Curation, Taste and the Writing Classroom

AI is the ill wind that solved my biggest problem teaching first-year college writing, which is that it’s not very fun. Teaching first-year composition presents a number of challenges, from students’ reflexive hatred of required courses to the administrative microtasks that clutter up any core course. The main drag, though, is that it’s dull. It’s hard to be inspired or inspiring about topic sentences.

But ever since ChatGPT arrived and demanded wholesale changes to how writing gets taught, I’ve been experiencing a renaissance of pedagogical creativity, sponsored by the evil robots here to kill us all.

My new pride-and-joy semester-length assignment for first years is called the Influences Project. In it, students draft a series of short pieces that tell the story of three generations of artistic influence inspired by a single work of their choosing. The assignment moves backward in time, asking students to pick a work of art that they love—from any genre and any era—and look for its artistic ancestors: Parent and grandparent works that foreshadow a thematic or stylistic development of their chosen work. Students then situate each of the three works within its historical context, telling the story of key events, forces and figures that created the conditions from which it emerged. It’s more or less Ancestry.com for art and ideas.

In a recent podcast, Anthropic co-founder Jack Clark uses the terms “schlep work” and “taste” to draw a line between labor that should be handled by AI and what should be done by humans—a distinction between predictable, formulaic tasks and projects that require creativity and feel worthwhile. While I remain freaked out about AI’s impact on the economy, I like Clark’s terms for describing the vibe shift I’ve tried to introduce in my writing classroom and the larger distinction I’d like my students to absorb.

Composing an essay in earnest is decidedly not “schlep work,” yet it’s clear that: 1) many of my students see it this way, 2) I am sometimes guilty of feeling this way about grading them and 3) Chatbots make mincemeat of traditional essay assignments. So, I’ve pivoted to what I call curation-based assignments.

Curation-Based Assignments

Curation is the work of selection and contextualization, a subtle art of argumentation and persuasion. It helps students understand and value their own experiences, perspectives and curiosity, and because it foregrounds the choices and rationales of the curator, it’s even capable of making AI a natural ally of student learning rather than an adversary. Locating these topics and texts is part of the intellectual work of the project, and we use AI to find them.

In fact, curation makes a live case for why human agency is so important—and why writing is a central vehicle for developing and expressing one’s perspective. Choice matters in essay work, too, but a curation project works like a mobile: shift one element and everything else must rebalance to restore the whole. And when the stakes of getting that balance right mean building a detailed, meaningful story around a text you love, students buy in.

Curation is also an effective way of incorporating AI into intellectual work in order to extend, rather than supplant, student agency. Using it openly allows you to discuss its capabilities and limitations, dispels the boogeyman, and makes the case for why human writing is important. AI can tell you all sorts of things about your favorite text and its period, but its prose has no resonance. It’s utterly forgettable and stripped of the personal context curation projects also require—that of a human mind processing the world through a work of art.

The Influences Project

Structurally, the Influences Project is like a museum exhibition of interconnected informative and reflective pieces that chart the pathways of a student’s interest and inquiry. Their initial enthusiasm for their anchor text becomes the trunk from which branches of curiosity extend toward kindred expressions of the past; as they research and write, the intergenerational bonds between history and art feel more real and more relevant.

The Influences Project foregrounds journaling and conferencing. Thoughts spill out into journals in the early days; as we add ancestor texts and historical context, we talk about writing as a tool for working through the unfamiliar. Over the term, we move between prompts that emphasize self-discovery through a text and ones designed to practice clarity, concision, and voice.

We conference at key stages: after the parent and grandparent texts are selected and the vertical axis thread is clear; as they select, research and write about their chosen historical context subjects; and finally in a grading conference where we review the completed project and discuss the journey of their inquiry.

Here’s a favorite example of how this works: A student selected One Day, a buzzy 2024 Netflix rom-com she was repeat-bingeing, which led her back to When Harry Met Sally (1989) and It Happened One Night (1934). The romantic obstacles of those utterly different eras startled her. She couldn’t wrap her mind around the “can men and women be friends?” debate of the 1980s; it seemed so obvious to her that they could.

Her journals about the films centered on the conflicts that arose for the three couples, so she focused her historical research on dating in each era. What began with a hazy affection for a feel-good comfort watch ended up as a study of how the central conflicts of romantic comedies mirror shifts in men and women’s social proximity, which has changed significantly over the 90 years of this cinematic sweep, as women entered professional workspaces and higher education at greater rates.

I loved that neither of us knew where this project would end up. College faculty know the circuitous routes and unforeseen outcomes of long-haul research; we trust the process to lead us somewhere meaningful. But that’s not how most students have experienced academic writing—they’re practiced in prescribed destinations and confining instructions. It takes a lot of iterative journaling to help them identify what about their anchor text actually calls to them: what makes them smile or think or feel something, not what is “most significant.” And it takes a lot of conferencing to help them trust that those interests are substantial enough for serious inquiry.

A Case for Using AI (a Bit)

I dabbled with curation projects in the past, but one of this scale wouldn’t have been as tractable in the Before Times. Three texts and nine historical context subjects is simply too large a research lift for most freshmen. We use AI to locate ancestor texts and identify historical context topics, then shift to traditional sources to actually learn about them. Without AI, students would burn through the semester just finding their subjects, poking through articles, interviews and forums before doing any real research or getting to the writing part. And I’d be wading through heaps of material to advise and verify along the way.

More importantly, students would probably default to safer, less exploratory options. AI not only delivers information quickly and concisely, it can also stretch the aperture of what we know. We prompt for range rather than the expected: not “what other horror films inspired Midsommar” but “give me a range of artistic influences.” Maybe a student picks The Wicker Man, but maybe brutalist architecture or Scandinavian folk art calls to them instead.

For most AI use, it’s best to use a really specific prompt to generate targeted results. But here I want expansive searches; I want the machine to take them somewhere weird. This helps me push students who are inclined to play it safe and work with selections they already know and to reassure those attracted to unusual connections that there’s probably something there. Does AI occasionally hallucinate an influence? Sure. But students read plenty of traditional articles on each of these texts and topics, so it’s pretty easy to catch any links that don’t fit.

Journaling and conferencing do a lot to prevent students from offloading cognitive labor. But obsessing over whether a robot or my student wrote this sentence or that one is a losing game, and it only makes comp class tedious. Instead, I want students to think cogently and critically about their stance on AI writing, and I want them to know how to distinguish between schlep work and taste. There’s plenty of writing tasks out there that are purely functional and may prevent us from spending time on things that matter; we might disagree on this point, but I feel no guilt when using AI to write a letter of recommendation for a student’s camp counselor summer gig.

Taste is what happens when our humanness matters, when selves formed from everything we love and remember reach out toward other selves, formed from their own passions and experiences. I want it to feel a little yucky to let AI have a voice in that. Asking them to use it here opens a real conversation about why our voices matter, why writing isn’t schlep work to be offloaded, but the very means by which we figure out what we think and who we are.

Most of my students report that this project meant something to them, that it was writing they wanted to do because it felt like their take mattered. That doesn’t mean they’re all perfect; there’s plenty of the inevitable rush jobs, uneven effort and yes, some prohibited GPTing. But I haven’t had a student yet who didn’t find something of themselves in this project. I now remember former students by their influences. When my middle-aged brain has deleted their names, I can still wave across campus with recognition at the student who watched Power for a vicarious sense of agency during COVID lockdown, the one who discovered loyalty and male friendship watching Teenage Mutant Ninja Turtles with his brother and the one who chose The Fran Lebowitz Reader because she grew up idolizing her humor and fashion.

Curation has given me a way of connecting with my first-year students more fully than I ever have before. I watch their curiosities pull them toward new ideas during a period when who they are is still so fully in the process of becoming. And that is a joy to teach.