Using AI as a Collaborator
Think of AI as a partner in a conversation, not a machine that gives perfect answers on the first try. When you send a prompt, the AI doesn’t just look at that one message — it also considers everything said earlier in the conversation. This helps it understand your intent, maintain consistency, and refine its responses based on context.
This back-and-forth process is called iterative prompting. You guide the AI step-by-step — asking, checking, and adjusting — until the response meets your goals.
Here are some examples showing how iterative prompting works in practice:
Image generated using Microsoft Copilot, November 24, 2025.
Science Example: Photosynthesis
In this example, notice how the AI is able to refine tone and format throughout the conversation.
Prompt 1: “Explain photosynthesis for middle school students.”
AI Response:
“Photosynthesis is the process by which plants make their own food using sunlight, water, and carbon dioxide. It helps plants grow and provides oxygen for animals and humans.”
Prompt 2: “Rewrite this in a fun, engaging way for 7th graders.”
AI Response:
“Imagine plants as little chefs! They take sunlight, water, and carbon dioxide and whip up their own food. This magical process, photosynthesis, helps plants grow and gives us oxygen to breathe.”
Prompt 3: “Break it into 3 points and give an example. Also, connect it to computational thinking.”
AI Response:
Capture sunlight – like following step-by-step sensor readings in a program.
Make food – combine water + CO₂, similar to executing instructions in a program.
Release oxygen – producing outputs for living things.
Example: A sunflower soaking up sunlight and producing oxygen.
CT Connection: Decomposition (breaking into steps) and algorithms (stepwise processes) connect biology to coding.
History Example: American Revolution
Notice how the AI can make connections between previous prompts.
Prompt 1: “Summarize the American Revolution for high school students.”
AI Response:
“The American Revolution was a war between the American colonies and Britain from 1775 to 1783. The colonies wanted independence and won the war, forming the United States of America.”
Prompt 2: “Explain why the colonies wanted independence and what the result was.”
AI Response:
“The colonies wanted independence because they were taxed without representation and felt their rights were being ignored. They won the war and established the United States of America with a new government.”
Prompt 3: “Combine the summary and cause-effect explanation into a structured study guide with nested bullets highlighting causes, key events, results, and important decisions.”
AI Response:
Cause: Taxation without representation
Key Events: Formation of militias, Boston Tea Party
Result: Colonists demand independence
Important Decisions: Organize resistance, protest through boycotts
Cause: Desire for self-governance
Key Events: Declaration of Independence (1776)
Result: Revolutionary War begins
Important Decisions: Draft and approve the Declaration, seek alliances
Cause: Oppression of rights
Key Events: Alliances with France, battles like Saratoga
Result: Victory and creation of the USA
Important Decisions: Negotiate peace treaties, establish a new government
Pedagogical Debate Example: Seating Charts
Notice how in this conversation the user challenges the AI's responses making it consider other options
Prompt 1: “Do you think fixed seating charts are a good idea in a middle school classroom? Explain.”
AI Response:
“I think fixed seating charts are a good idea. They help the teacher manage behavior and make the classroom more organized. Students know where to sit, which can reduce confusion.”
Prompt 2: “I see your points, but I’m not sure fixed seating is always better. Could you explain situations where it might actually hurt collaboration or student autonomy?”
AI Response:
“You’re right — fixed seating can have drawbacks. It can reduce student choice and limit interactions with different classmates. Some students might feel uncomfortable if they are always seated away from friends or grouped in ways that don’t fit their learning style.”
Prompt 3: “Okay, now weigh the benefits against the drawbacks. If you were a teacher, how would you decide whether to use a fixed seating chart or not?”
AI Response:
“If I were a teacher, I would weigh both sides. Fixed seating is useful for new students or classes with high behavioral needs because it provides structure. However, I would also allow flexibility: occasionally letting students choose seats or changing groups to support collaboration and student autonomy. Balancing structure with choice seems like the best approach.”
Using Multiple Prompts
The number of prompts needed depends on the complexity of the task. A simple recipe might take 4–5 prompts to get right. Planning a full lesson or unit could take 20 prompts, while a large, multi-part project might require 100 or more. Iterative prompting is similar to collaborating with a colleague: the more guidance, feedback, and discussion you provide, the closer the final result will match your vision.
When to Refine and When to Restart
While AI's ability to maintain context through converation is useful in refining output, sometimes the AI can get anchored to prior instructions. It may keep generating material based on old context or follow a narrow pattern, producing responses that don’t improve even after refinement.
Knowing when to refine and when to restart is important:
Refine: The response is close to what you want but needs tweaks — for example, adjusting tone, length, or adding details.
Restart: The AI is stuck in a pattern, you are switching tasks completely, or you want a fresh perspective.
Strategies for restarting:
Start a new chat
Tell the AI to ignore previous prompts
Image generated using Microsoft Copilot, November 24, 2025.