1 Prompt Mastery – Introduction (5 mins)
Module 1: AI Prompting 101 – How to Get Better Responses
Intro:
AI is only as good as the prompts you give it. In this module, you’ll learn what AI prompting is, how AI processes your instructions, and simple ways to refine your prompts for better results.
Learning Objectives:
- Understand how AI generates responses and why prompting matters
- Learn the difference between vague vs. precise prompts
- Discover how small wording changes can dramatically improve AI output
- Practise refining a basic prompt to make it more effective
Written overview of video tutorial
AI is only as good as the prompts you give it. The way you phrase your request directly impacts the quality of the AI’s response. If you’ve ever received a vague, unhelpful, or overly generic answer from AI, it’s likely because the prompt lacked clarity, context, or specificity.
In this guide, we’ll cover:
- What AI prompting is and why it matters
- How AI generates responses (and why it sometimes gets things wrong)
- The difference between weak vs. strong prompts
- A step-by-step exercise to refine your prompts for better results
By the end of this module, you’ll understand how to communicate more effectively with AI and start getting more useful and accurate responses.
What is AI Prompting?
AI prompting is the process of giving instructions to an AI model like ChatGPT or Claude to generate a useful response.
Think of AI as a highly knowledgeable but overly literal assistant—it only understands what you explicitly tell it. Unlike humans, AI doesn’t infer meaning beyond what is provided in the prompt.
AI Prompting Defined:
“AI prompting is the process of structuring and refining instructions to guide an AI model’s response toward a specific outcome.”
To get the most relevant responses, you need to be clear, specific, and intentional about how you phrase your prompts.
How AI Understands Prompts
AI models don’t “think” like humans do—they generate responses by predicting the most statistically probable next words based on vast amounts of training data.
Here’s what happens when you type a prompt into an AI tool:
- The AI breaks down your text into tokens (word chunks).
- It analyzes the patterns and context based on its training data.
- It predicts the most likely sequence of words that follow.
Since AI doesn’t actually understand meaning the way humans do, it relies on the quality of the prompt to generate useful responses.
Key AI Limitations to Keep in Mind:
- AI doesn’t have opinions—it generates responses based on patterns, not independent thought.
- AI is only as accurate as its training data—it can make errors or provide outdated information.
- AI doesn’t automatically know what you need—if a response is vague, it’s likely because the prompt was vague.
Good vs. Bad Prompting
The way you structure a prompt determines the quality of the AI’s response. Let’s compare some examples.
Example 1: Vague vs. Clear Prompting
❌ Weak Prompt: “Tell me about AI.”
- The response will be too broad and lack depth.
✅ Improved Prompt: “Explain AI as if I’m 10 years old, using simple language and real-world examples.”
- The response will be clear, engaging, and easy to understand.
Example 2: Refining for a Specific Use Case
❌ Weak Prompt: “Summarise this report.”
- AI won’t know what level of detail you want.
✅ Improved Prompt: “Summarise this report in three bullet points, keeping each under 20 words.”
- The response will be concise and structured.
Example 3: Controlling Output Format
❌ Weak Prompt: “Give me marketing advice.”
- AI may provide general advice rather than something actionable.
✅ Improved Prompt: “Give me five social media marketing strategies for a small handmade jewellery business. Focus on Instagram and TikTok.”
- AI will generate specific, relevant advice.
How to Improve a Prompt in Three Simple Steps
If an AI response isn’t useful, don’t assume AI is “wrong” – instead, refine your prompt. Here’s a simple three-step process to improve any prompt:
Step 1: Identify What’s Missing
Look at the AI’s response and ask:
- Is the response too vague or generic?
- Does it lack structure?
- Did it miss key details?
Step 2: Add Clarity, Context, or Constraints
- Be more specific about the topic or expected response.
- Provide more context (e.g., target audience, expertise level).
- Control the format (e.g., bullet points, structured lists, specific word limits).
Step 3: Test and Refine
- Try running the improved prompt and compare the output.
- Make additional tweaks if needed.
- Save the best version of your prompt for future use.
📌 Example Prompt Iteration:
1️⃣ Basic Prompt: “Tell me about AI ethics.”
2️⃣ Refined Prompt: “Give me a short, three-bullet summary of AI ethics, with real-world business examples.”
3️⃣ Final Prompt: “Summarise the three biggest AI ethics concerns in business today, with an example for each. Keep each answer under 30 words.”
Each iteration makes the response more targeted and useful.
Practical Exercise – Improve a Prompt in Three Iterations
Try this yourself! Follow the three-step process using the basic prompt below:
Starting Prompt: “Give me productivity tips.”
1️⃣ First improvement: Add specificity (e.g., “for remote workers” or “for a busy professional with ADHD”).
2️⃣ Second improvement: Add a format constraint (e.g., “List five tips in bullet points, under 15 words each”).
3️⃣ Final improvement: Add real-world relevance (e.g., “Focus on time-blocking and task prioritisation”).
Compare the different responses you get from AI before and after refining your prompt.
Summary & Next Steps
Key Takeaways:
✔️ AI doesn’t think—it generates responses based on patterns in data.
✔️ Vague prompts = vague answers. Specific prompts get better results.
✔️ Small tweaks in wording can completely change the AI’s output.
✔️ Refining prompts is an iterative process—test, tweak, and improve!
📌 Action Step: Take a prompt you’ve recently used with AI and refine it using the three-step process. Notice how each small change improves the response.
🚀 Next Module: Now that you understand prompting, in the next lesson, we’ll explore custom instructions—how much should you tell AI upfront to get the best results? Let’s find out!