing to a chatbot.

Multi-modal.  These days, the definition of “composition” has expanded to include a wide variety of expressive modalities.  At their best, assignments that ask students to represent what they know in the form of podcasts, videos, presentations, infographics, graphic design, and other media invite students to perform a kind of intellectual work that resembles the effect of “alphabetical” writing assignments.  Students take ownership of a body of knowledge by reconstituting it in their own “voice.”  When combined with a process-oriented instructional approach, writing itself can also play an important role in the development of these multimedia projects as students script out videos, compile notes for an infographic, or articulate feedback to their classmates’ work.  The emphasis that UDL places on celebrating multiple means of representation reflects an important insight into how engaging and inclusive these kinds of assignments can be. 

Personalized.  The most important ingredient in crafting relevant writing assignments in the post-AI environment is the students themselves – their perspectives, their voices, their thoughts, feelings, fears, and dreams.  We don’t have any further need for the five-paragraph objective essay on an abstract topic.  The chatbots are really good at writing them, and, as a writing teacher, I’m happy to cede this worthless ground.  What we desperately need are human beings, with their miraculous capacity for original ideas, inspired flights of fancy, creative possibilities, new connections, brave conversations, and wild speculations, their tentative insights and unconscious intuitions, their vulnerability and empathy and hard-won inner wisdom.  This may be the most important clue to rethinking WAC in the Age if AI, because, while English teachers tend to celebrate the expansive human values of personalized writing, content-area teachers may not see the point in asking students to write about questions like:

·        What is your favorite stage of cell division and why? 

·        Describe the most recent time you had to do a math problem in your head.  What did you do and how did you feel about it?

·        How do you think humans will have evolved 10 million years from now?  Use evidence to support your prediction.

·        Would you rather be in a covalent bond or an ionic bond?  Explain your answer.

Questions like these are likely to provoke students into engaging with content knowledge in ways that are closely connected to their personal self-conception, allowing them to cultivate a more intimate relationship with new pieces of information.  These questions invoke the students’ experience as human beings, inviting them to use the lens of language to focalize their own ideas, simultaneously crafting a clearer picture of both the ideas they’re writing about and their own identity as subjects and agents.

So PUMP up your adaptability to the age of AI, PUMP up your students’ level of intellectual engagement, and PUMP new life into academic writing assignments across the curriculum.