You have /5 articles left.
Sign up for a free account or log in.

An illustration of a woman dining with a robot.

simplehappyart/iStock/Getty Images Plus

Over the past year, I have been encouraging my students to “invite AI to dinner.” I’m an English professor, but I believe students in other disciplines can also benefit from extending a dinner invitation to a chat bot. As one of my students quipped about dinner with a chat bot, “AI spices things up!”

Let me explain.

“Dinner With AI” is the good-humored name I’ve given to a classroom assignment designed to cultivate certain habits of mind and help students write better essays. The class exercise is straightforward: We are discussing a text, in this case Ralph Waldo Emerson’s essay “Self-Reliance,” which the students have read for homework.

  1. I begin the discussion by providing background on Emerson, sharing my experience with the essay, outlining some of its historical context, mentioning what other writers and scholars have said about it, and connecting it with other materials we’ve read in class—fairly general stuff and a typical approach for us lit teachers.
  2. I then split the class into groups of two or three and ask each group to develop a couple of prompts about the Emerson essay for the chat bot (I use ChatGPT 4o). I must explain to students what a prompt means in the context of a large language model and how it differs from a traditional search engine query. The key difference: The AI prompt initiates an open-ended conversation, while a search term seeks specific information.
  3. Once each group has a couple of prompts, they choose one to write on the board. We then have about seven prompts to consider.
  1. This is where it gets interesting. We debate as a class which prompt to present to the AI. This part of the exercise is particularly engaging because it involves discussing what’s most significant to ask about the Emerson essay and the best way to phrase it. I often find that the students’ initial prompts are weak, requiring some coaching. Students struggle to move beyond the search-engine mindset. They tend to seek out a quick answer rather than formulate questions that lead to more thought-provoking, open-ended conversations. After revising the prompts, we vote on the prompt we want to submit to ChatGPT. In the case of the Emerson essay, this prompt won the contest (after a little extra crafting during our debate): “Could you help us better understand Ralph Waldo Emerson’s relationship to Romanticism in his essay ‘Self-Reliance’?”
  2. We include a little more information in the prompt (for instance, we tell the bot whom it is talking with—undergraduate literature students studying literary history—and that we are interested in having an ongoing conversation) and then I submit our prompt. A second later, we have a response. We read and discuss the response as a class. We ask, what’s most interesting in the AI’s response? What’s dull? What did the chat bot miss? What did we miss with our prompt? What direction should we take for our next prompt? Should we approach it with agreement or criticism? The goal is to push the chat bot into new areas, challenge it to explore different directions and make connections it hadn’t previously considered. Through class discussion, we develop a direction, craft a follow-up prompt and present it to the bot.
  3. The chat bot responds again and we repeat the process—discussing the bot’s response and deciding on the next step. This back-and-forth can continue for the rest of class and even into future classes.
  1. When we conclude, I download the chat bot discussion and share it with the students.
  2. Students then have a transcript of our discussion (or half a transcript, since it doesn’t capture all the verbal exchanges they had with each other in the classroom). They can use this material to begin writing their essays. I emphasize that their essays can emerge directly from the discussion—what position does your paper take within the larger conversation we’ve had among ourselves and with the chat bot?

Now, on to why I find this exercise valuable. Most significantly, it models a habit of mind, as well as academic writing, as a kind of conversation.

My students—like most students (and perhaps most people)—struggle with having good conversations. I don’t blame them, because having a conversation is extremely challenging. It demands higher-level skills: choosing a topic, understanding your relationship to it (do you agree or disagree with X or something else?), maintaining focus, considering the assumptions behind the topic, responding to others, ensuring the right tone, gathering and evaluating evidence … The list goes on. These are the very skills involved in writing an essay or engaging in any intellectual inquiry, regardless of the discipline.

However, I find that students rarely see their academic work as a conversation. They often think of knowledge as something their teacher, a textbook or now a chat bot delivers to them, or they believe they must somehow generate it from within themselves through a mysterious faculty called creativity or intellect. They often assume they need to discover a profound idea buried deep within Emerson’s essay, but this approach rarely succeeds.

Instead, I encourage them to view inquiry as a conversation with others. What do others in our classroom—fellow students, the professor, a chat bot—have to say about Emerson? Most important, how do you respond to them? The knowledge I aim to teach develops through responding to others, including a chat bot.

We could accomplish this end without the chat bot, but the bot makes the process a little easier. The exercise with the chat bot forces students to externalize the elements of a good conversation: identifying what’s interesting, noticing what’s missing, determining what comes next and figuring out how to respond. Plus, they see immediate responses to their decisions, which is pedagogically valuable.

I still hold the old-fashioned hope that one day my students won’t need a chat bot for conversations about Emerson or much else. But in a world where large language models may become a fact of life, one reigning fear is that students will stop thinking for themselves. However, I believe this assignment fosters the skills needed to think with (and even against) the chat bot, rather than letting the chat bot do the thinking for them.

There’s more to discuss, but I’ll conclude with a passage I always share with my classes. It is a passage often quoted by scholars of writing pedagogy who emphasize essay writing as a conversation. The passage makes more sense to my students after they’ve done the AI exercise. However, they won’t understand the sixth word of the first sentence— “parlor”—because they’ve never heard it before, so I change it to “dinner party” and, more recently, to “group chat”!

“Imagine that you enter a parlor. You come late. When you arrive, others have long preceded you, and they are engaged in a heated discussion, a discussion too heated for them to pause and tell you exactly what it is about. In fact, the discussion had already begun long before any of them got there, so that no one present is qualified to retrace for you all the steps that had gone before. You listen for a while, until you decide that you have caught the tenor of the argument; then you put in your oar. Someone answers; you answer him; another comes to your defense; another aligns himself against you … However, the discussion is interminable. The hour grows late, you must depart. And you do depart, with the discussion still vigorously in progress.”

Kenneth Burke, The Philosophy of Literary Form, 1941

Even though the discussion is interminable, I’ll sum up by saying I’ve found it helpful to invite the chat bot to the dinner party. There are countless ways to misuse this technology (and I discuss these with my students, too), but this approach has proven to be a productive way to use it.

Michael Millner is an associate professor of American studies and English at the University of Massachusetts Lowell.

Next Story

More from Teaching