1. A user repeatedly gets ambiguous answers from an AI tool.

Apply
prompt-engineering strategies to redesign the prompt to improve
clarity and intent.
Prompt engineering is the process of designing clear and structured inputs to guide AI
models to produce accurate and relevant outputs .
Ambiguous answers usually occur due to unclear, vague, or incomplete prompts.
1. Identify the Problem
Ambiguity happens when:
 The prompt is too general
 No clear instruction is given
 Lack of context or examples
 Output format is not specified
👉 Example (bad prompt):
“Explain pollution” → Too vague → AI gives broad answer
2. Apply Prompt Engineering Strategies
(a) Make Instructions Clear and Specific
 Clearly state what the AI should do
 Use simple and precise language
👉 Improved prompt:
“Explain the causes of air pollution in 5 points”
(b) Provide Context
 Add background information to guide the AI
 Helps produce relevant answers
👉 Example:
“Explain the causes of air pollution in urban cities for school students”
(c) Specify Output Format
 Mention how the answer should be structured
👉 Example:
“Explain the causes of air pollution in 5 bullet points with simple language”
(d) Use Examples (Few-shot / One-shot Prompting)
 Provide sample input-output to guide the model
👉 Example:
“Example:
Cause → Vehicle emissions
Now list 5 causes of air pollution”
(e) Use Instruction Prompt Technique
 Add rules for tone, style, or behavior
👉 Example:
“Explain the causes of air pollution in a formal and easy-to-understand manner”
(f) Break Down Complex Queries
 Divide into smaller parts
👉 Example:
1. Define air pollution
2. List causes
3. Give solutions
(g) Iterative Testing and Refinement
 Try different prompts and refine based on output
3. Final Redesigned Prompt
👉 Well-structured prompt:
“Explain the causes of air pollution in urban areas in 5 bullet points, using simple language
suitable for school students. Include examples for each point.”

2. Build a zero-shot prompt to classify comments as positive or negative.

Explain why it qualifies as zero-shot
Build a zero-shot prompt to classify comments as positive or negative. Explain why it
qualifies as zero-shot.
1. Definition of Zero-Shot Prompting
Zero-shot prompting is a technique where the AI is given only instructions without any
examples, and it generates output using its pre-trained knowledge .

2. Zero-Shot Prompt for Sentiment Classification

👉 Prompt:
“Classify the sentiment of the following comment as Positive or Negative:
‘The product quality is excellent and I am very satisfied.’”
👉 Expected Output:
Positive
3. Another Example
👉 Prompt:
“Determine whether the sentiment of this comment is Positive or Negative:
‘The service was slow and disappointing.’”
👉 Expected Output:
Negative
4. Why This is Zero-Shot
This prompt qualifies as zero-shot because:
 ❖ No examples are provided before asking the question
 ❖ Only a direct instruction is given to the model
 ❖ The model uses its existing knowledge of language and sentiment
 ❖ It does not rely on pattern learning from examples

3. Develop a writing prompt that helps a student begin an essay on “The

Impact of Social Media on Teenagers.” Explain how the prompt stimulates
the writing process
Writing Prompt
👉 Prompt:
“Write an essay on The Impact of Social Media on Teenagers. Begin by describing how
teenagers use social media in their daily lives. Discuss both the positive effects (such as
communication, learning, and creativity) and negative effects (such as addiction, mental
health issues, and distraction). Conclude with your opinion on how social media should
be used responsibly. Write in simple language in about 200–250 words.”
2. How the Prompt Stimulates the Writing Process
The prompt helps the student in the following ways:
(a) Provides Clear Direction
 It clearly states the topic and what to include
 Reduces confusion and helps the student start easily
(b) Breaks Down the Task
 Introduction → usage of social media
 Body → positive and negative impacts
  Conclusion → opinion
👉 This structure makes writing more organized
(c) Encourages Idea Generation
 Mentions specific points like communication, addiction, mental health
 Helps students think of relevant ideas quickly
(d) Sets Output Expectations
 Specifies word limit and simple language
 Guides the student on length and style
(e) Reduces Writer’s Block
 Gives a starting point (“Begin by describing…”)
 Makes it easier to begin writing without hesitation
(f) Improves Clarity and Focus
 Keeps the student focused on relevant aspects of the topic
 Avoids off-topic writing

4. You are building a customer – support assistant using a Large Language

Model (LLM). Develop a prompt that ensures the model responds politely
and provides step-by-step troubleshooting guidance. Explain how your
prompt uses principles of prompt engineering
Designed Prompt
👉 Prompt:
“You are a polite and professional customer-support assistant. Always respond respectfully
and use simple, clear language.
When a user asks a question or reports a problem:
1. Greet the user politely.
2. Understand and restate the issue briefly.
3. Provide step-by-step troubleshooting instructions in numbered points.
4. Ensure the explanation is easy to follow for non-technical users.
5. Offer additional help at the end.
User Query: ‘My internet connection is not working properly.’”
 How the Prompt Uses Prompt Engineering Principles
(a) Clear Instructions
 The prompt explicitly tells the model what to do (greet, explain, guide step-by-step)
 This improves accuracy and reduces ambiguity
(b) Role Prompting (System Behavior)
 Assigns a role: “polite and professional customer-support assistant”
 Ensures tone and behavior remain consistent
(c) Output Structure Specification
 Step-by-step numbered instructions are required
 Leads to organized and easy-to-understand responses
(d) Context Provision
 Includes a sample user query (internet issue)
 Helps the model understand the situation clearly
(e) Simplicity and Clarity
 Instructs the model to use simple language for non-technical users
 Improves communication effectiveness
(f) Instruction Prompt Technique
 Provides specific rules like politeness, clarity, and structured guidance
 Ensures output meets desired quality

5. Develop a one-shot example to help an LLM convert active voice to passive

voice. Illustrate with the example and the instruction.
1. Definition of One-Shot Prompting
One-shot prompting is a technique where the model is given one example to understand the
pattern and then perform a similar task .
2. One-Shot Prompt Design
👉 Prompt:
“Convert the following sentences from active voice to passive voice.
Example:
Active: The teacher explains the lesson.
Passive: The lesson is explained by the teacher.
Now convert:
Active: The chef cooks the meal.”
3. Expected Output
👉 Passive:
“The meal is cooked by the chef.”
4. Explanation
 The prompt includes one example, which shows the transformation pattern
 The model learns how to:
o Identify subject, verb, and object
o Rearrange them into passive voice
 The instruction clearly specifies the task: “Convert active voice to passive voice”
👉 This helps the model generate accurate and structured output
5. Why It is One-Shot
 Only one example is provided
 The model uses this example to understand the format and apply it to a new
sentence
 Improves performance compared to zero-shot prompting

6. Build a creative prompt that encourages ChatGPT to propose futuristic

transportation ideas. Explain how this promotes imaginative thinking
1. Creative Prompt
👉 Prompt:
“Imagine you are a futuristic transportation designer in the year 2050. Propose 3 innovative
transportation ideas that could solve problems like traffic congestion, pollution, and long
travel times. Describe how each system works, its unique features, and how it improves
people’s lives. Be creative and think beyond current technology.”
2. How the Prompt Promotes Imaginative Thinking
(a) Encourages Role-Based Creativity
 Assigns a role (futuristic designer)
 Helps the model think from a creative perspective
(b) Uses Open-Ended Instructions
 No fixed answer → allows multiple possibilities
 Encourages innovative and diverse ideas
(c) Introduces Future Context
 Setting the year 2050 pushes thinking beyond present limitations
 Promotes futuristic and visionary ideas
(d) Stimulates Problem-Solving
 Mentions real-world issues (traffic, pollution)
 Encourages creative solutions to practical problems
(e) Allows Freedom of Expression
 “Think beyond current technology” removes constraints
 Boosts originality and imagination
(f) Provides Structured Creativity
 Asking for 3 ideas + explanation ensures organized thinking
 Balances creativity with clarity

7. Apply ethical prompt construction to minimize bias in AI-generated

content.
1. Introduction
Ethical prompt construction involves designing prompts that ensure AI outputs are fair,
unbiased, and responsible. Poorly designed prompts may introduce bias, stereotypes, or
unfair representations .
2. Strategies for Ethical Prompt Construction
(a) Use Neutral and Inclusive Language
 Avoid biased or leading words
 Ensure representation of all groups fairly
👉 Example:
Instead of “Why are some communities less capable?”
Use: “Discuss factors influencing opportunities across different communities.”
(b) Avoid Stereotypes and Assumptions
 Do not include assumptions about gender, race, or culture
 Keep prompts objective and fact-based
(c) Provide Balanced Context
 Include multiple perspectives in the prompt
 Encourage fair and unbiased responses
👉 Example:
“Discuss both advantages and disadvantages of social media usage.”
(d) Specify Ethical Guidelines
 Clearly instruct the AI to avoid harmful or discriminatory content
👉 Example:
“Ensure the response is respectful, unbiased, and does not promote stereotypes.”
(e) Use Conditional Prompts Carefully
 Avoid conditions that lead to biased outputs
 Ensure conditions are fair and inclusive
(f) Encourage Evidence-Based Responses
 Ask for factual, data-driven answers
 Reduces opinion-based bias
👉 Example:
“Support your answer with facts or general observations.”
(g) Review and Refine Prompts
 Test prompts and improve them iteratively
 Identify and remove bias through experimentation
(h) Promote Transparency and Fairness
 Encourage clarity and accountability in responses
 Avoid misleading or manipulative instructions
3. Example of Ethical Prompt
👉 Biased Prompt:
“Why are women better leaders than men?”
👉 Ethical Prompt:
“Discuss the qualities of effective leadership without bias toward gender. Provide a balanced
perspective.”

8. Develop a use-case scenario where prompt engineering improves

communication clarity in cross-cultural conversations.
. Use-Case Scenario
A multinational company has employees from different countries (e.g., India, Japan, and the
USA) working together on a project.
They use an AI assistant to communicate and draft messages.
However, misunderstandings occur due to:
 Differences in language
 Cultural tone (formal vs informal)
 Misinterpretation of phrases
👉 To solve this, prompt engineering is applied.
2. Designed Prompt
“Translate and rewrite the following message for a Japanese colleague.
Ensure the tone is polite, respectful, and culturally appropriate. Avoid slang and informal
expressions.
Message: ‘Please finish this task quickly.’”
3. Improved AI Output
“Kindly complete this task at your earliest convenience. Your effort is greatly appreciated.”
4. How Prompt Engineering Improves Clarity
(a) Provides Cultural Context
 Specifies target audience (Japanese colleague)
 Helps AI adapt communication style
(b) Ensures Appropriate Tone
 Instructions like polite and respectful avoid unintended rudeness
(c) Removes Ambiguity
 Rewrites vague or harsh phrases into clear, meaningful communication
(d) Reduces Language Barriers
 Helps in translation and culturally correct phrasing
(e) Improves Understanding
 Makes communication clear and acceptable across cultures
(f) Promotes Effective Collaboration
 Prevents misunderstandings in global teams

9. Design a Few-Shot Prompt that trains ChatGPT to classify customer

feedback as positive, neutral, or negative
1. Definition of Few-Shot Prompting
Few-shot prompting is a technique where the model is given a few examples (2–5) to learn
the pattern before performing the task .
2. Few-Shot Prompt Design
👉 Prompt:
“Classify the sentiment of the following customer feedback as Positive, Neutral, or Negative.
Examples:
Feedback: ‘The product is amazing and works perfectly.’ → Positive
Feedback: ‘The product is okay, nothing special.’ → Neutral
Feedback: ‘I am very disappointed with the quality.’ → Negative
Now classify:
Feedback: ‘The delivery was fast but the packaging was average.’”
3. Expected Output
👉 Neutral
4. Explanation
 The prompt provides multiple examples (3)
 Each example clearly shows:
o Input (feedback)
o Output (sentiment label)
 The model learns the pattern and applies it to the new input
5. Why It is Few-Shot
 Includes more than one example (unlike one-shot)
 Helps improve accuracy and consistency
 Useful for tasks like classification and sentiment analysis

10. Apply creative prompts to generate innovative ideas for a sustainable

startup project.

1. Introduction
Creative prompts are designed to stimulate imagination, innovation, and problem-solving,
helping generate new ideas for real-world challenges like sustainability .
2. Creative Prompt Design
👉 Prompt:
“Imagine you are an entrepreneur building a sustainable startup in 2035. Generate 3
innovative business ideas that address environmental challenges such as waste
management, renewable energy, or water conservation.
For each idea, describe:
 1. The problem it solves
2. How the solution works
3. Its environmental and social impact
Think creatively and propose ideas beyond current technologies.”
3. Example AI Output (Sample Ideas)
1. Smart Waste Segregation System
o Uses AI-powered bins to automatically separate waste
o Reduces landfill waste and improves recycling efficiency
2. Solar-Powered Water Purification Units
o Portable systems for rural areas
o Provides clean water using renewable energy
3. Urban Vertical Farming Pods
o Compact farms using hydroponics in cities
o Saves space, water, and reduces transportation pollution
4. How Creative Prompts Generate Innovative Ideas
(a) Encourages Imagination
 Future setting (2035) pushes thinking beyond current limits
(b) Promotes Problem-Solving
 Focuses on real-world issues like pollution and sustainability
(c) Uses Open-Ended Thinking
 No fixed answer → allows diverse and unique ideas
(d) Provides Structured Guidance
 Asking for problem, solution, and impact ensures clarity
(e) Stimulates Innovation
 “Think beyond current technologies” promotes originality
(f) Enhances Idea Generation
 Multiple ideas (3) increase creativity and variety

11. Develop imaginative prompts to enhance product design creativity in

engineering students.
Introduction
Imaginative prompts help engineering students think beyond traditional designs by
encouraging creativity, innovation, and problem-solving.
2. Imaginative Prompts
👉 Prompt 1:
“Imagine you are designing a product for a city in 2050 with zero pollution. Create a device
that helps reduce environmental impact. Explain its features and working.”
👉 Prompt 2:
“Design a smart product that can solve a daily problem faced by students (e.g., time
management, energy saving). Include unique features using modern technology.”
👉 Prompt 3:
“Think of a product inspired by nature (biomimicry). Describe how natural systems influence
your design and improve efficiency.”
👉 Prompt 4:
“Create a low-cost, sustainable product for rural areas that solves a real-life problem such as
water scarcity or electricity shortage.”
These Prompts Enhance Creativity
(a) Encourage Future Thinking
 Prompts like 2050 push students to think beyond current limitations
(b) Promote Problem-Solving Skills
 Focus on real-world challenges (pollution, daily problems)
(c) Stimulate Imagination
 Open-ended questions allow multiple innovative ideas
(d) Introduce Interdisciplinary Thinking
 Combines engineering with nature, sustainability, and technology
(e) Encourage Practical Innovation
 Prompts include real-life applications and usability
(f) Improve Design Thinking
 Students consider features, function, and impact

12. Apply LLM-based writing prompts to produce creative content for

digital marketing campaigns.
1. Introduction
 LLM-based writing prompts help generate creative, engaging, and audience-specific
marketing content by giving clear instructions, context, and tone.
2. Writing Prompt
👉 Prompt:
“You are a digital marketing expert. Create creative content for a social media
campaign promoting a new eco-friendly water bottle.
Include:
 A catchy tagline
 A short caption (50 words)
 3 hashtags
Use a friendly and persuasive tone targeting young users.”
3. Example Output
 Tagline: “Drink Clean. Live Green.”
 Caption: Stay hydrated in style with our eco-friendly bottle! Designed for modern
lifestyles, it’s reusable, durable, and planet-friendly. Make a smart choice today and
join the green movement.
 Hashtags: #EcoLife #SustainableStyle #GoGreen
4. How Prompt Helps Generate Creative Content
 Clear instructions → ensures relevant output
 Defined audience → improves targeting
 Specified format → structured content
 Creative cues (catchy, persuasive) → boosts engagement
 Context (product details) → meaningful content

13. Differentiate between Zero-shot, one-shot and few shot prompting

1. Definitions
 Zero-shot prompting:
The model is given only instructions without any examples and generates output
using its prior knowledge
 One-shot prompting:
The model is given one example to understand the task before answering
 Few-shot prompting:
The model is given multiple examples (2–5) to learn patterns and improve accuracy
 2. Comparison Table

Feature Zero-shot One-shot Few-shot

Examples
None One Few (2–5)
Provided

Guidance
Low Moderate High
Level

Better than zero-

Accuracy Lower (sometimes) Highest among three
shot

Learning Uses general Learns from one Learns patterns from

Method knowledge example multiple examples

Moderately Less flexible but more

Flexibility Very flexible
flexible precise

Simple tasks, Pattern-based Complex tasks,

Use Cases
general queries tasks reasoning

3. Example
 Zero-shot:
“Classify: ‘This product is great’ → ?”
 One-shot:
Example: “Bad service → Negative”
Now classify: “This product is great → ?”
 Few-shot:
Example 1: “Good quality → Positive”
Example 2: “Not worth it → Negative”
Now classify: “This product is great → ?”

14. Explain self-consistency prompting and how does it improve model

performance
1. Definition of Self-Consistency Prompting
Self-consistency prompting is a technique where the model generates multiple possible
reasoning paths or answers for a given problem and then selects the most consistent and
logical result.
2. How It Works
 The same question is processed multiple times
 The model produces different answers or reasoning steps
 The most frequent or logically correct answer is selected
 Ensures consistency and reduces random errors
3. Example
👉 Prompt:
“Solve the problem and verify your answer using multiple reasoning steps: What is 15 ×
4?”
👉 Model may generate:
 15 × 4 = 60
 15 + 15 + 15 + 15 = 60
 (10 × 4) + (5 × 4) = 60
👉 Final Answer: 60 (consistent across all methods)
4. How It Improves Model Performance
(a) Increases Accuracy
 Multiple reasoning paths reduce chances of wrong answers
(b) Reduces Random Errors
 Avoids single incorrect output
(c) Improves Logical Consistency
 Ensures answers follow correct reasoning
(d) Enhances Reliability
 Produces more trustworthy results
(e) Better for Complex Tasks
 Useful in problem-solving, decision-making, and analysis

15. Explain the steps for Effective Prompt Learning

1. Introduction
Effective prompt learning involves designing and refining prompts so that AI models
generate accurate, relevant, and high-quality outputs. It is an iterative process that
improves results step by step .
2. Steps for Effective Prompt Learning
(a) Understand the Task
 Clearly define what you want the model to do
 Identify the type of output (summary, explanation, classification)
👉 A clear goal leads to better prompts
 (b) Design Clear and Specific Prompts
 Use simple, precise, and unambiguous language
 Avoid vague instructions
👉 Improves clarity and reduces confusion
(c) Provide Context and Details
 Include relevant background information
 Helps the model understand the situation better
(d) Specify Output Format
 Mention how the answer should be structured (points, paragraph, table)
👉 Ensures organized responses
(e) Use Examples (One-shot / Few-shot)
 Provide sample inputs and outputs
 Helps the model learn patterns and improve accuracy
(f) Test and Experiment
 Try different variations of prompts
 Compare outputs to find the best one
(g) Refine and Improve
 Modify prompts based on results
 Remove errors and improve clarity
(h) Evaluate Performance
 Check output for:
o Accuracy
o Relevance
o Clarity
👉 Make final improvements if needed