Basics-of-Artificial-Intelligence/C2/Prompting:-Persona-and-Giving-Examples/English
Tutorial Name: Persona and Giving Examples
Keywords: Prompting, Prompt Engineering, Artificial Intelligence, AI Education, AI Literacy, Communication with AI, Persona Prompting, Zero-Shot, Few-Shots,Spoken Tutorial, video tutorials
Pre-requisite Tutorial: The 3 Cs of Prompting
| Visual Cue | Narration |
| Slide 1
Title slide |
Welcome to this Spoken Tutorial on Persona & Giving Examples. |
| Slide 2
Learning Objectives Slide |
In this tutorial, you will learn to:
Use Persona Prompting Apply Zero-Shot Prompting, One-Shot Prompting, and Few-Shot Prompting Compare the results of different prompt types |
| Slide 3
System Requirements Slide |
Here I will be using:
A browser on a computer or mobile device; An AI tool such as ChatGPT, Gemini To follow this tutorial, you must be connected to the internet. No coding knowledge is needed. |
| Slide 4
Pre-requisite Slide Person sitting at laptop, light bulb animation |
To follow this tutorial, you should be familiar with the 3 Cs of Prompting.
'Clarity, Context, and Constraints. Pause whenever needed and practice each step with me. |
| Slide 5
Pre-requisite |
For the prerequisite Artificial Intelligence tutorials, please visit this website. |
| Graphic: Thinking human being | You may have noticed that AI answers are sometimes vague.
This happens when the AI is unsure how it should respond. We can fix this by guiding the AI to think and speak like an expert. |
| Section heading animation: Persona Prompting | To understand how, let us begin with Persona Prompting. |
| Screen showing a blank prompt box | Persona Prompting means telling the AI who to be while answering a query.
This helps the AI reply in a certain style or from a specific point of view. For example, it can speak like a teacher, a doctor or a chef. It can also respond like a designer or a travel agent. By giving a persona, you help the AI to focus its thinking. |
| Section heading animation: Science teacher persona | Let’s understand this with an example.
Open any browser. I am using the Firefox browser here. Next, open any AI tool. I am using ChatGPT here. Type the following prompt. “Create a lesson plan for teaching the water cycle to 6th graders.” Click the up arrow icon to begin the search. |
| Screen showing AI typing a basic plan | Look at the output.
Observe how it is organised. |
| Prompt box animation | Now, let me give the AI a persona and rewrite the prompt.
“Act as an expert science teacher. Create a 40-minute lesson plan to teach the water cycle to 6th graders. Include learning objectives, activities, and a short quiz.” |
| Screen showing structured lesson plan output
Side-by-side output (generic vs detailed lesson plan) |
Compare the two outputs.
When a persona is added, the AI response becomes more focused and useful. The clear role helps the AI give a better and more organized answer. This shows the power of Persona prompting. |
| Section heading animation: Giving Examples | Now, let’s learn the second technique, Examples-based prompting.
It means giving the AI examples to show the kind of answer you want. This technique helps AI understand required patterns and formats. It has three types. Zero-shot, one-shot, and few-shot prompting. Let’s understand each one with simple examples. |
| Whiteboard animation showing “Zero-Shot = No examples” | Zero-shot prompting means you DON’T give any examples in the prompt.
You just ask for what you want. Most prompts we use every day are like this. |
| Prompt box animation | Let’s try this prompt.
“Summarize this content into key points.” The AI answers directly, without seeing any example. |
| Whiteboard animation showing “One-Shot = One example” | The next type of example-based prompting is One-shot prompting.
Here, you give one example to AI of the type of response you want. |
| Prompt box animation
“Turn this sentence into a question like this sentence: ‘The sky is blue.’ → Output: ‘Is the sky blue?’ Use this format for the following sentence. Sentence: ‘Birds can fly.’ →” |
Now, I will enter the prompt shown on the screen in the prompt bar.
Submit it, and observe the output. Observe the output. |
| Screen showing well-structured AI output | By giving one example, the AI understands the desired format and follows it. |
| Whiteboard animation showing “Few-Shot = A few examples” | Finally, let’s understand Few-shot prompting.
It means you give two or more examples with the prompt. These examples show the AI what kind of output you want. Suppose your task is to write short product descriptions. You want them to sound catchy and engaging. First, give a simple or boring description. Then, give an improved and catchy version. From these examples, the AI learns the style and format you expect. This works well for creative or complex writing tasks. |
| Prompt box animation showing multiple examples | With more examples, the AI understands the pattern you want.
As a result, the answers become more accurate. |
| Side-by-side outputs (zero, one, few) | See how the output’s quality improves from zero-shot to one-shot to few-shot. |
| Summary Slide | Let us recap what we learned today.
Persona prompting means giving the AI a clear role to guide its response. Zero-shot prompting needs no example. One-shot prompting needs one example. Few-shot prompting needs two or more examples. More examples lead to clearer and better AI answers. |
| Assignment | For your assignment, open any AI tool you use.
Pick a topic you know well, like food, travel, or your professional activities. Try a zero-shot prompt by simply asking the question. Rewrite it using a persona to shape the response. Give one example to make it a one-shot prompt. Provide two or more examples to make it a few-shot prompt. Notice how the AI’s answers change at each step. |
| EduPyramids logo | This Spoken Tutorial is brought to you by EduPyramids Educational Services Private Limited.
Thank you for joining! |
