Prompt Engineering

Introduction

Two colleagues Alex and Taylor are exploring the fascinating world of prompt engineering through a conversation. One of them is an expert in the field, while the other is a novice eager to learn. Through their interactions, the expert will teach the novice various techniques of prompt engineering. As they discuss and practice these techniques, the novice will gain a deeper understanding of how to craft effective prompts that can elicit accurate and useful responses from AI models.

The Basics of Prompt Engineering

This video showcases The Basics of Prompt Engineering.

Writing a clear, concise, and specific prompt:

  1. Instead of just saying “Summarize this,” you could say, “Please provide a concise summary of the main points in this document.”
  2. Another example of Summarization- “Rewrite the introduction and conclusion of this document to be more concise while preserving the main point”
  3. While asking a question - “Explain how user authentication works in web applications for a beginner in web development”

Providing clear context, and specifying the format of the output:

  1. Providing Context: “As a 7th-grade science student, explain the process of photosynthesis in simple terms”
  2. Transcribe your meetings the right way: “Analyze this meeting transcript and summarize the meeting, especially all the key points made by a particular person.
  3. Specify the output format: “Provide a detailed summary of my recent emails about Copilot, output as a table with 4 columns: time, sender, description, did I respond to it”

Common Mistakes to Avoid

This video showcases Common Mistakes to Avoid in Prompt Engineering.

  1. Being too vague: When a prompt is too vague, the LLM may provide a broad or irrelevant response. It’s important to be specific to get the desired output.

Vague prompt: “Tell me about technology.”

Specific prompt: “Explain the impact of artificial intelligence on healthcare.”

Vague prompt: “Give me an image of a car.”

Specific prompt: “Give me an image of a red Volvo XC90 from 2013.”

  1. Asking multiple questions in one prompt: Asking multiple questions in one prompt can confuse the LLM and result in incomplete or mixed responses. It’s better to split them into separate prompts.

Multiple questions: “What is LLM, and how does it work?”

Separate prompts: “What is LLM?” and “How does LLM work?”

Multiple questions: “Write a blog post about climate change and include recent statistics.”

Separate prompts: “List recent climate change statistics from 2023.” and “Create an outline for a climate change blog post.”

Advance Prompt Engineering

This video showcases Advance Prompt Engineering.

1. Prompt Chaining

Prompt chaining involves breaking down a complex task into a series of simpler, sequential prompts. Each prompt builds on the previous one to guide the model through a multi-step process.

Example:

Task: Generate a detailed travel itinerary for a week-long trip to Japan.

Step 1: Generate a list of must-visit cities in Japan. List the top 5 cities to visit in Japan for a week-long trip.

Step 2: For each city, list the top 3 tourist attractions. For each of the following cities: Tokyo, Kyoto, Osaka, Hiroshima, and Nara, list the top 3 tourist attractions.

Step 3: Create a day-by-day itinerary incorporating the attractions. Create a 7-day travel itinerary for Japan, including visits to the top attractions in Tokyo, Kyoto, Osaka, Hiroshima, and Nara.

2. Using Examples in a Prompt

Providing examples in a prompt helps the model understand the desired format and content of the response.

Example:

Task: Generate a product description for an e-commerce website.

Prompt:

Write a product description for a new smartphone. Here are some examples of product descriptions:

Example 1: Product: XYZ Smartphone Description: The XYZ Smartphone features a stunning 6.5-inch OLED display, a powerful octa-core processor, and a long-lasting 4000mAh battery. With its advanced camera system, you can capture stunning photos and videos in any lighting condition. The sleek design and premium build quality make it a perfect choice for tech enthusiasts.

Example 2: Product: ABC Laptop Description: The ABC Laptop is designed for productivity and performance. It comes with a 15.6-inch Full HD display, an Intel i7 processor, and 16GB of RAM. The 512GB SSD ensures fast boot times and ample storage for all your files. Whether you’re working on a project or streaming your favorite shows, the ABC Laptop delivers exceptional performance. Now, write a product description for a new smartphone.

Give us your feedback about this learning resource