AI Logic School

Empowering Students with AI & Computational Thinking

Class 10 | Chapter 1 AI Reflection, Project Cycle & Ethics

Class 9 AI — Chapter 1: AI Reflection, Project Cycle & Ethics | AI Logic School
📘 Class 9 · Subject Code 417 · CBSE 2024-25

Chapter 1
AI Reflection, Project Cycle
& Ethics

Everything starts here! Understand what AI really is, explore its 3 domains, master the 6-step AI Project Cycle, and learn why AI Ethics is the most important topic in all of technology today.

⏱️ 30 Theory Hours 🎯 Part B · Unit 1 📝 10 Marks 📚 Sub-Code 417
🧠 Human Intelligence Learns · Feels · Creates + 💻 Machine Processing Fast · Accurate · Tireless = 🤖 Artificial Intelligence Learns · Decides · Improves
🤖

What is Artificial Intelligence?

Understanding the basics — definition, types and real-world examples

📖 Definition

Artificial Intelligence (AI) is the simulation of human intelligence by machines — especially computers. It allows machines to perform tasks that normally require human intelligence such as learning, reasoning, problem-solving, perception and language understanding.

The word "Artificial" means made by humans, and "Intelligence" means the ability to learn and solve problems. Together — AI is a computer system that humans have taught to be smart!

✅ AI CAN do these things

  • Recognise faces in photos
  • Translate between languages
  • Suggest videos you'll like
  • Detect diseases in X-rays
  • Play chess better than any human
  • Generate images from text
  • Drive cars (partially)

❌ AI CANNOT do these things

  • Feel emotions (love, sadness, joy)
  • Be truly creative from nothing
  • Have common sense naturally
  • Learn without being trained
  • Understand sarcasm perfectly
  • Make moral judgements reliably
  • Replace all human jobs (yet!)
🌟
The term "Artificial Intelligence" was first coined by John McCarthy in 1956 at the Dartmouth Conference — making AI about 70 years old! But the real AI revolution happened in the 2010s when computers became fast enough and data became plentiful enough for AI to actually work well.

🧠

AI vs Machine Learning vs Deep Learning

These three are nested — each one is a subset of the previous!

The Nesting Relationship — like Russian Dolls!
Artificial Intelligence (AI) Broad concept: machines simulating human intelligence Machine Learning (ML) Machines that learn from data Deep Learning Neural networks 🤖 📊 🧬

📊 AI vs ML vs DL — Simple Explanation

  • 🤖 AI (Artificial Intelligence) — The big umbrella. Any machine that mimics human intelligence. Includes rule-based systems AND learning systems.
  • 📈 ML (Machine Learning) — A subset of AI. Instead of programming rules, machines learn from data. The more data, the smarter it gets.
  • 🧠 DL (Deep Learning) — A subset of ML. Uses neural networks (layers of calculations inspired by the human brain). Powers face recognition, ChatGPT, image generation.
🍕
Pizza Analogy: AI = all food. ML = pizza (a type of food). DL = wood-fired Neapolitan pizza (a specific type of pizza). Every deep learning system is ML, every ML system is AI — but not every AI uses ML!

🌐

The 3 Domains of AI

As per CBSE Curriculum — Data Science, Computer Vision & NLP

CBSE introduces AI through three core domains. Understanding these helps you identify which type of AI is being used in any application you encounter in daily life!

📊

Data Science

Finding patterns in numbers and information to make predictions and decisions.

🎮 Try: Rock Paper Scissors AI
👁️

Computer Vision

Teaching computers to see and understand images, videos and the visual world.

🎨 Try: Quick Draw by Google
💬

Natural Language Processing

Teaching computers to understand and generate human language — spoken or written.

🎯 Try: Semantris by Google
Where Do You See Each Domain in Real Life?
📊 Data Science 🎵Spotify suggestions 🛒Amazon recommendations 🌦️Weather prediction 📈Stock market AI 🏏Cricket match prediction 👁️ Computer Vision 😊Face unlock (Phone) 🚗Self-driving cars 🏥X-ray disease detection 📸Photo filter / beauty mode 🔍Google Lens search 💬 NLP 🌐Google Translate 🎙️Siri / Google Assistant 📧Email spam filter 🤖ChatGPT / Gemini 🔎Google Search

🔄

The AI Project Cycle

6 steps every AI project follows — from problem to real-world solution!

🔄 Why a Cycle?

AI development is not a straight line — it's a cycle! After evaluating your model, you might go back and collect better data, or re-scope the problem. The cycle is iterative — meaning you keep improving until you're satisfied with the results.

🔄 Click each step to explore — use this on your blog post!
AI Project Cycle 🔭1.Scope 📦2.Data 🔍3.Explore 🤖4.Model 📊5.Evaluate 🚀6.Deploy

👆 Click any step in the diagram above to learn about it!

The AI Project Cycle has 6 stages. Each stage is important — skipping one leads to a poor AI model. Click each numbered circle above to understand that step in detail with an example.


🔭

Step 1: Problem Scoping — The 4W Canvas

The most important step — a clear problem = a great AI solution!

🗺️ What is Problem Scoping?

Problem Scoping means clearly defining exactly what problem you want to solve before writing a single line of code or collecting any data. A vague problem leads to a useless AI model!

CBSE uses the 4W Problem Canvas — four questions that help you define your problem completely:

The 4W Problem Canvas — Fill This For Every AI Project!
WHAT? What is the problem? Example: "Farmers cannot detect crop diseases early enough to prevent loss." WHO? Who is affected? Stakeholders: Farmers, agriculture officials, food supply chain, consumers. WHERE? Where does it occur? Rural farming communities in India. Limited access to agricultural experts. WHY? Why solve this? SDG Goal 2: Zero Hunger. Crop diseases cost farmers ₹50,000+ per season.
💡
CBSE Tip: After the 4W Canvas, you also need to identify Stakeholders (everyone affected by the problem) and think about Ethical Issues your AI solution might create. For example, a crop disease AI that only works in English excludes non-English speaking farmers — that's an ethical problem to consider!

📦

Steps 2 & 3: Data Acquisition & Exploration

Collecting the right data and understanding what it tells you

📦 What is Data Acquisition?

After scoping the problem, you need to collect the data that your AI will learn from. This is called Data Acquisition. Getting the right data is the most critical step — bad data = bad AI.

  • What features do you need? (e.g., for crop disease: leaf colour, texture, size)
  • Where will you get the data? (surveys, sensors, online datasets, cameras)
  • How much data do you need? (more = better, usually thousands of examples)
  • How often must data be updated? (daily weather, monthly sales?)
  • What if you don't have enough data? (can you generate synthetic data?)

🔍 Data Exploration & Visualisation

Once you have data, you must explore and understand it before training any model. This is done through Data Visualisation — turning numbers into charts!

  • 📊 Bar Charts — Compare values across categories
  • 📈 Line Charts — Show trends over time
  • 🥧 Pie Charts — Show proportions of a whole
  • 🔵 Scatter Plots — Show relationship between two variables
  • 📦 Box Plots — Show distribution and outliers
Rule-Based vs Learning-Based AI — Two Types of Modelling
📋 Rule-Based AI → Humans write all the rules → IF temperature > 38°C THEN fever → Predictable & explainable → Cannot handle new situations Example: Early spam filters, thermostats 🧠 Learning-Based AI → Machine learns from examples → Trained on thousands of spam emails → Adapts to new patterns → Can handle unknown situations Example: ChatGPT, face recognition, Siri

📊

Step 5: Evaluation — How Good is Your AI?

True Positive, False Positive, True Negative, False Negative

📊 Evaluation Metrics

After building your model, you need to test it. The CBSE curriculum introduces these important terms:

  • True Positive (TP) — AI says YES, and the answer is actually YES. Correct!
  • True Negative (TN) — AI says NO, and the answer is actually NO. Correct!
  • False Positive (FP) — AI says YES, but the answer is actually NO. False Alarm!
  • ⚠️ False Negative (FN) — AI says NO, but the answer is actually YES. Dangerous Miss!
Easy Way to Remember — Disease Detection Example
Reality: HAS Disease Reality: NO Disease AI says: HAS Disease AI says: NO Disease True Positive (TP) Correctly detected disease ⚠️ False Positive (FP) False alarm — healthy patient 🚨 False Negative (FN) Dangerous! Missed real disease True Negative (TN) Correctly cleared healthy patient

⚖️

AI Ethics — Fairness, Privacy & Responsibility

The most important aspect of AI — making sure it helps everyone equally

⚖️ What is AI Ethics?

AI Ethics is the set of moral principles and rules that ensure AI systems are developed and used in ways that are fair, safe, transparent and beneficial to all people — not just some.

As AI becomes more powerful and more widely used, the ethical questions become more important. CBSE includes AI ethics because you — today's students — will build tomorrow's AI systems. Understanding ethics now is critical!

⚖️

Fairness — Treat all people equally

AI must not discriminate based on gender, religion, language, caste or geography. Example: A job-screening AI must evaluate all candidates equally regardless of their name or hometown.

🔍

Transparency — Explain decisions

AI systems must be able to explain WHY they made a decision. If an AI rejects a loan application, it must give a reason. Black-box AI with no explanations is unethical.

🔒

Privacy — Protect personal data

AI must not collect or misuse personal data without consent. Your photos, location, health data, and messages are private. AI applications must ask permission and protect this data.

🛡️

Safety — Do no harm

AI used in critical systems like hospitals, cars or planes must be tested exhaustively before deployment. A bug in a self-driving car's AI could be fatal. Safety must come first.

👤

Human Control — Humans stay in charge

AI must always have a human override option. No AI system should be able to make irreversible decisions without human approval. Humans must remain responsible for AI outputs.


😮

AI Bias — When AI is Unfair

Understanding what bias is, why it happens, and how to prevent it

😮 What is AI Bias?

AI Bias occurs when an AI system produces unfair or prejudiced results because of biased training data or a flawed algorithm. Simply put — AI learns from data. If the data is biased, the AI will be biased too.

This is one of the most serious problems in AI today because biased AI can affect millions of people — in hiring, lending, medical care, and law enforcement.

How AI Bias Happens — Step by Step
📁 Biased Data Only city students, mostly boys, English only 🤖 AI Training AI learns the biased patterns as "normal" 😢 Biased Output Rural students rejected, girls scored lower unfairly The Fix Use diverse, representative data!

Click each card to reveal if the AI decision is ethical or biased:

🏥

Medical Diagnosis AI

An AI trained on X-rays from patients across India — rural & urban, all ages and genders — to detect tuberculosis.

👮

Crime Prediction AI

An AI that predicts who might commit crimes based primarily on their residential neighbourhood and past arrest data.

📚

Personalised Learning AI

An AI that adjusts the difficulty and style of questions based on each individual student's learning pace and progress.

💼

Job Recruitment AI

A hiring AI trained on 10 years of a tech company's past hires — which were 85% male — to shortlist new candidates.

🌍 AI Access — Not Everyone Benefits Equally

Even when AI is unbiased, there is the problem of AI Access — not everyone has equal access to AI technology or its benefits:

  • 🌐 AI requires the internet — 40% of India's rural population lacks reliable connectivity
  • 📱 AI apps are mostly in English — India has 22 official languages
  • 💰 Premium AI tools are expensive — not affordable for everyone
  • 📚 AI literacy requires education — digital divide affects AI benefits
🎭
CBSE Balloon Debate Activity: Split into teams. One team argues AI is beneficial for society. The other argues it is harmful. Both must use real examples. Also explore the Moral Machine at moralmachine.net — it shows the ethical choices self-driving cars must make!

🔑

Key Terms & Definitions

Must-know vocabulary for your exam — every term explained clearly

Artificial Intelligence (AI)
Simulation of human intelligence by machines to think, learn and solve problems.
Machine Learning (ML)
A subset of AI where machines learn patterns from data automatically without explicit programming.
Deep Learning (DL)
A subset of ML using multi-layered neural networks inspired by the structure of the human brain.
Data Science
The domain of AI that deals with collecting, processing and finding meaningful patterns in data.
Computer Vision (CV)
AI domain that enables computers to interpret and understand images, videos and visual content.
Natural Language Processing (NLP)
AI domain that enables computers to understand, interpret and generate human language.
AI Project Cycle
6-step process: Problem Scoping → Data Acquisition → Exploration → Modelling → Evaluation → Deployment.
4W Problem Canvas
A framework to define an AI problem using: What (problem), Who (affected), Where (location), Why (importance).
Rule-Based AI
AI that follows human-written IF-THEN rules. Predictable but cannot learn from new situations.
Learning-Based AI
AI that trains on data and discovers patterns on its own. Forms the basis of modern machine learning.
True Positive (TP)
AI correctly predicts a positive result. e.g., AI correctly identifies a patient with a disease.
False Negative (FN)
AI predicts negative when reality is positive. Most dangerous error — e.g., missing a real disease.
AI Bias
Unfair outputs from an AI system due to prejudices or imbalances in training data or algorithm design.
AI Ethics
Moral principles ensuring AI is fair, transparent, safe, privacy-respecting and under human control.
AI Access
The issue of unequal access to AI technology across different communities, languages and economic levels.
Stakeholder
Any person or group affected by an AI project — users, operators, communities, decision-makers.

📌 Chapter 1 — Complete Revision Checklist

  • AI = Artificial Intelligence. Simulation of human thinking by machines.
  • AI ⊃ ML ⊃ DL — Deep Learning is the smallest/most specific subset
  • 3 Domains: Data Science (patterns in data), Computer Vision (images), NLP (language)
  • AI Project Cycle: Scope → Data → Explore → Model → Evaluate → Deploy
  • 4W Canvas: What + Who + Where + Why
  • Two model types: Rule-Based (human writes rules) vs Learning-Based (learns from data)
  • Evaluation terms: TP, TN, FP, FN — know the confusion matrix!
  • AI Ethics: Fairness · Transparency · Privacy · Safety · Human Control
  • AI Bias = unfair AI due to biased training data
  • AI Access = unequal access to AI benefits across society
  • Real example: Amazon's hiring AI in 2018 was shut down for gender bias
  • CBSE activity: Moral Machine at moralmachine.net

✍️ Exam Tips — Chapter 1

1 MARK Expect "Define AI" or "What is ML?" questions. Always write the full form + one-line definition + one example. Never just give the definition alone — always add an example.
2 MARKS Common question: "Differentiate between Rule-Based and Learning-Based AI." Draw a table — 2 points each. Also prepare "What is the 4W Problem Canvas? Give an example."
3 MARKS "Explain any 3 steps of the AI Project Cycle." Mention the step name + what it does + one real example for each step. Use the crop disease AI as your running example throughout.
5 MARKS "Explain AI Ethics with examples" or "What is AI Bias? Give 2 real-world examples." For 5-mark answers: define the term, give 4–5 points, include a real example (Amazon hiring AI), and write a concluding line.
DIAGRAM Practice drawing the AI Project Cycle as a circular diagram with arrows. Also practice the Confusion Matrix (2×2 table with TP, TN, FP, FN). These are high-probability diagram questions!

Comments

Chat on WhatsApp