IDEAL COLLEGE OF ARTS AND SCIENCES

(AUTONOMOUS & NAAC A)

BSc, BCA, BBA, BCOM HONORS First Year - Semester- I
(A.Y- 2025-2026)
Paper Title: Introduction to Artificial Intelligence
Question Bank
Unit I:
Essays Questions:
1. Discuss about primary features and benefits of AI.
2. Discuss about Industry Applications of AI.
3. Write the Evaluation of AI.
4. Explain the concept of Machine Learning (ML).
5. Write about Natural Language Processing (NLP).
Short Answer Questions:
1. Define AI? List its types.
2. What are the major Challenges in AI.
3. Write a short note on Artificial General Intelligence (AGI).
4. What is Computer Vision?
5. What is Robotics?
Unit II:
Essays Questions:
1. Explain about the Applications of AI.
2. Discuss about the Applications of AI used in Healthcare.
3. Write the Applications of AI used in Transport.
4. Explain the role of AI in Finance?
5. Write the applications of AI in Agriculture?
Short Answer Questions:
1. Examine the challenges of AI in Healthcare.
2. Discuss the role of AI in Financial risk management.
3. Write the Applications of AI in Education.
4. Explore the impact of AI in e-commerce sector.

Unit III:
Essays Questions:
1. Describe ethics in AI and why is it important?
2. Discuss about bias and Fairness in AI Systems and how can it affect?
3. How does Transparency and accountability in AI systems?
4. How can we keep AI systems safe from hacks or attacks?
5. Why should AI protect (secure) our personal information and privacy?
Short Answer Questions:
1. Write importance of ethics in AI.
2. What is bias in AI?
3. How AI to be sustainable and help the environment?
4. How can AI models be made robust against attacks?
Unit IV:
Essays Questions:
1. Explain the role of AI in experimentation and multi-disciplinary research.
2. Explain about various generative AI Tools.
3. Describe the role of Prompt Engineering in AI/ML Interaction.
4. Explore the Emerging trends and Future Directions in AI.
5. Discuss about prompt engineering and its importance?
Short Answer Questions:
1. Write about Gemini and ChatGPT AI tools?
2. How does Generative AI create new things like text, images, or music?
3. How does AI help in different research fields?
4. Write about different deep learning models used in Generative AI?

Unit V:
Essays Questions:
1. Write Applications of Prompt Engineering used to help in Education.
2. What ways can businesses use prompt engineering to improve customer service?
3. How does AI help in Content Creation and creative writing?
4. What is SlidesGPT and how does it use prompts to make presentations?
5. Discuss about Designing Thumbnails and Channel Branding with AI?
Short Answer Questions:
1. Define Prompt Engineering in business applications.
2. How can AI generate video scripts using prompt engineering?
3. How does AI help in Creative designing?
4. Explain Applications of prompt engineering?
Unit- I
Q) Define AI? Discuss its primary features and benefits.
AI is the field of building machines or computer that can simulate human intelligence —
Like decision-making, learning, and problem-solving.
AI allows computers to perform tasks such as recognizing speech, making predictions,
solving problems, understanding language, and even creating new ideas.
Types of AI:
1. Narrow AI: Specific task (e.g., Siri, ChatGPT)
2. General AI: Human-level intelligence (still theoretical)
3. Super intelligent AI: Beyond human capability (hypothetical)
Examples for AI:
 Voice assistants like Siri, Alexa.
 Chatbots like ChatGPT, Gemini.
 Image and speech recognition systems
 Knowledge Engineering for answering questions.
Features of AI
 Automation – To perform tasks without human intervention.

 Learning and Adaption – AI systems improve with experience.

 Problem Solving and decision making

 Natural Language processing – Understand and processes human language.

 Computer vision – Recognising objects, faces, images, etc.

 Predictive Analysis - Forecasts trends based on historical data.

Advantages
 Accuracy and efficiency

 Automation i.e. performing repetitive tasks without fatigue.

 24X7 availability and continuously working without breaks.

 Handles large datasets quickly.

 Provides data driven solutions.

Q) Discuss about Industry Applications of AI.
AI is the field of building machines or computer that can simulate human intelligence —
Like decision-making, learning, and problem-solving.
The following are some of the applications of AI in various industries.

Banking and Finance

AI used to identify fraud transactions.

It provides 24X7 customer service.

To manage credit risk in banks.

Retail and E – commerce

AI used for product recommendation.

24X7 customer service.

For supply chain optimization.

Manufacturing

To maintain quality of the products by detecting defects.

Managing inventory problems.

Healthcare

Assist in medical diagnosis by analyzing medical images like X – rays, CT scans.

AI helps in drug discovery and development.

Provides robotic surgeries.

Education
Provide online tutoring.
Useful for language translation and learning.
Transportation
AI helps in traffic data management.
It is used to develop autonomous vehicles.
Provides dynamic pricing.
Agriculture
Helps in crop monitoring and in precise farming.
Predicting weather conditions.

Q) Write the History (Evaluation) of AI.

AI is the field of building machines or computer that can simulate human intelligence
— Like decision-making, learning, and problem-solving.
AI allows computers to perform tasks such as recognizing speech, making predictions,
solving problems, understanding language, and even creating new ideas.

HISTORY OF AI
1. Before Computers (Before 1940s)
 People dreamed about machines that could think.
 Example: Greek myths had robots made by gods.
 Mathematicians like George Boole created logic (true/false rules) — the basic
language of computers!
2. The Beginning (1940s–1950s)
 Computers were invented — now machines could store and process data.
 Alan Turing (1950) asked: “Can machines think?”
→ He created the Turing Test to check if a computer can act like a human.
 First AI programs were made to play chess and solve math problems.
3. The Birth of AI (1956)
 In 1956, scientists met at Dartmouth College in the USA.
 They officially used the name “Artificial Intelligence.”
 Early AIs could:
o Solve puzzles, Play simple games
o Prove math theorems
4. The First Boom (1956–1974)
 Everyone was excited!
 Governments gave money for AI research.
 AI programs learned to play checkers and translate languages.
 But computers were slow, and memory was tiny — progress stopped.
5. The AI Winter (1974–1980)
 People were disappointed
 Projects failed because AIs couldn’t solve real-world problems.
 Funding stopped — it was called the AI Winter (a “cold time” for AI)
6. Expert Systems (1980s)
 AI became useful again!
 Scientists made expert systems — programs that used knowledge to make decisions
like doctors or engineers.
Example:
 MYCIN helped diagnose diseases.
 AI was used in companies and factories.
7. Another Slowdown (Late 1980s–1990s)
 Expert systems were expensive to maintain.
 AI hype dropped again.
 But something good was coming...
8. Machine Learning Rise (1990s–2010s)
 Instead of programming every rule, computers started learning from data!
 AI could recognize speech, images, and handwriting.
 Example: Google, Facebook, and Amazon started using AI.
9. Modern AI (2010s–Now)
 Thanks to Big Data and powerful computers, AI got super smart!
 Deep Learning and Neural Networks can:
o Recognize faces
o Drive cars
o Translate languages
o Chat like humans
 AI assistants: Siri, Alexa, ChatGPT.
10. Future of AI
AI will help in:
o Education
o Medicine
o Robotics
o Space exploration
 But we must make sure it’s safe, fair, and used for good.
Summary Timeline
Period Stage Key Highlights
Before 1940s Imagination Myths, Logic, Math
1940s–1950s Early Ideas Turing Test, First Programs
1956 Birth of AI Dartmouth Conference
1956–1974 AI Boom Puzzles, Games
1974–1980 AI Winter Lost Funding
1980s Expert Systems Business Use
1990s–2010s Machine Learning Data & Learning
2010s–Now Deep Learning Modern AI Revolution

 Discuss the Major Challenges in AI

The major challenges in AI are as follows:

 Data Privacy:
Protecting people's personal information so AI doesn't misuse or expose it.
 Bias:
AI can learn unfair habits if trained on biased data, treating some people
unfairly.
 Transparency and Explainability:
AI often works like a black box, and it's hard to understand how it makes
decisions, losing people's trust.
 Ethical Concerns:
Making sure AI decisions are good and fair, without hurting people or breaking
rules.
 Job Displacement:
AI might take over some jobs, making people worried about losing work.
 Regulation and Standards:
Creating rules that keep AI safe and fair is tough because technology changes
fast.
  Security Risks:
AI systems can be hacked or tricked, which can cause big problems.
 Resource Intensiveness:
AI needs lots of power and data, which can be expensive and bad for the
environment.
 Lack of Skilled Workforce:
Not enough experts exist to build and manage AI properly.
 Interoperability:
Different AI systems often can't work well together, making it hard to combine
their powers.
These challenges need smart solutions to ensure AI benefits everyone safely and fairly while
avoiding harm or unfairness.

Q) Explain about various subfields of AI

Various AI Subfields are as follows

1. Knowledge Engineering:
Knowledge Engineering is about making computers think like human experts. It involves
collecting human knowledge, storing it, and creating rules so the computer can use that
knowledge to make decisions.
Example:
 In a medical expert system, the computer uses rules like “If temperature is high and
throat is sore, then possible fever.”
2. Machine Learning:
Machine Learning is a way to teach computers to learn from data instead of giving them
step-by-step instructions. To finds patterns, and makes predictions or decisions.
Types of Machine Learning:
 1. Supervised Learning – learn from labeled data (e.g., “this is a cat”, “this is a dog”).
2. Unsupervised Learning – find hidden patterns without labels (e.g., grouping similar
customers).
3. Reinforcement Learning – learn by trial and error (e.g., teaching a robot to walk).
Examples:
 Spam email detection.
 Movie recommendations on Netflix.
 Predicting weather or sales.
3. Natural Language Process:
NLP is the study of how computers can understand, interpret, and respond to human
languages. It connects computers with human communication.
Steps involved:
 Speech or text input (e.g., you say or type something).
 Language understanding (AI analyzes grammar, meaning, etc.).
 Response generation (AI gives an answer).
Examples:
 Chatbots like ChatGPT or Siri.
 Google Translate.
 Voice assistants (Alexa, Google Assistant).

4. Computer Vision:
Computer Vision allows computers to see and understand images or videos.
It uses cameras, sensors, and algorithms to analyze what is happening in an image.
Steps involved:
1. Image capturing (using a camera or sensor).
2. Image processing (cleaning or enhancing the picture).
3. Feature extraction (finding shapes, colors, or objects).
4. Understanding (recognizing what’s in the image).
Examples:
 Face recognition on phones.
 Traffic sign recognition in self-driving cars.
 Counting people in a CCTV camera feed.

5. Robotics:
Robotics is a branch of AI that deals with designing and building robots that can
perform tasks automatically.
Robots use sensors (to sense surroundings), actuators (to move), and AI algorithms
(to make decisions).
Types of Robots:
 Industrial Robots – used in factories for welding, painting, or assembling.
 Service Robots – used in hospitals, restaurants, or homes.
 Mobile Robots – like drones or self-driving cars.
Examples:
 Robot vacuum cleaners.
 Mars Rover exploring other planets.
 Surgical robots used in hospitals.
Q) Explain the concepts of Machine Learning (ML)
Machine Learning is a way to teach computers to learn from data instead of giving
them step-by-step instructions. To finds patterns, and makes predictions or decisions.
Types of Machine learning:

 Supervised Learning
In Supervised Learning, a model is trained on labeled data for prediction.
Features:
• Requires a training dataset.
• Model accuracy evaluated on test dataset.
Advantages:
• High accuracy if enough labeled data is available.
• Clear evaluation metrics.

Example:
 Spam email detection.
 Movie recommendations on Netflix.
 Predicting weather or sales.
 Disease diagnosis (presence or non-presence).
 Unsupervised Learning
Unsupervised Learning used to identify patterns, clusters and relationships in unlabelled data.

Features:
• No labeled data is required.
• Focuses on clustering, association and dimension reduction.
Advantages:
• Discover hidden patterns in the data.

• Useful to data exploration.

Examples:
• Market analysis by finding patterns in product purchases.
• Observing customers buying behavior.
• Detects credit card fraud.
Reinforcement Learning
In Reinforcement Learning, an agent (computer software/ machine/robot) learns to behave in
an environment by trial and error to make decisions.
Features:
• Based on reward feedback rather than labeled data.

• Focuses on sequential decision making.

Advantages:

• It handles sequential decision-making problems.

• Learns optimal strategies in dynamic environments.

Examples:

• Algorithms in playing games.

• Autonomous vehicles can improve decision making in traffic.

• Robots learn to navigate obstacles.

Q) Write about Natural Language Processing (NLP)
Natural Language Processing (NLP) is a field of artificial intelligence that enables
computers to understand, interpret, and generate human language, whether spoken or written.
NLP bridges the gap between computer understanding and natural human communication,
powering many technologies we use every day.

Steps involved in NLP:

 Speech or text input (e.g., you say or type something).
 Language understanding (AI analyzes grammar, meaning, etc.).
 Response generation (AI gives an answer).

Examples:
5. Chatbots like ChatGPT or Siri.
6. Google Translate.
7. Voice assistants (Alexa, Google Assistant).

Tasks of NLP:
1. Tokenization: Dividing text into smaller units such as words, phrases, sentences.
2. Named Entity Recognition (NER): NER detects and identifies named entities such as
names of persons, organizations and places.
3. Text summarization: Selecting key points, relevant content and rearranging important
sentences.
4. Parts of speech tagging: It identifies the grammatical role of each word in the sentence
(noun, verb, adjective, etc.)
5. Language translation: Converting text from one language to another.

Applications of NLP:
• Virtual assistants like Siri, Alexa, Google Assistants.
• Useful for language translation like Google translate.
• NLP useful for Sentiment analysis for determining customer reviews and opinions.
• Content generation i.e. writing reports, articles and summaries.
Q) What is Computer Vision
Computer Vision allows computers to see and understand images or videos.
It uses cameras, sensors, and algorithms to analyze what is happening in an image.
Steps involved:
• Image capturing (using a camera or sensor).
• Image processing (cleaning or enhancing the picture).
• Feature extraction (finding shapes, colors, or objects).
• Understanding (recognizing what’s in the image).
Examples:
 Face recognition on phones.
 Traffic sign recognition in self-driving cars.
 Counting people in a CCTV camera feed.
Applications:
• Computer vision useful in facial recognition systems to maintain security and to detect
suspicious activities.
• It is useful for autonomous vehicles to detect obstacles and traffic signals.
• It assists in medical imaging for diagnosing diseases from X – rays, MRI’s and CT scans.
• It inspects the quality of the products in industries.
Limitations:
• Computer vision requires high quality data. So poor images reduce accuracy.
• High computational resources are needed for real time applications.
• Complex environments (such as lighting variations) challenge object detection.
Q) What is Robotics?
Robotics is a branch of AI that deals with designing and building robots that can perform
tasks automatically. Robots use sensors (to sense surroundings), actuators (to move), and
AI algorithms (to make decisions).
Types of Robots:
 Industrial Robots – used in factories for welding, painting, or assembling.
 Service Robots – used in hospitals, restaurants, or homes.
 Mobile Robots – like drones or self-driving cars.
Examples:
 Robot vacuum cleaners.
 Mars Rover exploring other planets.
 Surgical robots used in hospitals.
Advantages:
• Performs dangerous tasks safely and efficiently.
• Reduces human error in critical tasks.
• Operates continuously without fatigue.
Limitations:
• High cost for development
• Requires skilled operators.
• Difficult to maintain.

Unit 2 Applications of AI
Q) Explain about the Applications of AI.
AI is the field of building machines or computer that can simulate human intelligence —
Like decision-making, learning, and problem-solving.
The following are some of the applications of AI in various industries.

Banking and Finance

AI used to identify fraud transactions.
It provides 24X7 customer service.
To manage credit risk in banks.
Retail and E – commerce
AI used for product recommendation.
24X7 customer service.
For supply chain optimization.
Manufacturing
To maintain quality of the products by detecting defects.
Managing inventory problems.
Healthcare
Assist in medical diagnosis by analyzing medical images like X – rays, CT scans.
AI helps in drug discovery and development.
Provides robotic surgeries.
Education
Provide online tutoring.
Useful for language translation and learning.
Transportation
AI helps in traffic data management.
It is used to develop autonomous vehicles.
Provides dynamic pricing.
Agriculture
Helps in crop monitoring and in precise farming.
Predicting weather conditions.

Q) Discuss about the Applications of AI used in Healthcare.

1. Medical Imaging and Diagnostics
AI algorithms, particularly deep learning models, analyze medical images such as X-rays,
MRIs, CT scans, and ultrasounds.
Applications:
• Detecting tumors, fractures, and abnormalities earlier and more accurately.
• Assisting radiologists with automated image interpretation.
• Enhancing image quality and reducing diagnostic errors.
2. Disease Prediction and Risk Assessment
AI analyzes patient data to predict the likelihood of developing diseases.
Examples:
• Predicting risks for diabetes, heart disease, stroke, or cancer.
• Monitoring chronic conditions through AI-driven predictive analytics.
• Identifying high-risk patients for early intervention.
3. Personalized Treatment and Precision Medicine
AI helps tailor treatment plans according to an individual’s genetics, lifestyle, and clinical
data.
Benefits:
• Optimizing medication doses.
• Predicting patient responses to treatments.
• Improving outcomes in oncology, cardiology, and neurology.
4. Drug Discovery and Development
AI accelerates the development of new drugs by analyzing biological data and predicting
molecular interactions.
Uses:
• Identifying promising drug candidates faster.
• Reducing research costs and time.
• Repurposing existing drugs for new diseases.
5. Virtual Health Assistants and Chatbots
AI-powered chatbots assist patients with routine health questions and guidance.
Functions:
• Symptom checking and triage.
• Appointment scheduling.
• Medication reminders and therapy support.
6. Robotic Surgery
AI-driven surgical robots assist surgeons with minimally invasive procedures.
Advantages:
• Higher precision and reduced human error.
• Shorter recovery times for patients.
• Real-time adjustments during surgery based on visual data.
7. Electronic Health Records (EHR) Management
AI streamlines the processing and analysis of patient records.
Tasks:
• Automated data entry and documentation.
• Identifying patterns in patient history.
• Improving billing and administrative workflows.
8. Remote Patient Monitoring
AI systems analyze data from wearable devices and sensors.
Applications:
• Continuous monitoring of vital signs.
• Early detection of abnormal health conditions.
• Supporting telemedicine for chronic care patients.
Q) Write the Applications of AI used in Transport.
1. Self-Driving Vehicles
 AI enables autonomous cars, trucks, and buses to sense their environment.
 Uses sensors (cameras, radar) to detect lanes, traffic, pedestrians.
2. Traffic Prediction
 AI analyzes real-time traffic data to control traffic lights.
 Predicts traffic jams and suggests alternative routes.
3. Route Optimization
 GPS apps (Google Maps) use AI to suggest the fastest routes.
 Considers traffic, road conditions, accidents, and weather.
4. Public Transport
 AI predicts passenger demand.
 Optimizes bus/train schedules to reduce waiting time.
5. Predictive Maintenance
 AI detects early signs of mechanical failures in vehicles (trains, buses, airplanes).
  Reduces breakdowns and costly repairs.
 Increases safety and reliability.
6. Logistics & Supply Chain Optimization
 AI optimizes delivery routes for trucks.
 Speeds up loading/unloading processes.
 Reduces transportation costs and fuel usage.
7. Smart Parking Systems
 AI finds empty parking spots in real time.
 Reduces time spent searching for parking.
8. Air Traffic Control
 AI predicts flight delays and optimizes flight paths.
 Increases safety and reduces fuel consumption.
Q) What is role of AI in Finance?
Artificial Intelligence (AI) plays a major role in improving accuracy, speed and decision-
making in the financial sector.
It helps banks, financial institutions and customers in various ways.
The main roles of AI in finance are:
1. Fraud Detection:
AI analyzes transaction patterns to identify suspicious activities and prevent fraud in
real time.
2. Risk Assessment:
AI helps evaluate the creditworthiness of customers by analyzing financial data,
behavior and repayment history. This improves loan approval decisions.
3. Customer Service:
Chatbots and virtual assistants powered by AI answer customer queries provide
account information and offer support 24/7.
4. Trading:
AI uses large amounts of data to make fast and accurate trading decisions. It helps
maximize profits by predicting market trends.
5. Financial Advice:
AI-based financial advisors (robo-advisors) help customers choose investments based
on their income, goals and risk level.
6. Fraud-Resistant:
AI secures online payments by detecting unusual login patterns, locations and devices.
7. Process Automation:
AI automates routine tasks like data entry, document verification, and compliance
checks, reducing human errors.
 8. Credit Scoring:
AI can assess credit worthiness more accurately by analyzing a wider range of
data points than traditional methods.

Q) Write the applications of AI in Agriculture?

In agriculture, AI enhances farming efficiency by monitoring crop through drones,
increasing crop health by detecting pests and diseases, optimum use of infrastructure and
irrigation facilities. AI enables farmers to maintain food safety, security and sustainability.
How it works
• AI powered drones and sensors collect agricultural related data.
• AI analyzes crop data to predict yield and optimize irrigation or fertilizer use.
Here are the major applications, organized clearly:

Applications:
1. Crop monitoring:
For detecting pests, diseases and to check crop if they need more nutrients.
Ex: Use drones for Spraying Fertilizers & Pesticides.
2. Precise farming:
For optimizing irrigation and finding best time to plant seeds.
3. Weather predictions:
AI analyses weather data to provide accurate forecasts.
4. Harvesting robots:
AI powered robots can pick fruits and vegetables without damaging.
5. Soil analysis:
AI checks soil quality and recommend the right type of the fertilizer for healthy plant
growth.
6. Smart Watering:
AI decides when plants need water so they don’t get too much or too little—like a smart
watering can.
7. Livestock management:
AI monitors the health of animals and its weight and temperature.
Autonomous tractors and machinery perform filed operations without human supervision.
Benefits
• Reducing wastage of water and resources.
• Early detection of Problems.
• Increases crop yield and quality.
• Less wastage and improves work efficiency.
• Improve live stock health
• Faster and accurate field monitoring

Short Answer Questions:

Q) Examine the challenges of AI in Healthcare.
1. Missing Data:
Hospitals have tons of data, come from different places like- hospitals, clinics, labs:
 written differently in different places
 missing data
 hard for computers to understand
2. Unfair Learning:
If the robot only sees examples from some kinds of people it may not learn how to help
everyone. This can make it unfair without meaning to be.
3. Hard to Understand:
AI sometimes gives answers without explaining why.
Doctors need reasons, not just guesses.
Example: if AI says “Don’t eat that!” but won’t say why frustrating
4. Rules and Safety:
There are strict rules to keep patients safe.
AI has to follow them, and the rules for AI are still being figured out.
5. Privacy Problems:
AI needs lots of patient data to learn, but that data must stay private.
Keeping patient information safe while still training AI is a big challenge.
6. Expensive and Hard to Set Up
AI systems need:
 special computers
 experts to run them
 training for doctors

Q) Discuss the role of AI in Financial risk management.

1. Checks if people can repay loans
AI looks at money habits to guess if they’ll pay back a loan.
2. Catches fraud
It spots strange or suspicious transactions, like sudden big purchases.
3. Predicts market problems
AI studies lots of numbers to warn when the stock market might crash.
4. Prevents mistakes inside the bank
It finds errors or system problems before they cause trouble.
5. Makes sure the bank has enough money
AI helps plan daily cash needs so the bank doesn’t run out.
6. Stops illegal money activity
AI finds patterns that might show money laundering or other crimes.

Q) Write the Applications of AI in Education.

In education, AI is used to adapt learning content enhances teaching, automate grading
and monitoring student progress.

Applications
1. Personalized learning: Adapts personalized learning materials and adapts contents based
on student progress.
2. Virtual tutors: Offers interactive feedback, support and explain tricky concepts and
answer questions.
3. Language translation: AI can translate languages and used to learn new languages.
4. Automated grading: automatically assigning grades and conducting exams.
5. Learning analytics: AI analyses students’ academic data and predict their performance.

Benefits:
• Improves learning outcomes
• Provides access to remote learning.
Q) Explore the impact of AI in e-commerce sector.
1. Personalization & Recommendation Systems
AI analyzes user behavior (clicks, searches, purchases) to deliver personalized product
suggestions.
Example: Amazon’s “Recommended for You” and Netflix-style product feeds in online
stores.
2. Intelligent Search & Product Discovery
AI improves search accuracy using NLP, image recognition, and voice input.
Example: Myntra and Pinterest allow “search by image” to find similar products.
3. Customer Service Automation
AI chatbots handle FAQs, order tracking, returns, and support 24/7.
Example: Flipkart’s AI chatbot helps users check delivery status instantly.
4. Dynamic Pricing
AI adjusts product prices in real time based on demand, competition, and stock levels.
Example: Uber surge pricing and Amazon’s frequent price changes.
5. AI-Driven Supply Chain & Inventory Management
AI forecasts demand, predicts stock shortages, and optimizes warehouse operations.
Example: Amazon uses AI to predict which items will be in high demand before holidays.
6. Fraud Detection & Security
AI identifies unusual transactions, fake reviews, and account fraud using pattern recognition.
Example: UPI’s AI models detect suspicious payment activity instantly.
Unit 3
Bias and Fairness in AI Systems
Q) What does ethics in AI and why is it important?
Ethics in AI refers to the principles, guidelines and standards that responsible, ethical
development and uses of AI systems by minimizing potential harm.
It involves addressing the moral and societal implications of AI technologies.
Ethics in AI systems include bias, fairness, transparency, accountability, privacy,
security, inclusivity, sustainability, robustness and reliability.

Ethics in AI:
 treats people fairly
 does not harm anyone
 protects people’s privacy
 is honest and transparent
 can be trusted
 is used responsibly
Importance of Ethics in AI
1. To prevent unfairness
AI can treat people unfairly if it learns from biased data.
Ethics helps ensure everyone is treated equally.
2. To protect privacy
AI uses lots of personal data.
Ethics makes sure that data is kept safe and not misused.
3. To avoid harm
AI decisions can affect health, jobs, safety, and finances.
Ethics ensures AI doesn’t cause accidents or hurt people.
4. To build trust
People will only use AI if they believe it is reliable, honest, and safe.
5. To make AI transparent
 Ethics encourages systems that explain how they make decisions, instead of acting like
secret “black boxes.”
6. To guide responsible use
Powerful technology needs rules.
Ethics ensures AI is used in ways that benefit society and not for harmful purposes.
Q) Discuss about bias and Fairness in AI Systems and how can it affect?
Bias in AI Systems:
Bias in AI means the system makes decisions that are unfair or favor one group over
another. This usually happens because the AI learned from data that was already biased or
incomplete.
Example:
If an AI was trained mostly on photos of light-skinned faces, it may not recognize dark-
skinned faces as well.
This is data bias, and the AI simply repeats what it learned.
Causes of bias:
 Biased or incomplete data
This happens when the data used to train the AI does not represent everyone fairly.
Ex: An AI trained on unfair data may start preferring male applicants, even when
women are equally or more qualified.
 Human choices
Sometimes bias comes from the choices made by developers
Ex: Labeling Data with Human Error,
When The AI learns:
“Doctor = male”
“Nurse = female”
It may incorrectly label a photo of a female doctor as a nurse.
Bias Affect AI
Bias can cause serious problems, such as:
1. Unfair Treatment
People may be wrongly denied jobs, loans, housing, or medical care because the AI learned
biased patterns.
2. Wrong Predictions
Biased AI may misidentify people, give incorrect scores, or make bad recommendations.
3. Discrimination
When an AI system treats certain groups unfairly
4. Loss of Trust
Users and communities may stop trusting AI systems if they see unfair or harmful decisions.
5. Legal and Ethical Issues
Companies can face lawsuits, fines, or public backlash for using biased AI systems.
Fairness in AI System:
Fairness in AI means making sure the system treats everyone equally, consistently,
and without discrimination—regardless of race, gender, age, disability, or background.
A “fair” AI gives similar results to different groups without favoring or harming anyone.
Measuring and detecting strategies of Bias and Fairness
Fair AI attempts to:
 Give similar accuracy across groups
 Avoid unfair advantages
 Prevent harmful outcomes
Q) How does Transparency and accountability in AI systems?
Transparency means being open and clear about how AI system works, how it makes
decisions, and what data it uses.
This helps people trust the system and check for fairness, bias, and ethical issues.

Main Parts of Transparency and Accountability

1. Explainability
This means the AI should be able to explain why it made a certain decision.
Ex: An AI that predicts diseases should show which symptoms or patterns made it give that
result.
2. Data Visibility
We should know:
 What data the AI was trained on
 Where the data came from
 If the data is good and fair
This helps us understand if the AI might be biased or limited.
3. Algorithmic Transparency
This means being open about:
 What type of model the AI uses
 How it works
 How it learns
Open-source tools help make algorithms more understandable.
4. Decision-Making Process
A transparent AI shows the steps it uses to reach a decision.
In important fields like banking or healthcare, users should know how the AI reached its final
answer.
5. Ethical Considerations
Developers should clearly mention:
 What ethical rules they followed
 What risks exist
 How they reduced unfairness or bias
6. User Interface
The AI system should have a simple, easy-to-understand design so users can see:
 What the AI can do
 What it cannot do
7. Documentation and Reporting
Good transparency requires:
 Clear documents about how the AI was built
 Regular updates about how it is performing
 Reports if anything goes wrong
8. Laws and Regulations
AI systems should obey laws about:
 Privacy
 Security
 Fairness
Following these rules builds accountability and trust.
Q) How can we keep AI systems safe from hacks or attacks?
1. Protect the Training Data
AI learns from data. If hackers change the data, the AI may learn wrong information.
How to protect it:
 Use secure storage
 Check data for tampering
 Limit who can access the data
2. Secure the Model Itself
Hackers might try to steal or modify the AI model, AI may learn Biased information.
How to protect it:
 Encrypt the model (lock it with strong security)
 Use access controls
 Monitor activity for suspicious behavior
3. Defend Against “Adversarial Attacks”
These are special tricks where attackers change the input slightly so the AI gets confused.
Example: Changing a few pixels in an image to make an AI misidentify it.
How to protect it:
 Use adversarial training
 Test the AI with tricky examples
 Add security layers that check inputs
4. Keep Software Updated
Old software can have security issues. They may breaks security and privacy.
How to protect it:
 Install updates regularly
 Patch vulnerabilities quickly
5. Use Strong Authentication
Only trusted people should be able to use or modify the AI. To prevent the unauthorized
access.
How to protect it:
 Use passwords and 2 factor authentication
 Limit admin access
6. Monitor the System for Strange Activity
Hackers often leave signs, like sudden changes in behavior. Like multiple signs may occur at
the same time.
How to protect it:
 Use continuous monitoring
 Set alerts for unusual behavior
 Review logs
7. Protect the AI During Deployment
When the AI is used publicly, it becomes easier to attack. It may fails security and privacy
feature.
How to protect it:
 Add firewalls
 Secure APIs
 Limit how much information the AI reveals
8. Test the AI Against Threats
Security experts can test how easy it is to attack the system. To know the ways to attack AI
System.
How to protect it:
 Do regular penetration tests
 Run red-team exercises (simulated attacks)
9. Use Privacy Techniques
Some attacks target personal data used by AI. To causes the Data Breach Risks.
How to protect it:
 Use techniques like differential privacy and security.
 Remove unnecessary personal data
10. Build AI with Safety in Mind
Include safety steps from the beginning, not after something goes wrong.
How to protect it:
 Follow secure design principles
 Use risk assessments
 Comply with cyber security standards
Q) Why should AI protect (secure) our personal information and privacy?
AI should protect personal information because privacy protects people, and without it, AI
could easily become dangerous, unfair, or un-trusted.
1. Prevent Harm
Personal data can be misused for identity theft, fraud, discrimination, and manipulation.
If AI systems leak or misuse data, people can be directly harmed in very real ways.
2. Maintain Trust
People will only use AI if they believe it handles their data responsibly. Strong privacy
protection builds public trust, which is essential for AI adoption and long-term success.
3. Respect Human Rights
Privacy is a fundamental human right. AI systems must operate in ways that respect
autonomy, dignity, and personal freedom—otherwise they can become tools of surveillance
or control.
4. Prevent Bias and Discrimination
If private data is improperly collected or used, AI systems may reinforce unfair patterns
related to race, gender, health, or socioeconomic status. Protecting privacy helps prevent
misuse of sensitive data.
5. Follow Laws and Regulations
Governments worldwide have privacy laws that require safeguarding personal information.
AI systems must comply with these laws to avoid legal and financial consequences.
6. Reduce Data Breach Risks
AI systems often handle large amounts of information. Without strong protection measures,
they become high-value targets for attackers.
Securing data reduces the likelihood and impact of breaches.
7. Give Users Control
Good privacy practices ensure that people know:
 what information is collected,
 why it’s used,
 How they can opt out.
This empowers users and supports informed decision-making.
Short Answer Questions:
Q) Write importance of ethics in AI.
Importance of Ethics in Artificial Intelligence (AI)
1. Ensures Fairness and Prevents Bias
Ethical AI helps reduce bias in algorithms so that AI systems treat all individuals
fairly, without discrimination based on gender, race, religion, or background.
2. Protects Privacy and Personal Data
Ethics in AI ensures responsible collection, storage, and use of personal data,
safeguarding users’ privacy and preventing misuse of sensitive information.
3. Promotes Transparency and Accountability
Ethical principles require AI systems to be understandable and explainable, so
decisions made by AI can be traced and responsibility can be assigned when something
goes wrong.
4. Builds Trust in Technology
When AI follows ethical guidelines, people are more likely to trust and accept AI
systems in areas like healthcare, education, and transportation.
5. Prevents Harm and Misuse
Ethics guide developers to design AI that benefits society and avoids harmful
applications such as manipulation, surveillance abuse, or unsafe automation.
6. Supports Human Values and Rights
Ethical AI ensures that technology respects human dignity, rights, and social values,
keeping humans in control of important decisions.
These points highlight why ethics are essential for the safe, fair, and beneficial use of
Artificial Intelligence.
Q) What is bias in AI?
Bias in AI means the system makes decisions that are unfair or favor one group over
another. This usually happens because the AI learned from data that was already biased or
incomplete.
Example:
If an AI was trained mostly on photos of light-skinned faces, it may not recognize dark-
skinned faces as well.
This is data bias, and the AI simply repeats what it learned.
Causes of bias:
 Biased or incomplete data
This happens when the data used to train the AI does not represent everyone fairly.
Ex: An AI trained on unfair data may start preferring male applicants, even when
women are equally or more qualified.
 Human choices
Sometimes bias comes from the choices made by developers
 Ex: Labeling Data with Human Error,
When The AI learns:
“Doctor = male”
“Nurse = female”
It may incorrectly label a photo of a female doctor as a nurse.

Q) How AI to be sustainable and help the environment?

How AI Can Be Sustainable and Help the Environ ment
1. Efficient Use of Energy
AI can optimize energy consumption in buildings, factories, and cities by controlling
lighting, heating, and cooling systems, reducing energy waste.
2. Support for Renewable Energy
AI helps predict weather patterns to improve solar and wind energy production,
making renewable energy sources more reliable and efficient.
3. Climate Change Monitoring
AI analyzes satellite images and climate data to track deforestation, melting glaciers,
rising sea levels, and air pollution, helping governments take timely action.
4. Smart Agriculture
AI enables precision farming by monitoring soil, water, and crop health, reducing the
use of water, fertilizers, and pesticides while increasing crop yield.
5. Waste Management and Recycling
AI-powered systems can sort recyclable waste more accurately and optimize garbage
collection routes, reducing pollution and fuel use.
6. Conservation of Wildlife
AI helps track endangered species, detect illegal poaching, and protect natural habitats
through data analysis and image recognition.
7. Sustainable Transportation
AI improves traffic management, supports electric vehicles, and enables efficient route
planning, which lowers fuel consumption and emissions.
8. Designing Sustainable Products
AI helps companies design eco-friendly products by analyzing materials, production
methods, and life cycles to minimize environmental impact.
Q) How can AI models be made robust against attacks?
AI should protect personal information because privacy protects people, and without it,
AI could easily become dangerous, unfair, or un-trusted.
1. Prevent Harm
Personal data can be misused for identity theft, fraud, discrimination, and
manipulation. If AI systems leak or misuse data, people can be directly harmed in very
real ways.
2. Maintain Trust
People will only use AI if they believe it handles their data responsibly. Strong privacy
protection builds public trust, which is essential for AI adoption and long-term success.
3. Respect Human Rights
Privacy is a fundamental human right. AI systems must operate in ways that respect
autonomy, dignity, and personal freedom—otherwise they can become tools of
surveillance or control.
4. Prevent Bias and Discrimination
If private data is improperly collected or used, AI systems may reinforce unfair patterns
related to race, gender, health, or socioeconomic status. Protecting privacy helps
prevent misuse of sensitive data.
5. Follow Laws and Regulations
Governments worldwide have privacy laws that require safeguarding personal
information. AI systems must comply with these laws to avoid legal and financial
consequences.
Unit-4
Long Questions

1. Explain the role of AI in experimentation and multi-disciplinary research.

A. Artificial Intelligence (AI) plays a major role in modern research by automating

processes, analyzing large datasets, generating insights, and helping researchers make better
decisions. It supports scientific discovery across all fields—science, engineering, medicine,
social sciences, and humanities.

AI in Experimentation :-

AI in Experimentation refers to the use of Artificial Intelligence tools to design, automate,

analyze, and optimize scientific experiments. It increases accuracy, reduces time, and
improves decision-making in modern research.

1. Automation: AI systems can automatically control laboratory equipment, collect data, and
manage experimental conditions. This reduces human involvement in repetitive tasks,
minimizes human error, and ensures consistency of results.

2. Data Analysis: Machine Learning (ML) techniques help analyze large experimental
datasets. AI algorithms can detect hidden patterns, classify data, make predictions, and assist
in interpreting complex results that are difficult to analyze manually.

3. Hypothesis Generation: Generative AI models analyze existing research papers, datasets,

and scientific literature to suggest new ideas, research questions, or hypotheses. This speeds
up the early stages of research by helping scientists identify potential directions quickly.

4. Simulation & Modeling: AI simulates experiments virtually before physical trials. These
simulations save time, reduce cost, and allow researchers to test many conditions in a risk-
free environment (e.g., drug behavior, climate modeling, material testing).

5. Optimization: AI helps determine the most efficient experimental setup. Techniques like
reinforcement learning, evolutionary algorithms, and search optimization identify the best
parameters and conditions to achieve the desired outcome with minimal resources.

AI in Multi-disciplinary Research :-

AI in multi-disciplinary research refers to using Artificial Intelligence to connect different

fields, integrate diverse data, and support collaborative problem-solving. AI accelerates
innovation by combining knowledge from science, engineering, healthcare, social science,
environment, and more.
1. Bridging Disciplines: AI connects multiple domains by offering common computational
tools. Examples include bioinformatics, where biology and computer science merge, and
computational social science, which applies AI to study human behavior.

2. Decision Support: AI models help experts from different fields make informed decisions.
They combine data from health, environment, transportation, and smart city systems to
support planning, resource allocation, and policy-making.

3. Knowledge Transfer: AI enables “transfer learning,” where techniques from one

discipline are applied to another. For instance, image recognition models used in computer
vision are adapted for medical imaging to detect tumors or abnormalities.

4. Collaboration: AI simplifies complex scientific information using visualization tools,

summaries, and natural language explanations. This helps researchers from different fields
communicate better and work together effectively.

5. Multi-modal Integration: AI integrates multiple types of data such as text, images, audio,
signals, and numerical measurements. This gives researchers a holistic understanding of
complex problems—useful in fields like healthcare, climate science, and robotics.

2. What is Generative AI? Explain about various Generative AI Tools.

A. Generative AI

It is a type of artificial intelligence that can create new content such as text, images, audio,
video, and code based on learned patterns from data. These models are trained on large
datasets and learn the structure of data. When given a prompt, they generate new content that
resembles human-created output.
Generative AI uses deep learning models like:
 Neural Networks
 Transformers
 Large Language Models (LLMs)

Various Generative AI Tools:

1. ChatGPT (OpenAI)
ChatGPT is a text-based generative AI tool designed for conversational interaction.
Features:
 Generates human-like text responses
 Answers questions and explains concepts
 Writes essays, reports, and code
 Supports learning and research assistance
Applications:
 Education (notes, tutoring)
 Content writing
  Programming help
 Customer support
2. Google Gemini
Gemini is Google’s advanced generative AI model designed for multimodal tasks.
Features:
 Handles text, images, and code
 Strong reasoning and analysis abilities
 Integrated with Google services
Applications:
 Research assistance
 Data analysis
 Code generation
 Image understanding
3. Hugging Face
Hugging Face is an open-source platform providing access to thousands of AI models.
Features:
 Hosts models for NLP, vision, and audio
 Supports model training and deployment
 Open-source and community-driven
Applications:
 Research and experimentation
 Custom AI model development
 NLP tasks like translation and summarization
4. Perplexity AI
Perplexity AI is an AI-powered search and answer engine.
Features:
 Provides concise answers with sources
 Combines search with generative AI
 Real-time information retrieval
Applications:
 Academic research
 Fact-based queries
 Quick information lookup
5. SlidesGPT
SlidesGPT is a generative AI tool used to create presentations.
Features:
 Converts text prompts into slides
 Generates structured content
 Saves time in presentation design
Applications:
 Academic presentations
 Business reports
 Teaching materials
3. Describe the role of Prompt Engineering in AI/ML Interaction.
A. Prompt Engineering plays a crucial role in effective interaction between humans and
Artificial Intelligence (AI) / Machine Learning (ML) models. It refers to the practice of
designing clear, structured, and meaningful prompts to guide AI systems in generating
accurate, relevant, and useful outputs.

1. Acts as a Communication Bridge

Prompt Engineering serves as a bridge between human intent and AI understanding. Since AI
models respond strictly based on input instructions, well-designed prompts help translate
human requirements into a format the model can process effectively.

2. Improves Accuracy and Relevance

Clear prompts reduce ambiguity and help AI generate to Correct answers, Context-aware
responses, Task-specific outputs. For example, specifying format, tone, and depth leads to
more accurate results.

3. Controls Output Style and Format

Prompt Engineering allows users to control Length of response, Writing style

(formal/informal), Output structure (points, tables, code) etc,. This is essential in academic,
business, and technical applications.

4. Enhances AI Performance Without Model Changes

Instead of retraining models, prompt engineering optimizes performance using better inputs,
making AI systems more efficient and cost-effective.

5. Enables Complex Task Execution

With advanced prompting techniques, AI can perform Step-by-step reasoning, Data analysis,
Content generation and Code writing and debugging. This expands the practical usability of
AI tools.

6. Reduces Errors and Bias

Well-crafted prompts help to Minimize hallucinations, Reduce biased outputs, Encourage

factual and neutral responses

7. Supports Personalization and Adaptability

Prompt Engineering enables AI to adapt responses based on User expertise level, Domain
requirements, Specific use cases.
8. Facilitates Research and Experimentation

In AI-assisted research, prompt engineering helps in Literature review, Hypothesis

generation, Data interpretation and Multi-disciplinary analysis. Prompts become part of the
research methodology.

9. Supports Ethical and Responsible AI Use

Prompt engineering can guide AI to Avoid biased responses, Follow ethical constraints and
Respect privacy and fairness etc,.

4. Explore the Emerging Trends and Future Directions in Artificial Intelligence?

A. Artificial Intelligence (AI) is rapidly evolving and transforming almost every sector.
Recent advancements indicate several emerging trends and future directions that will shape
the next generation of intelligent systems.

Emerging Trends and Future Directions in AI

1. Multimodal AI
Multimodal AI systems can process and generate multiple types of data such as text, images,
audio, and video simultaneously.
Examples: Image captioning, Text-to-video generation, Voice-enabled assistants
Future Direction: More human-like interaction and richer AI applications.
2. Generative AI Expansion
Generative AI is advancing beyond text generation to include:
 Image, audio, and video creation
 Code generation
 Synthetic data generation
Future Direction: AI-assisted creativity, design automation, and content production at scale.
3. Autonomous AI Agents
AI agents are being developed to:
 Plan tasks independently
 Make decisions
 Learn from feedback
Examples: AutoGPT, task automation bots.
Future Direction: AI systems that act with minimal human intervention.
4. Explainable AI (XAI)
Explainable AI focuses on making AI decisions transparent and interpretable.
Importance: Builds user trust
Essential for healthcare and finance
Supports ethical compliance
Future Direction: Widespread adoption in regulated industries.
5. Ethical and Responsible AI
Growing emphasis on:
 Bias reduction
  Data privacy
 Fairness and accountability
Future Direction: Global AI regulations and ethical frameworks.
6. AI in Scientific Discovery
AI is accelerating research in:
 Drug discovery
 Climate modeling
 Material science
Future Direction: Faster innovation and scientific breakthroughs.
7. Edge AI
Edge AI enables processing data locally on devices rather than cloud servers.
Benefits:
 Low latency
 Improved privacy
 Reduced bandwidth usage
Future Direction: Smart IoT devices and real-time AI applications.
8. Human–AI Collaboration
Future AI systems will focus on augmenting human capabilities rather than replacing
humans.
Examples:
 AI-assisted decision-making
 Creative collaboration
Future Direction: Balanced and ethical AI–human partnerships.
5. What is prompt engineering and its importance?

A. Prompt Engineering is the systematic process of designing, structuring, and optimizing

prompts (inputs or instructions) given to AI/ML models so that they generate accurate,
relevant, and high-quality outputs.

A prompt may include:

 Instructions
 Context
 Examples
 Constraints
 Output format

Prompt Engineering ensures that the AI clearly understands what to do, how to do it, and in
what format.

Example:

Basic prompt: “Write about AI”

Engineered prompt:
“Write a 300-word explanation of Artificial Intelligence in simple language, include
advantages, applications, and one real-world example.”

Components of Prompt Engineering:

 Instruction – What the AI should do

 Context – Background information
 Input Data – Information to process
 Constraints – Word limit, format, tone
 Output Specification – Table, points, code, paragraph

Types of Prompting Techniques

 Zero-shot Prompting- No examples provided

Example: “Explain Machine Learning”

 Few-shot Prompting- Few examples provided to guide output

Example: Providing sample questions and answers

 Chain-of-Thought Prompting- Encourages step-by-step reasoning

Example: Explain step-by-step how AI makes decisions

 Role-based Prompting- Assigning a role to AI

Example: Act as a data science professor

Importance of Prompt Engineering

1. Improves Accuracy of Output-

Well-structured prompts reduce ambiguity and help AI generate precise and correct
responses.
2. Enhances Human–AI Interaction-
It acts as a communication bridge between humans and AI systems, ensuring effective
understanding.
3. Controls Output Style and Format-
Prompt engineering helps specify:
o Length of response
o Tone (formal/informal)
o Format (points, table, code)
4. Increases Efficiency and Productivity-
Users can get desired results quickly without repeated trials.
5. Enables Complex Task Execution-
Advanced prompts allow AI to perform tasks like:
 o Step-by-step reasoning
o Data analysis
o Content and code generation
6. Reduces Errors and Bias-
Clear prompts guide AI toward factual, neutral, and reliable outputs.
7. No Need to Retrain Models-
Better prompts improve performance without modifying or retraining the AI model.

Short Questions

1. Write about Gemini and ChatGPT AI tools?

A. ChatGPT Tool

ChatGPT is a conversational AI tool developed by OpenAI. It is based on Large Language

Models (LLMs) and the Transformer architecture, designed to understand user inputs and
generate meaningful text responses.

Working of ChatGPT

1. User provides a prompt (question or instruction)

2. ChatGPT analyzes the context and intent
3. It predicts the next sequence of words using probability models
4. Generates a coherent and relevant response

Features

 Natural language understanding

 Text generation and summarization
 Code writing and debugging
 Question answering
 Context-aware conversations

Gemini Tool

Gemini is a powerful generative AI tool developed by Google. It is designed as a

multimodal AI model, capable of processing text, images, audio, and code.

Working of Gemini

Gemini combines advanced neural networks with Google’s AI infrastructure. It can:

 Understand text prompts
 Analyze images
 Generate code and explanations
 Integrate with Google services
Features

 Multimodal understanding
 Advanced reasoning and problem-solving
 Integration with Google tools (Docs, Search, Gmail)
 Supports text, image, and code inputs

2. How does Generative AI create new things like text, images, or music?

A. Generative AI creates new content by learning patterns from large amounts of data and
then using those patterns to produce original outputs such as text, images, and music. It does
not copy existing data directly; instead, it generates new content that follows learned
structures and styles.

1. Training on Large Datasets: Generative AI models are trained on massive datasets like

 Text data (books, articles, websites)

 Image datasets (photos, drawings)
 Audio and music datasets

During training, the model learns: a. Grammar and sentence structure for text.
b. Shapes, colors, and patterns for images.
c. Rhythm, melody, and harmony for music.

2. Use of Deep Learning Models: Generative AI uses deep neural networks, especially
Transformer models for text and code, Diffusion models for images and Neural audio models
for music etc., These models learn complex relationships between data elements.

3. Pattern Learning and Probability Prediction: Instead of memorizing data, Generative AI

Learns probability distributions Predicts the most likely next word, pixel, or sound unit

Example (Text):Given the prompt “Artificial Intelligence is”

The model predicts the most probable next words based on learned patterns.

4. Prompt-Based Generation

When a user provides a prompt, the AI Understands the context, Uses learned patterns and
Generates content step by step. For images, the prompt is converted into a mathematical
representation that guides image creation.

5. Iterative Refinement Process: Many generative models generate output gradually

 Text is generated word by word

 Images are refined from noise to clear visuals
 Music is composed note by note
6. Human Feedback and Fine-Tuning: Models are often fine-tuned using Human feedback
and Reinforcement learning. This helps improve relevance, safety, and usefulness of
generated content.

3. How does AI help in different research fields?

A. Artificial Intelligence (AI) plays an important role across various research domains by
analyzing large datasets, discovering patterns, and supporting decision-making. It improves
accuracy, speed, and efficiency in research activities.

1. Healthcare and Medical Research: AI is widely used in healthcare research for

 Medical image analysis (X-rays, MRI, CT scans)

 Disease diagnosis and prediction
 Drug discovery and clinical trials
 Personalized treatment planning

2. Biological and Life Sciences: In life sciences, AI helps in

 Gene sequencing and bioinformatics

 Protein structure prediction
 Understanding biological processes
 Analyzing complex biological data

3. Engineering and Technology: AI supports engineering research through

 System modeling and simulation

 Fault detection and quality control
 Predictive maintenance
 Robotics and automation

4. Social Sciences: AI is used in social science research to

 Analyze human behavior patterns

 Perform sentiment analysis on social media
 Study economic and social trends
 Support policy analysis

5. Environmental and Climate Research: AI helps researchers in

 Climate modeling and weather forecasting

 Pollution monitoring
 Natural disaster prediction
 Sustainable resource management

6. Education Research: In educational research, AI assists with

  Learning analytics
 Personalized education models
 Student performance prediction
 Curriculum design

7. Business and Economic Research

AI supports business research by:

 Market trend analysis
 Customer behavior prediction
 Financial forecasting
 Risk assessment

4. Write about different deep learning models used in Generative AI?

A. Generative AI uses different types of models to create new content by learning patterns
from large datasets. The major models are:

1. Large Language Models (LLMs)

 Designed to generate human-like text.

 Trained on massive text datasets.
 Capable of writing, summarizing, translating, and answering questions.
Examples: ChatGPT, Gemini
Applications: Content creation, chatbots, research assistance.

2. Generative Adversarial Networks (GANs)

 It Consists of two neural networks:

o Generator: creates new data
o Discriminator: checks authenticity
 Used to generate images, videos, and audio.
Applications: Image synthesis, deepfakes, art generation.

3. Variational Autoencoders (VAEs)

 Encode data into a compressed latent space.

 Generate new, similar data by sampling from this space.
 Commonly used to produce synthetic data.
Applications: Data augmentation, anomaly detection, medical data generation.

4. Diffusion Models

 Generate data by gradually removing noise from random input.

 Known for producing high-quality and realistic images.
 More stable than GANs.
Applications: Image generation, image editing, creative design.
Unit -5
Long Questions

1. Write Applications of Prompt Engineering used to help in Education.

A. Applications of Prompt Engineering in Education

Prompt engineering plays a vital role in improving how Artificial Intelligence (AI) tools
support teaching and learning. By designing clear, structured, and goal-oriented prompts,
educators and students can obtain accurate, relevant, and meaningful outputs from AI
systems.

1. Personalized and Adaptive Learning- Prompt engineering helps AI customize learning

materials based on a student’s like

 Grade level
 Learning speed
 Prior knowledge
 Learning style
This allows students to receive explanations, examples, and exercises tailored specifically to
their needs, improving understanding and retention.

2. Intelligent Tutoring Systems- Well-crafted prompts enable AI to act as a virtual tutor by

 Explaining concepts step-by-step

 Providing hints instead of direct answers
 Asking follow-up questions to test understanding
 Offering multiple solution approaches
This encourages critical thinking and independent problem-solving.

3. Lesson Planning and Curriculum Design- Teachers can use prompt engineering to
generate

 Detailed lesson plans

 Learning objectives aligned with Bloom’s Taxonomy
 Teaching strategies and activities
 Curriculum outlines based on syllabus requirements
This saves time and improves teaching quality.

4. Assessment Creation and Evaluation- AI can generate various assessment tools using
effective prompts

 Multiple-choice questions (MCQs)

 Short and long-answer questions
  Case studies and problem-based questions
 Automated grading criteria and rubrics
Prompt engineering also enables AI to provide instant and constructive feedback.

5. Homework and Assignment Support- Students can use AI to

 Understand assignment requirements

 Get structured outlines
 Improve clarity, grammar, and coherence
 Receive suggestions for improvement
Prompt engineering ensures that AI supports learning without promoting plagiarism.

5. Language Learning and Communication Skills-

6. Prompt engineering allows AI to assist in

 Grammar correction and vocabulary building

 Pronunciation guidance
 Conversation practice
 Translation and paraphrasing
This is especially useful for second-language learners.

7. Exam Preparation and Revision- Using tailored prompts, AI can help generate

 Practice tests and mock exams

 Chapter-wise summaries
 Flashcards and key points
 Time-management strategies
This enhances exam readiness and confidence.

8. Special Education and Inclusive Learning- Prompt engineering can adapt content for

 Students with learning disabilities

 Simplified explanations
 Visual or story-based learning
 Multiple difficulty levels
This ensures equal learning opportunities.

9. Research and Academic Writing Assistance- AI can support higher education students
by

 Summarizing research papers

 Explaining complex theories
 Formatting citations
 Comparing concepts
Effective prompts ensure accuracy and academic relevance.
10. Teacher Training and Professional Development- Educators can use prompt
engineering to

 Learn new teaching methodologies

 Design classroom activities
 Improve assessment strategies
 Explore innovative educational technologies

2. What ways can businesses use prompt engineering to improve customer service?

A. Prompt Engineering helps businesses design effective instructions for AI systems,

enabling them to deliver accurate, consistent, and personalized customer support. By
carefully crafting prompts, companies can significantly enhance customer service quality and
efficiency.
Different ways for Businesses:

1. Intelligent Chatbots and Virtual Assistants- Prompt Engineering enables chatbots to

 Understand customer queries clearly

 Provide accurate and relevant answers
 Handle FAQs, order tracking, and service requests
Example prompt:
“Act as a customer support agent and provide polite, step-by-step solutions.”

2. Personalized Customer Interaction- AI can tailor responses based on

 Customer history
 Preferences
 Language or tone
Well-designed prompts ensure empathetic and personalized communication.

3. Faster Issue Resolution- Prompt Engineering helps AI

 Identify the problem quickly

 Suggest step-by-step troubleshooting
 Reduce response time
This leads to higher customer satisfaction.

4. 24/7 Customer Support- AI-powered customer service systems can operate continuously.
Proper prompts help maintain

 Consistent tone
 Accurate information
 Professional behavior

5. Multilingual Customer Support- Prompt Engineering allows AI to

  Respond in multiple languages
 Translate customer queries accurately
This helps businesses serve global customers.

6. Automated Ticket Handling

 Categorize customer complaints

 Assign priority levels
 Route tickets to the correct department
Clear prompts ensure correct classification.

7. Feedback Analysis- AI analyzes customer feedback and reviews to

 Identify common issues

 Understand customer sentiment
 Suggest service improvements

3. How does AI help in Content Creation and creative writing?

A. Artificial Intelligence (AI) has become a powerful tool for content creation and creative
writing. It assists writers, educators, marketers, and students by improving productivity,
creativity, and content quality while reducing time and effort.

AI in Content Creation and creative writing:

1. Idea Generation and Brainstorming

AI helps writers by:

 Generating story ideas, themes, and plots
 Suggesting titles, hooks, and headlines
 Overcoming writer’s block
This supports creativity and inspires new perspectives.

2. Drafting and Writing Assistance

 Generate initial drafts of articles, stories, poems, and scripts

 Expand short ideas into full content
 Rewrite content in different tones or styles
This helps writers focus more on refining creativity than starting from scratch.

3. Creative Storytelling Support

In creative writing, AI assists with:

 Character development
 World-building and setting descriptions
  Plot progression and dialogue creation
Writers can experiment with multiple story directions easily.

4. Editing and Refinement

AI improves content quality by:

 Correcting grammar, spelling, and punctuation
 Enhancing sentence flow and clarity
 Suggesting stylistic improvements
This ensures polished and professional writing.

5. Style and Tone Adaptation

AI can adjust writing to:

 Formal or informal styles
 Academic, narrative, or conversational tones
 Specific audiences (children, students, professionals)
This flexibility is valuable in both content creation and creative writing.

6. Language Learning and Vocabulary Enhancement

AI helps writers by:

 Suggesting synonyms and richer vocabulary
 Improving sentence variety
 Supporting multilingual writing and translation
This enhances expression and creativity.

7. Content Personalization

AI tailors content based on:

 Reader preferences
 Age group or reading level
 Cultural context
Personalized content improves reader engagement.

8. Multimedia Content Support

AI contributes to creative projects by:

 Writing scripts for videos and podcasts
 Generating captions, subtitles, and descriptions
 Assisting in image and visual content generation
This supports storytelling across multiple formats.

9. Productivity and Time Efficiency

AI:
  Speeds up content production
 Reduces repetitive tasks
 Enables quick revisions and multiple versions
This allows writers to focus on originality and creativity.

10. Learning and Skill Development

AI acts as a learning partner by:

 Providing writing tips and feedback
 Explaining literary techniques
 Helping students practice creative writing

4. What is SlidesGPT and how does it use prompts to make presentations?

A. SlidesGPT:

SlidesGPT is an AI-based presentation tool that automatically generates PowerPoint or

Google Slides presentations from user-provided text prompts. It uses Generative AI to
convert ideas or topics into well-structured slides with headings, bullet points, and
summaries. SlidesGPT relies heavily on prompt engineering to understand the user’s
requirements and generate effective slides.
Representation of SlidesGPT Working

Prompts to make Presentations:

1. Understanding the Topic
The user provides a prompt such as:
“Create a 10-slide presentation on Artificial Intelligence for undergraduate students.”
SlidesGPT identifies the topic, audience, and depth required.
2. Structuring the Presentation
Based on the prompt, SlidesGPT:
 Divides content into slide titles
 Organizes points logically
 Creates introduction and conclusion slides
3. Content Generation
The AI generates:
 Clear bullet points
 Key definitions and explanations
 Examples where needed
Prompts may specify word limits or slide count.
4. Customization Using Prompts
Users can control:
 Number of slides
 Level of detail
 Tone (formal/educational/business)
Example:
“Create concise slides with 4 bullet points per slide.”
5. Time and Productivity Benefits
By using prompts, SlidesGPT:
 Saves time
 Reduces manual effort
 Ensures consistent formatting
5. Discuss about Designing Thumbnails and Channel Branding with AI?

A. Designing Thumbnails and Channel Branding with AI

In 2025, AI has shifted from being a simple image generator to a comprehensive
"performance engine" for YouTube creators. 1 Designing thumbnails and branding with AI is
no longer just about making something look "cool"—it’s about using data to maximize your
Click-Through Rate (CTR).2

a. AI-Powered Thumbnail Design

Traditional design tools like Photoshop are being replaced or augmented by AI tools that
understand YouTube-specific psychology.

Top AI Thumbnail Tools (2025)

  1of10: Currently considered a top-tier tool because it doesn't just design; it analyzes
your channel's performance data, facial expressions, and lighting styles to predict
what will get the most clicks.
 Canva (Magic Media & Dream Lab): The most accessible option. You can use
"Magic Edit" to swap backgrounds or objects and "Magic Eraser" to clean up
distractions instantly.4
 Ideogram 2.0: Highly recommended if your thumbnail needs accurate text. While
many AIs struggle with spelling, Ideogram is specifically optimized for graphic design
and typography.
 Face Swap Tools (e.g., Replicate/Remix): Many creators use AI to place their own
faces (with exaggerated expressions) into AI-generated epic scenes, maintaining a
personal connection while using high-budget visuals.

The AI Workflow for Thumbnails

1. Ideation: Feed your video transcript into ChatGPT or Claude and ask for "5 high-
contrast thumbnail concepts based on the 3-second rule."5
2. Generation: Use Midjourney (using the --ar 16:9 parameter) or [Link] to
create the background or main subject.6
3. Refinement: Use Adobe Express or Canva to add bold text.7 Ensure the text is
"mobile-safe" (not covered by the time stamp in the bottom right corner).

b. Channel Branding with AI

Branding is about consistency.8 AI helps you maintain a "visual signature" across your
banner, profile picture, and video watermarks.

Visual Identity

 Logo & Icon Generation: Tools like Looka or Microsoft Designer (powered by
DALL-E 3) can generate vector-style logos based on your niche.9
 Consistent Banners: Use Venngage’s AI Banner Generator to create layouts that
work for both mobile and desktop.10 You can prompt for specific "safe zones" to
ensure your text isn't cut off on phones.
 Color Palettes: AI tools like Khroma learn which colors you like and generate
palettes that evoke specific emotions (e.g., neon for gaming, earthy tones for
gardening).

Maintaining Consistency with LoRAs

For advanced creators, 2025 has seen the rise of LoRAs (Low-Rank Adaptation).11 In tools
like [Link], you can "train" the AI on 10–15 images of your own brand style. Once
trained, the AI will always generate images that match your specific color grade, lighting, and
"vibe," ensuring every video looks like it belongs to the same brand.
c. Best Practices & Psychological Hooks

 The 3-Second Rule: A viewer should understand the "story" of your video within 3
seconds of seeing the thumbnail. 12 Use AI to simplify busy backgrounds.
 Emotion Enhancement: Use AI to subtly enhance facial expressions—brightening
eyes or slightly widening a "surprised" mouth can subconsciously increase curiosity. 13
 A/B Testing: Tools like Test My Thumbnails allow you to run AI simulations to see
which version of a thumbnail is likely to perform better before you even upload.

Short Questions

1. Define Prompt Engineering in business applications.

A. In business, Prompt Engineering is the strategic practice of designing, refining, and

optimizing inputs (prompts) to guide Generative AI models to produce high-quality, brand-
aligned, and actionable results.

While a casual user might ask an AI to "write an email," a business prompt engineer treats the
AI as a high-speed professional assistant that requires a precise "Statement of Work" to be
effective.

The Core Pillars of Business Prompting

For an AI output to be "business-ready," it must move beyond generic text. Prompt

engineering in a corporate context relies on four key pillars:

 Contextual Anchoring: Providing the AI with background information (e.g., "You are
a senior supply chain analyst for a Fortune 500 retail company").
 Objective Clarity: Defining exactly what the success metric is (e.g., "Summarize this
50-page contract into 5 bullet points focused only on liability and termination
clauses").
 Persona & Tone Management: Ensuring the AI speaks in the company’s specific
"brand voice"—whether that is authoritative and legalistic or friendly and customer-
centric.
 Constraint Setting: Establishing boundaries to prevent "hallucinations" (e.g., "Only
use the provided PDF as a source; if the answer isn't there, state that the information is
unavailable").

2. How can AI generate video scripts using prompt engineering?

A. AI can generate complete video scripts by using prompt engineering, which means
giving clear, structured instructions to guide the AI in creating relevant, engaging, and well-
organized content.

Various Aspects in AI Video scripts:

1. Role of Prompt Engineering- Prompt engineering helps the AI understand

 Topic of the video

 Target audience
 Length and format
 Tone and style like educational, promotional, storytelling

2. Step-by-Step Process of AI Video Script Generation

Step 1: Prompt Input-

Example Prompt:
“Write a 3-minute educational YouTube video script explaining Prompt Engineering for
undergraduate students, including an introduction, examples, and conclusion.”

Step 2: AI Understanding & Processing- The AI Analyzes keywords, Identifies structure

(intro, body, outro) and Applies language patterns from training data

Step 3: Script Structuring- AI creates Hook/opening lines, Scene-wise narration, Transitions

and Call-to-action

Step 4: Refinement Through Prompts- Users can refine outputs by adding Word limits,
Visual cues and Speaker roles

Refinement Prompt:
Rewrite the script with a friendly tone and include on-screen text suggestions.

3. Practical Applications: AI-generated video scripts are used for

 YouTube educational videos

 Marketing and advertisements
 Corporate training videos
 Social media reels

4. Tools Used

 ChatGPT – Scriptwriting and narration

 Gemini – Context-rich and factual scripts
 Jasper AI – Marketing scripts
 Synthesia – Script + AI avatar videos
3. How does AI help in Creative designing?

AI helps creative design by automating repetitive tasks like resizing, background removal,
acting as a brainstorming partner for new ideas (mood boards, logo concepts), speeding up
workflows, enabling hyper-personalization, and ensuring brand consistency, freeing
designers to focus on complex, strategic, and truly innovative aspects of their work. It
 generates variations, suggests layouts, and helps overcome creative blocks, making the entire
process faster, more efficient, and more imaginative.

How AI enhances the design process:

 Automation & Efficiency: Handles tedious tasks like image resizing, background removal,
color correction, and template adaptation, significantly speeding up production.
 Inspiration & Ideation: Acts as a powerful brainstorming tool, generating numerous logo
concepts, color palettes, patterns, and mood boards from simple prompts, breaking through
creative blocks.
 Rapid Prototyping & Iteration: Quickly creates multiple design iterations, allowing
designers to explore many directions and concepts efficiently.
 Personalization at Scale: Analyzes user data to help create highly customized designs and
experiences for specific audiences, notes LinkedIn.
 Consistency & Quality: Ensures designs adhere to brand guidelines, maintaining high
quality and consistency across large-scale projects.
 Accessibility: Automates checks for color contrast and readability, making designs more
accessible faster, according to Adobe.
 New Product Creation: Can generate novel product designs, clothing, or furniture concepts,
accelerating innovation, say [Link] and Amadine.
4. Explain Applications of prompt engineering?
Prompt engineering's applications span numerous fields, enabling AI to generate creative
content, power intelligent chatbots, assist developers with code, personalize education,
analyze data for insights, and improve customer support by providing context-aware,
accurate, and efficient responses across industries like marketing, healthcare, and finance.
Essentially, it's about crafting precise instructions (prompts) to guide Large Language Models
(LLMs) to perform specific tasks better, from simple text generation to complex problem-
solving.

Key Applications
 Content Creation: Generating articles, social media posts, ad copy, product descriptions,
and creative writing.
 Customer Support: Building smarter chatbots and virtual assistants that understand intent
and provide accurate, on-brand support.
 Software Development: Assisting with code generation, debugging, explaining complex
code, and generating documentation.
 Education: Creating personalized learning materials, quizzes, lesson plans, and simplifying
complex topics for students.
 Data Analysis & Research: Extracting insights from datasets, summarizing research papers,
and identifying trends.
 Language Translation: Achieving more contextually relevant translations by specifying
dialect, tone, and script.
 Healthcare: Aiding in diagnosis by processing symptoms and researching treatment options
from medical literature.
 Marketing & E-commerce: Creating targeted marketing campaigns, personalized product
recommendations, and analyzing customer feedback.