One Task, Three Conversations

Subtitle: Ask naturally, add what matters, and choose the next turn

What you will make

Create a short comparison record showing how one low-risk task changes across three conversation moves: a natural first request, a clearer request, and a focused follow-up. The goal is not a perfect prompt or a polished answer. It is to notice what each move changes, what the chatbot assumes, and what still belongs to human judgment.

Lab materials

• 📓 Notebook · Lab Notebook — An optional place to record the work and download a local copy.
• 📎 Resource · Learner guide
• 📎 Resource · One Task, Three Conversations: Why the First Prompt Is Not the Whole Story
• 📎 Resource · Gallery
• 📎 Resource · Facilitator guide

Before you start

Level and format: Beginner; individual, pair, observation, or facilitated demonstration

You need the 📎 Resource · worksheet or 📓 Notebook · local Lab Notebook. You may use an approved general-purpose chatbot, compare the supplied examples, or observe a facilitator. A paid account is not required.

Use a fictional or ordinary low-risk task. Do not enter private, client, medical, legal, financial, employment, account, credential, safety, or another person’s information. If your task becomes consequential or uncomfortable, switch to the supplied game-night scenario.

The core idea

A prompt is a starting point, not a spell. A chatbot can only respond to the request, context, and conversation available to it. You can improve the fit by naming the goal, relevant context, desired output, and a useful boundary. You can then inspect the answer and choose a focused next move.

For a fair comparison, keep the task the same. Change how you ask for help—not the underlying goal.

Three conversation moves

Complete the activity here. Open the optional 📓 Notebook · Lab Notebook beside this page when you want a place to record your work.

1. Choose one low-risk task

   Pick an ordinary task such as drafting a fictional invitation, organizing public notes, brainstorming hobby options, or explaining a common term. Write one sentence describing what you are trying to accomplish.

   If you do not want to choose a task, use this supplied scenario:

   Draft an invitation for a fictional neighborhood game night. It will last 90 minutes. The audience may include adults and families. The date, location, food, accessibility arrangements, and contact details have not been confirmed.

   Use the 📎 Resource · conversation worksheet or Lab Notebook to keep the task steady across all three rounds.

2. Round one: ask naturally

   Write the request as you normally would. For the supplied scenario, begin with:

   Write an invitation for a neighborhood game night.

   Save the request and response. Do not repair it yet. Mark anything the response assumed, invented, left vague, or handled usefully.

3. Round two: add what matters

   Create a clearer request for the same task. Include:

   • the goal
   • only the relevant context
   • the desired output
   • one boundary that prevents an unhelpful or unsupported choice

   For the supplied scenario, use:

   Draft a friendly invitation for adults and families in a neighborhood. The game night will last 90 minutes. Use plain language and keep it under 120 words. Do not invent the date, location, food, accessibility arrangements, or contact details; put missing details in brackets.

   If the tool allows it, start this round in a fresh thread so the comparison reflects the new request rather than hidden context from round one. Save the new response and note what changed.

4. Round three: continue deliberately

   Stay in the round-two conversation. Ask one focused follow-up based on something you noticed. You might request a shorter opening, ask for assumptions, reorganize the answer, identify uncertainty, or request one alternative.

   For the supplied scenario, use:

   List the assumptions you made. Then make the opening shorter without adding any details.

   Save the follow-up and response. A useful next turn should address a visible need; it should not hand the entire task back to the chatbot.

5. Compare the three rounds

   For each round, record:

   • what became more useful
   • what assumption or unsupported detail appeared
   • what important information remained missing
   • what you chose not to share
   • what would need checking before use

   Variation between tools and fresh conversations is expected. This comparison shows visible behavior in one exercise; it does not prove that the same wording will always produce the same result.

Check your result

• Fit: Does the last response serve the goal and intended audience better?
• Support: Did the tool avoid presenting missing details as facts?
• Selection: Can you name the detail or boundary that mattered most?
• Ownership: Did you keep, reject, or revise suggestions deliberately?
• Verification: Can you name what must be confirmed outside the chatbot?
• Privacy: Can you explain what you withheld and why it was unnecessary?

A fluent answer is not evidence that its facts are correct or that the task was appropriate for AI. If the output could affect health, money, employment, reputation, rights, or safety, stop this beginner exercise and use an appropriate human or approved process.

You are done when

You have retained each part of this record:

• The low-risk task or supplied scenario
• The natural first request and what I noticed
• The clearer request and what changed
• The focused follow-up in the same conversation
• One assumption or unsupported detail
• One item requiring human review or verification
• One thing I deliberately did not share

Try another version

These are optional extensions, not additional completion requirements.

• Change only the audience and compare what the tool changes
• Ask for two formats—such as a short email and a checklist—without changing the facts
• Give the three requests to a partner without the responses and ask what each request makes easier or harder
• Repeat the exercise later or with another tool and record variation without declaring one prompt universally best