Module 2: Prompt Engineering Basics

Lesson 2.3 — Chains of Prompts / Multi-Step Interactions

Lesson Objective

By the end of this lesson, you will be able to:

• Explain what a prompt chain is and why it’s essential for complex ChatGPT tasks.
• Design multi-step interactions to guide ChatGPT’s reasoning process.
• Compare how ChatGPT-3.5, 4, and 5 handle multi-step logic and memory.
• Create structured workflows using prompt chaining for research, writing, coding, or planning.

1. What Are Chains of Prompts?

A chain of prompts means breaking down a complex task into a sequence of smaller, connected prompts.

Instead of giving ChatGPT one long, confusing instruction, you guide it step by step — just like teaching a student.

Example:

❌ Single, overloaded prompt:

“Write a business plan for a café, including market analysis, marketing strategy, budget, and risks.”

✅ Prompt chain:

1. “List five things to consider before starting a café.”
2. “Create a target market analysis for the café.”
3. “Write a marketing strategy based on that analysis.”
4. “Estimate a basic budget and potential risks.”

Each step builds on the previous one — producing deeper, more logical results.

2. Why Prompt Chaining Works

ChatGPT is strongest when it reasons incrementally.

Each answer it gives creates a “context memory” that it builds on.

By chaining prompts, you:

• Reduce confusion by tackling one idea at a time.
• Improve accuracy — fewer hallucinations or skipped steps.
• Gain flexibility — you can refine one stage without restarting.
• Mimic human logic — like outlining, drafting, and editing.

💬 Analogy:

Prompt chaining is like teaching a student to write an essay — you don’t say “Write the whole essay now.”

You say, “Start with the outline → write the intro → build the arguments → conclude.”

3.Types of Multi-Step Interactions

💡 Pro Tip: You can mix and match chain types for flexible, iterative workflows.

4.Real-World Examples of Prompt Chaining

🧑‍🏫 Teaching / Education

Step 1: “List key concepts in photosynthesis.”

Step 2: “Explain each concept in simple terms for a 10-year-old.”

Step 3: “Create a 5-question quiz to test understanding.”

👩‍💻 Coding

Step 1: “Plan a Python program that tracks expenses.”

Step 2: “Write the code for step 1.”

Step 3: “Debug and explain any issues in the code.”

🧾 Business Planning

Step 1: “Brainstorm 3 business ideas in eco-friendly fashion.”

Step 2: “Do a quick SWOT analysis for the best idea.”

Step 3: “Create a 6-month marketing plan for that business.”

✍️ Writing / Content Creation

Step 1: “Outline a blog post about the future of AI.”

Step 2: “Write the introduction.”

Step 3: “Write the body using data from step 1.”

Step 4: “Add a conclusion and call-to-action.”

5.How ChatGPT-3.5, 4, and 5 Handle Multi-Step Logic

💬 Tip:

If you’re using GPT-3.5, summarize the output of each step manually.

With GPT-4 or 5, you can simply say:

“Using everything we discussed so far, proceed to the next step.”

6.How to Build a Multi-Step Chain

Step 1️⃣ — Define the End Goal

“I want to create an online course outline about AI basics.”

Step 2️⃣ — Break It into Logical Steps

1. “List main topics in AI basics.”
2. “Create 3 subtopics for each main topic.”
3. “Suggest key learning objectives for each subtopic.”

Step 3️⃣ — Give Instructions at Each Step

At each stage, ask for a specific, limited output (lists, summaries, tables).

Step 4️⃣ — Connect the Steps

Reference the previous response:

“Using the topics you just listed, create subtopics.”

Step 5️⃣ — Refine and Summarize

At the end, ask ChatGPT to merge all steps into one cohesive output:

“Combine everything we’ve created into a final course outline.”

7. Sample Prompt Chain Exercise

Goal: Write a short research summary on “The Impact of AI on Healthcare.”

1. “List five ways AI is transforming healthcare.”
2. “Choose one area from your list and explain it in 150 words.”
3. “Find potential challenges with that application.”
4. “Suggest future improvements or innovations.”
5. “Summarize everything into a 200-word mini report.”

✅ Result: A clear, structured, well-reasoned essay — created collaboratively with ChatGPT.

8. Learn More with the Videos Below

They can learn additional concepts with the videos below:

🎥 How to Chain Prompts for Complex Tasks (Prompt Engineering Guide)

🎥 Building Multi-Step Workflows with ChatGPT

🎥 Reasoning in Steps: How to Get Smarter Responses from ChatGPT

Lesson Quiz 2.3

Please complete this quiz to check your understanding of the lesson. You must score at least 70% to pass this lesson quiz. This quiz counts toward your final certification progress.

Answer the quiz using the Google Form below.

Click here for Quiz 2.3

Conclusion

Chaining prompts is the key to turning ChatGPT into a true thinking partner.

By designing a structured conversation — step by step — you create more accurate, personalized, and reusable outputs.

This technique also helps ChatGPT-5 excel at multi-layered reasoning, from lesson planning to software development.

In the next lesson, 2.4 — Avoiding Pitfalls & Common Mistakes, we’ll look at how to recognize weak prompts, prevent hallucinations, and troubleshoot common errors in ChatGPT interactions.

Previous and Next Lessons:

Module 2: Prompt Engineering Basics, Lesson 2.1 — Anatomy of a Good Prompt

Module 2: Prompt Engineering Basics, Lesson 2.4 — Avoiding Pitfalls & Common Mistakes