Minus AI, Plus AI, Times AI
Thinking about the OpenAI Educator Forum and When (and Why) to Use AI
Key Takeaways from the TLDR Bot:
Students are caught between two AI worlds: professors who ban it and those who experiment with it. Both reveal how uneasy higher ed still is about AI’s role.
Treat AI like a “minus, plus, times” equation: subtract it to recover human presence, add it to expand possibility, and multiply it to boost creativity.
Banning AI isn’t the answer—thoughtful assignment design is. Personalized, engaging prompts make cheating with AI unnecessary.
The goal isn’t to make classes AI-proof, but AI-literate. Students need to know when and why to use it, not just whether they’re allowed to.
The future of learning might look a lot like the past: people gathered to talk about meaning—only now, with smarter tools in the mix.
As both an English Ph.D. student and a writing instructor, I’m living in two very different AI worlds right now. I’m already on the record for declaring that the college essay is dead–but writing and the humanities are still very much alive, and perhaps more critical to learning than ever.
Professor and data artist Greg Niemeyer described this tension at the October 20, 2025 OpenAI Educator Forum using a “minus, plus, times” framework. According to Niemeyer, AI is a “living ecology” that demands an education capable of distinguishing truth from fiction. “If we lose that,” he warned, “we lose meaning.” His challenge to educators captures the essential questions we all need to be asking ourselves:
Where should AI be subtracted to recover presence? Where should it be added to expand possibility? And where should it multiply our capacity to create and connect?
ChatGPT picked my “smart casual” outfit for the OpenAI forum. It was particularly insistent on the gold sneakers.
As a student, I’m seeing these questions play out in real time with two courses I am taking this semester: Philosophy of Language and Medieval Mystics. I love both classes, but my professors’ approaches to AI are very different, so I am experiencing firsthand the whiplash many students have right now as they try to navigate the range of AI policies in their classes.
In my Philosophy of Language course, AI is a forbidden fruit proscribed on the syllabus, a policy that the instructor carefully polices through in-class exams and Google Docs version histories. In Medieval Mystics, AI is an uninvited but intriguing guest at the seminar table, treated with a mix of curiosity and occasional well-deserved side-eye. Between these two pedagogical poles—resistance and exploration—I find myself constantly negotiating what it means to learn and teach in the age of generative AI.
Because here’s the thing: I don’t know that either approach is right or wrong. And as a writing instructor, I’ve taken a third path, integrating AI tools in my own English 102 courses in an attempt to multiply what students are capable of achieving on their own, with interesting results.
Minus AI: The Resistor
My Philosophy of Language professor is a proud resistor. He forbids AI use in his syllabus and attempts to design his course like a fortress against digital infiltration. Exams are handwritten. Essays are version-tracked in Google Docs, as if the timestamp might still guarantee authenticity (spoiler: it doesn’t.) I reminded my professor that the best way to get students to skip the ChatGPT approach is to explain why their authentic human interaction with the difficult texts we are reading (Wittgenstein or Kripke, anyone?) matter. I’ve also asked him to help me test out a potential new surveillance tool, My Course ID, with our midterm exam, because let’s face it: effective surveillance tools will likely be critical to the survival of asynchronous online courses in the age of LLMs. I’ll keep you posted on what that testing experience is like for me.
Plus AI: The Centaur
At the other end of the spectrum, my Medieval Mystics professor takes a “plus AI” approach—perhaps unknowingly echoing what In my Medieval Mystics class, AI is a tool, sometimes visible, sometimes not. We subtract it during discussions, where students lead questions and conversation—no slides, no scripts, just presence. And we add it when experimenting with research tools like Deep Research to explore medieval texts (here’s an example report Gemini “wrote” on The Pearl) .
As an example of how AI might be used as a multiplier, I shared Chaucerbot with my professor. Chaucerbot is a custom GPT I created to help me translate Middle English. I wasn’t outsourcing the hard work and friction necessary to learning. Instead, I was accelerating it, tracking my errors, and engaging with the language more deeply, improving my translations much more quickly than I would have with an old-fashioned dictionary (I also like the “gamified” aspect of learning that creating custom GPTs provides).
This is an image of an interaction with Chaucerbot.
Revising our Writing Assessments to be AI-Literate
This spectrum—from subtraction to multiplication—mirrors what three of my ISU colleagues, Suhaib H. Malkawi, Ify Ndukuba, and Mahnaz Poorshahidi, found when they studied faculty perceptions of AI in writing classes in the summer of 2025. Their research revealed that AI “misuse” correlates less with student intent than with assignment design. Instructors who personalize prompts, scaffold writing processes, and ground tasks in local or contextual material report far less AI reliance from their students.
The real vulnerability to AI misuse lies in generic assignments—those timeless “compare and contrast” or “summarize this article” tasks that practically beg to be automated. Honestly, I use AI for rough drafts of “boring” writing at work, so why should I be surprised if students take the same approach in my classes?
My colleagues’ findings reinforce something we’ve all suspected: thoughtful design reduces misuse and increases engagement. I strongly believe that the future of teaching isn’t AI-proofing—it’s AI literacy. Students need to know why they should or shouldn’t use AI, and we need to design assignments that make authentic thinking visible. The goal isn’t to ban the bots; it’s to make sure we know when, where, and how they can help–or harm–learning (plus, minus, multiply).
I guess I am learning to live in the tension between my two professors—the resistor and the experimenter, even as my own approach to integrating AI in my courses is a cyborg one. All three approaches are helping us to learn more about how AI impacts learning. Subtracting AI restores presence. Adding it expands possibility. I think we are all still learning what multiplying learning with AI looks like.
Of course, my humanities training makes me also want to imagine the future of education in the age of artificial intelligence as a return to a mythical past: Sitting under a tree, reading a book and talking with students about what makes life meaningful.
The future of education, as imagined by ChatGPT and me. ChatGPT says, “the professor under the neural tree isn’t rejecting AI; she’s dwelling thoughtfully within it. It’s an image of stewardship, not control. That’s the sweet spot for the humanities in this moment—witness, interpret, question, preserve meaning.”
What about you? Did you attend the OpenAI Educator Forum? If so, what stood out to you? Let me know on LinkedIn! And happy prompting!
AI disclosure: I’d put this one at 75% me (including 75% of the em dashes), 25% Liza Long persona bot. The rough draft was a conversation with the bot based on a lengthy prompt. I heavily revised and reorganized that draft to emphasize my own voice and experiences. I am so freaking sick of AI slop right now!