Activity 3.7a: AI Narrative Evaluation Activity
Learning Targets
I can evaluate AI-generated narrative writing and explain what human storytelling does differently.
You've spent this unit learning how narrative structure creates emotional impact. Now you'll examine whether AI can do what you've learned to do. You'll compare an AI-generated narrative scene with a student-written one and analyze the differences. This is more than identifying which is which — you're developing a critical eye for understanding what human storytelling does that AI cannot replicate.
Scene A: Dragon Encounter
The great dragon descended upon the battlefield with terrifying speed, its massive wings beating against the wind. The party members readied their weapons, knowing that this moment would determine their fate. The dragon's scales glinted in the afternoon sun as it released a torrent of flame from its maw. The barbarian charged forward with fierce determination, sword raised high, while the mage began weaving protective spells with urgent gestures. The ranger positioned herself at a distance, drawing her bow with steady hands. The tension in the air was palpable as each fighter awaited their chance to strike. The dragon roared, a sound so loud it shook the ground beneath them, and the battle commenced in earnest. Steel clashed against scale as the party advanced, each member executing their role in the coordinated assault. The dragon was mighty, but they were united in purpose. Victory was possible, and they would fight with all their strength to achieve it.
Scene B: Dragon Encounter
The dragon dropped from the clouds so fast Kess didn't see it until its shadow swallowed them whole. She was still reloading when it landed, and the impact threw her sideways into the cleric. Both of them hit hard. Breathe, she told herself. Reload. But her hands shook and she couldn't get the bolt in straight. The barbarian wasn't supposed to charge—they'd talked about this—but he did anyway because he was running on something that looked like courage and felt like recklessness. The mage screamed something about falling back, but the dragon was moving between them now, and retreat wasn't possible anymore. It wasn't like the maps they'd studied. It wasn't like the campfire plans. The dragon's breath came yellow and smelled like sulfur and meat, and Kess realized with a sick finality that they were going to lose. Not heroically. Messily. And she was going to die next to a cleric whose name she'd never learned, reloading a weapon that felt like nothing in her hands.
Analysis Directions
Respond to each of the following in your submission. Write 2–3 sentences for each question.
- Which scene creates more genuine tension? Point to specific moments.
-
Where does Scene A follow the "rules" of tension building but still feel flat? Why?
Consider: Does it show tension or tell it? Are the details generic? Does it feel like something actually happening to real people, or a recipe for a battle scene?
-
What does Scene B do that Scene A cannot replicate? Be specific.
Think about: unexpected choices, character-specific details, the writer's perspective, moments where structure bends or surprises.
- If you were running each encounter as a DM, which would your players find more memorable? Why?
- Final Reflection (3–4 sentences): Based on this analysis, what does human storytelling do that AI storytelling cannot? How does this change how you think about using AI in your writing process?
The Reveal:
Scene A was generated by an AI tool prompted to "write a 200-word dragon encounter scene with action and tension." Scene B was written by a high school student in this course.
If you identified them correctly, you're developing an important critical eye. If you got them mixed up, that's worth reflecting on too — what fooled you?
Sentence Frames for Support
Use these frames to help organize your analysis:
Requirements
- Responses to all 5 analysis questions (2–3 sentences each, 3–4 for the final reflection)
- Specific evidence from both scenes to support your analysis
- Thoughtful final reflection connecting analysis to your own writing practice
Rubric
| Criterion | Developing (1–2) | Proficient (3) | Exemplary (4) |
|---|---|---|---|
| Textual Evidence | Responses are vague or lack specific references | Responses reference specific moments from both scenes | Precise, well-chosen evidence supports clear analytical claims |
| Critical Analysis | Surface-level observations | Identifies meaningful differences in craft and effect | Connects craft analysis to larger ideas about storytelling and AI |
| Reflection | Oversimplified view of AI's role | Nuanced understanding of human vs. AI storytelling | Evidence-based position with implications for own writing practice |
How to Submit Your Work
Choose how to present your work:
Option A: Written Responses
Submit your responses to all 5 questions as a document.
Option B: Written + Annotated Text
Submit your responses along with annotated copies of both scenes, highlighting the specific moments you reference.
Option C: Audio + Written
Submit an audio recording (3–5 minutes) discussing your analysis, plus written notes for each question.
AI Guidance for This Activity
This is an evaluation activity—you are analyzing AI output, not using AI as a tool. Do NOT use AI to write your analysis or generate your responses. Your critical thinking about these scenes is the entire point. Do NOT use AI to help you think through these questions. The goal is to develop YOUR ability to see what AI can and cannot do. You're building judgment through practice, and that only happens when the analysis is genuinely yours.
What's Next
After completing this activity, move to Unit 3 Reflection to reflect on your overall growth in narrative structure, conflict, and pacing throughout this unit.