How to Write Discussion Prompts Using ChatGPT

Before you prompt ChatGPT, decide what you want the discussion to accomplish. Are you trying to surface diverse perspectives on a topic? Help students apply a concept to their own situation? Create peer accountability for a practice exercise? The goal shapes the prompt. A "share your experience" discussion and a "debate the tradeoffs" discussion require fundamentally different questions. Tell ChatGPT the goal explicitly: "I want students to share how they applied this week's framework to a real client situation."

Paste in a summary of the lesson content — the key concepts, any frameworks or models you taught, the specific skills students practiced. ChatGPT writes better prompts when it understands the material students are responding to. A prompt that says "Discuss the concept of boundaries" is generic. A prompt that says "You just learned the three-part boundary statement framework. Describe a situation where you'd use it and what you'd actually say" is specific because it references what students just learned.

Be explicit in your instructions: "Generate discussion prompts where every question has multiple valid answers. No question should have a single correct response. Each prompt should invite students to draw on their own experience or take a position they can defend." Without this constraint, ChatGPT defaults to comprehension questions disguised as discussions — "What are the key benefits of active listening?" is not a discussion prompt no matter how you format it.

Read every prompt ChatGPT generates and ask two questions. First: does this question have multiple valid answers? If you can imagine only one reasonable response, it's a quiz question in disguise. Second: would you want to answer this? If you read the prompt and feel nothing — no curiosity, no urge to share your own take — your students will feel the same way. A good discussion prompt creates a slight itch to respond. If you don't feel that itch, revise until you do.

ChatGPT writes for a generic student. Your students are yoga teachers, or executive coaches, or watercolor artists, or dog trainers. Take each prompt and ground it in the specific world your students inhabit. Replace "a recent professional challenge" with "a client session that didn't go the way you planned." Replace "an example from your experience" with "a moment in your studio practice this week." The more specifically the prompt speaks to your students' daily reality, the more likely they are to respond with something real.

The best discussion prompts connect course content to students' real lives — and ChatGPT does not know your students' lives.

It doesn't know that your yoga teacher trainees are nervous about teaching their first public class. It doesn't know that your coaching students are struggling to charge what they're worth. It doesn't know that your health coaching cohort includes three single parents who do their coursework at midnight. These details shape which prompts land and which ones fall flat.

When you read ChatGPT's output and think "this is fine but kind of lifeless," that's usually the problem. The structure is sound but the emotional specificity is missing. Your students respond to prompts that name the thing they're actually feeling — the uncertainty, the excitement, the resistance. You have to add that layer yourself, because you're the one who has sat in those conversations and heard what your students really care about.

Discussion forums go quiet when students feel exposed. One way to lower the barrier: start your prompt with a brief personal disclosure. "I used to avoid raising my rates because I was afraid clients would leave. This week's lesson covers pricing strategy — share a moment when fear influenced a business decision you made." When you go first, students feel permission to be honest. ChatGPT can draft these vulnerability openings if you tell it what you're willing to share.

The richest discussions happen when students respond to each other, not just to the prompt. Add an explicit invitation: "After posting your response, read two classmates' answers and reply to one with a follow-up question." This turns a discussion board from a collection of parallel monologues into an actual conversation. You can ask ChatGPT to generate prompts with this structure built in.

"Reflect on your leadership style" is too broad to answer well. "Describe one decision you made this week where you chose between efficiency and inclusion" is answerable in two paragraphs. Narrow prompts produce longer, more thoughtful responses because students don't have to figure out where to start. When in doubt, make the prompt more specific, not less.

ChatGPT's default discussion prompts sound like they were written for a graduate seminar. "Analyze the implications of empathic listening in the context of professional coaching relationships" is technically open-ended, but no one outside a university would feel compelled to answer it. The academic register creates distance when you need intimacy. Rewrite prompts in the language your students actually use — the words they'd say to a colleague over coffee, not the words they'd write in a term paper.

It also generates yes/no questions disguised as open-ended ones. "Do you think active listening is important in coaching?" sounds open-ended but has only one socially acceptable answer. Everyone will say yes and add a sentence about why. That's not a discussion — it's a compliance exercise. Every prompt needs to create genuine space for disagreement or divergent experience.

The third pattern to watch for: missing emotional resonance. ChatGPT prompts engage the intellect but rarely engage the heart. The discussions that students remember — the ones they cite months later as turning points — are the ones that asked them to feel something, not just think something. "What scared you most about starting your practice, and how has that fear changed?" produces a fundamentally different conversation than "Describe a challenge you faced as a new practitioner." Same topic, different emotional depth.