Lesson 2: Prompt Engineering Fundamentals: Talking to AI Effectively

Lesson Overview & Activities

1. The Art of Asking: What is Prompt Engineering? (15 min)

* Activity: “Bad Prompt, Bad Result”: Show a very vague prompt (e.g., “Write something”) and its unhelpful AI output. Then show a slightly better one and its improved output.

* Mini-Lecture: * Definition: Prompt engineering is the process of designing and refining inputs (prompts) to guide AI models to generate specific, desired outputs.

* Why it Matters: AI is powerful but literal. It interprets your words, not your intentions. Good prompts unlock better results, save time, and reduce frustration.

* The “Garbage In, Garbage Out” Principle for AI.

* Discussion: “Have you ever asked an AI something and gotten a weird or unhelpful answer? What do you think went wrong?”

2. Anatomy of an Effective Prompt: Your AI’s Instruction Manual (25 min)

* Mini-Lecture: Breaking down the key elements of a strong prompt:

* Role/Persona: “Act as a…” (e.g., marketing expert, friendly assistant, critical editor).

* Task/Goal: What do you want the AI to do? (e.g., summarize, explain, brainstorm, draft).

* Context/Background: What relevant information does the AI need to know? (e.g., audience, topic, specific details).

* Constraints/Format: Any limitations or specific output requirements? (e.g., word count, bullet points, tone, language).

* Examples (Optional but powerful): Providing specific examples of the desired output.

* Interactive Exercise: “Prompt Builder”: In real-time, as a class, collaboratively build a prompt from scratch using a simple scenario (e.g., “Draft a short email to a colleague about a team meeting”). Instructor types as suggestions come in, demonstrating each component.

3. Hands-On Lab: Your First Prompts (40 min)

* Tool Introduction: Briefly review access to an LLM tool (e.g., ChatGPT free tier, Google Gemini, Copilot).

* Guided Practice – Task 1: Summarization: * Scenario: You have a long article/email.

* Prompt Elements Focus: Task, Context, Constraints.

* Challenge: “Summarize the following article in 3 bullet points, suitable for a busy executive.” (Provide a short text snippet). * Individual Practice: Participants try this, then share their prompts and results.

* Guided Practice – Task 2: Brainstorming: * Scenario: You need ideas for a marketing campaign.

* Prompt Elements Focus: Role, Task, Context, Constraints.

* Challenge: “Act as a creative marketing director. Brainstorm 5 unique ideas for a social media campaign promoting a new sustainable coffee brand. Each idea should include a catchy hashtag.”

* Individual Practice: Participants try this, then share and discuss the variety of ideas generated.

* Troubleshooting & Refinement: Instructor guides participants on how to tweak their prompts if the initial output isn’t ideal (“How can we make this better? What did the AI miss?”).

4. Iteration and the “Chat” Nature of AI (10 min)

* Mini-Lecture: * AI conversations are iterative: You don’t have to get it perfect in one prompt.

* Follow-up questions and refinements: “Expand on point #3,” “Make it sound more formal,” “Can you provide examples in a different format?”

* The importance of telling AI what not to do, as well as what to do.

* Quick Demo: Show how to refine a previous output with a follow-up prompt.

🛠️ Tools & Resources

• Presentation Slides: Key concepts, prompt breakdown, examples.

• Access to a free LLM: (e.g., ChatGPT, Google Gemini, Microsoft Copilot)

• “Prompt Engineering Checklist” Handout: A simple guide for building prompts.

• Text snippets for summarization practice.

• Scenario ideas for brainstorming practice.