Class 9

Introduction to Artificial Intelligence

_________________________________________________

Intelligence
Definition
● Intelligence is the ability to:

○ Learn from experience

○ Recognize problems

○ Solve problems

● Sternberg's Definition:
Intelligence is the capacity to learn from experience using metacognitive processes and
the ability to adapt to the environment.

Examples
● Students are considered intelligent if they:

○ Understand subject content

○ Score above-average grades

Artificial Intelligence (AI)

Meaning
● Artificial = Man-made

● Intelligence = Ability to think and make decisions

● AI is a branch of computer science that simulates human intelligence in machines.

Purpose of AI
● Make machines:

○ Think like humans

○ Learn and reason

○ Perceive surroundings

○ Make decisions

Components of AI
● Perception: Acquiring and interpreting data

● Reasoning: Drawing conclusions using logic

● Learning: Improving through experience and self-correction

Brief History of AI
Year Event

1950 Alan Turing introduces the Turing Test

1956 John McCarthy coins the term Artificial Intelligence at the Dartmouth
Conference

1974– First AI Winter – decline in funding and interest

1980

1980s Rise of Expert Systems, growth in research and learning algorithms

2000 AI enters homes (e.g., Roomba), Deep Learning emerges

2010s Introduction of Chatbots, Google Now, IBM Watson

2014 Chatbot Eugene Goostman passes the Turing Test

2020s Popularity of Big Data, Machine Learning, Data Science; widespread

education in AI
Turing Test
Definition
● Introduced by Alan Turing in 1950

● Determines whether a machine can exhibit human-like intelligence

● Originally called the “Imitation Game”

Process
● Involves 3 terminals:

1. Interrogator (human)

2. Human respondent

3. Computer respondent

● The interrogator asks questions and must determine which is the human.

● If the computer's responses are indistinguishable from human responses, it is

considered intelligent.

Example Questions
● What is the difference between a basketball and a football?

● Which season gets a lot of snowfall?

● What is the difference between a pencil and a pen?

Uses of Turing Test

● Helps assess how advanced AI has become

● Ensures unbiased judgments in controlled environments

● CAPTCHA: A practical Turing Test used on websites to differentiate humans from bots
Importance of Turing Test
● Key tool to evaluate AI development

● Helps identify whether machines can mimic human reasoning and conversation

● Indicates progress in Artificial Intelligence research

Applications and Benefits of AI

1. AI in Medical/Health Care
● Diagnosis: IBM Watson used for medical diagnosis.

● Eye Disease Detection: Google’s DeepMind analyzes retinal scans for early symptoms.

● Lab Work: AI helps in analyzing tissue samples with high accuracy.

2. AI in Education
● Automated Grading: Provides real-time feedback to students and teachers.

● Fair Assessment: AI is used to ensure accuracy and fairness in grading.

○ Example: Telangana State Board uses AI for exam results.

○ Pearson: Uses AI-based systems for English Language Testing (ELT).

3. AI in Business
● Sales & Forecasting: Helps in advertising, target marketing, and decision-making.

● Chatbots: Used for customer support and FAQs.

● Efficiency: Enhances business processes and profitability.

4. AI in Agriculture
● Smart Farming: AI helps maximize yield and efficient resource use.

● Examples:

○ See and Spray robot by Blue River Technology: Monitors and sprays pesticides
accurately.

○ Plantix app: Detects plant diseases using AI image recognition.

5. AI in Banking & Finance

● Stock Trading: Analyzes large data sets to predict market trends.

● Example: Nomura uses AI to learn from experienced traders.

● Fraud Detection: AI used in banking to identify anomalies and credit card fraud.

● Customer Support: Chatbots assist customers 24/7.

6. AI in Gaming Industry
● Game Intelligence: AI creates intelligent game characters.

● Example: AlphaGo defeated world #1 Go player, Fan Hui.

7. AI in Space Exploration
● Data Analysis: AI helped discover an 8-planet solar system using Kepler data.

● Mars Missions: NASA uses AEGIS, an AI robot for autonomous camera targeting on
Mars.
8. AI in Transport
● Airline Operations:

○ Ticket pricing based on real-time data.

○ Manages food services, boarding, staff schedules, and baggage handling.

● Customer Experience: Enhanced through predictive services.

9. AI in Social Media
● Personalized Feeds: ML algorithms design feeds based on user interests.

● Face Recognition: Used by Facebook to tag people.

● Content Moderation:

○ Twitter uses AI to detect hate speech and extremist content.

○ Uses Machine Learning, Deep Learning, and NLP.

10. AI in Artificial Creativity

● Music Generation:

○ MuseNet: Creates compositions with 10 instruments in different styles.

● Automated Writing:

○ Wordsmith: Generates narratives from data using NLP.

11. AI in Industry
● Efficiency & Accuracy: Used in manufacturing and logistics.
 ● COVID-19 Response: Startups developed AI-based innovations for healthcare and
logistics.

● Hazardous Job Replacement: AI is used in jobs risky for humans (medicine, defense,
environment, etc.)

12. AI in Global Defense

● Security Robots:

○ AnBot: Developed by China’s National Defense University for security patrols

and threat detection.

13. AI in Entertainment
● News Anchors:

○ AI Anchor Lisa reads headlines, allowing human anchors to focus on more

engaging tasks.

● Film Production: Enhances graphics and special effects.

● Chess & Games: Challenges and improves professional players’ skills.

14. AI in Security
● Surveillance Systems: AI-based CCTV monitoring to detect criminal activity.

● Social Media Intelligence: Predicts threats using public data and online behavior.

Disadvantages of AI

1. Expensive Technology
 ● AI systems require:

○ Constant hardware and software upgrades.

○ Complex engineering and setup.

● High development and maintenance costs.

2. Leads to Unemployment
● Automation of repetitive and manual jobs in:

○ Manufacturing

○ Agriculture

○ Food service

○ Retail

○ Transportation

○ Hospitality

● Jobs may be replaced, not just reduced.

3. Lacks Emotions and Creativity

● AI cannot replicate human:

○ Emotions

○ Intuition

○ True creativity

● AI's creative output is limited to its programmed scope.

4. Does Not Learn from Experience (Like Humans)

 ● AI is programmed, not intuitive.

● Cannot handle unprogrammed or unexpected situations unless reprogrammed.

Ethical Issues in AI

1. Fairness
● Human bias may be embedded in AI algorithms.

● Example: Biased data leads to discriminatory outputs.

2. Transparency
● Users need to understand how AI systems work:

○ How they are trained

○ How decisions are made

3. Interpretability & Explainability

● Interpretability: Can we predict what the AI will do?

● Explainability: Can we understand how it arrived at its result?

● Both are critical in:

○ Healthcare

○ Law

○ Security
4. Prejudice and Bias
● AI may inherit or amplify social biases:

○ Gender, race, language, wealth

● Examples:

○ "Doctor" often shows male images, "nurse" female.

○ Virtual assistants (Siri, Alexa) default to female voices.

5. Unemployment
● AI replaces many routine and hard-skill jobs.

● Soft-skill jobs still in demand (e.g., teaching, therapy).

● Technological shift leads to new types of jobs, but transitional job losses are a concern.

6. Inequality
● Owners of AI systems earn more → wealth concentration.

● AI advances may widen the gap between:

○ Developed and developing nations

○ Tech-rich and tech-poor individuals

7. Gender Gap
● Only 22% of AI professionals are female.

● AI systems may reinforce gender roles:

 ○ Female-voiced assistants

○ Gendered search results

8. Humanity and Social Confusion

● AI chatbots and avatars can imitate human interaction.

● Raises the question: Are we talking to a person or a machine?

● Can be helpful or harmful, depending on use.

9. Accountability
● AI decisions based on training data—not always adaptable to real-world
unpredictability.

● Can be misused by malicious actors.

● Issues in areas like:

○ Security

○ Surveillance

○ Law enforcement

10. Unforeseen Consequences

● AI may misinterpret objectives:

○ e.g., To eliminate a virus, it might try to eliminate the host (humans).

● AI lacks common sense and ethical reasoning.

11. Threat to Human Dignity
● Some roles should not be replaced by AI due to their emotional and ethical nature:

○ Customer service reps

○ Elderly caregivers

○ Therapists

○ Soldiers

○ Police officers

○ Judges

● Such roles require empathy, trust, and human connection.

AI Bias – Notes

What is AI Bias?
● AI Bias refers to irregular or unfair outcomes from AI algorithms.

● It often stems from:

○ Prejudiced assumptions during development

○ Biased training data

● AI systems process data and draw conclusions but lack understanding of why a
conclusion is correct or ethical.

Main Cause
The difference between facts and their interpretation in data.

Types of AI Bias
 1. Perceptive Biases

○ Biases based on group association or stereotypes.

○ Around 180 human cognitive biases exist.

○ These biases can unintentionally influence machine learning algorithms

through:

■ Designers’ assumptions

■ Data collection processes

2. Incomplete Data Biases

○ Caused by insufficient or unrepresentative data.

○ Example: Training AI on data from only one group or demographic.

○ Output becomes unfair or inaccurate.

3. People Bias (Designer Bias)

○ Bias introduced by developers themselves.

○ May happen when:

■ Developers focus only on specific goals.

■ Broader ethical or demographic considerations are ignored.

Popular Examples of AI Bias

1. Amazon Recruitment Tool (2014)

○ AI used to filter resumes.

○ Trained on 10 years of past hiring data dominated by men.

○ Result: AI favoured male candidates and ignored resumes with terms like
“women’s hockey”.

○ Outcome: System was scrapped for being gender-biased.

2. COMPAS (U.S. Judicial System)

 ○ Software to predict recidivism (repeat offences).

○ Found to be biased against people of colour.

○ Many U.S. states stopped using it.

3. Healthcare Risk Algorithm (USA)

○ Designed to identify patients needing extra care.

○ Biased against Black patients, who were underrepresented.

○ Selected white patients disproportionately.

4. Facebook Ads (2019)

○ Allowed targeted housing/employment ads excluding:

■ Races, genders, religions, etc.

○ Sued by U.S. Department of Housing and Urban Development.

○ Facebook pledged to remove such discriminatory ad targeting.

5. Facebook Face Recognition

○ Underperformed with:

■ Non-white

■ Non-male individuals.

○ Facebook is working on fixing this using open-source training data.

Prevention of AI Bias
1. Awareness of bias during design and development.

2. Inclusive & large training datasets that fairly represent all groups.

3. Testing and validation to identify biases in output.

4. Continuous monitoring during AI system operation to catch bias over time.

Sources of AI Bias
1. Data

○ AI systems are only as good as the data they are trained on.

○ Real-world data often contains:

■ Human subjectivity

■ Historical inequalities

○ Example: Amazon’s AI system learned to prefer male resumes due to male-

dominated historical data.

2. Algorithms

○ Algorithms may amplify bias present in the data.

○ Example: An image classifier trained on kitchen photos (mostly of women) may

conclude that “people in kitchens are women”, reinforcing stereotypes.

3. People

○ Developers may inadvertently embed their biases in the design.

○ Focus on achieving functional goals may cause ethical concerns to be

overlooked.

○ Reminder: Bias is not the machine’s fault—it’s the human influence behind it.

Key Takeaways
● AI bias is not just a technical issue—it's a social and ethical one.

● Preventing it requires:

○ Ethical awareness

○ Careful data curation

○ Diverse teams

○ Transparent testing and accountability

A. Short Answer Type Questions

1. Name any three important areas where AI has made a remarkable impact.
● Healthcare – AI helps in diagnosis, treatment planning, and medical imaging.

● Education – AI supports automated grading, personalized learning, and real-time

feedback.

● Business – AI is used for sales forecasting, customer service (chatbots), and data-
driven decision-making.

2. How can AI be helpful in space exploration?

● AI helps analyze large datasets from space telescopes like Kepler.

● AI robots like AEGIS assist in autonomous navigation and targeting for NASA’s Mars
missions.

● AI can detect new planets, predict space events, and assist in deep space
communications.

3. How can AI lead to unemployment?

● AI automates repetitive and manual jobs, reducing the need for human workers.

● Industries like manufacturing, retail, logistics, and customer service are especially at
risk.

● Jobs are replaced by machines that perform tasks more efficiently and cost-
effectively.

4. What do you think Artificial Intelligence is?

Artificial Intelligence is the simulation of human intelligence in machines that are designed to
think, learn, and make decisions like humans. It involves capabilities like reasoning, learning,
perception, and problem-solving.
B. Long Answer Type Questions

1. Explain in detail the application of AI in Autonomous Vehicles.

AI plays a critical role in the development and functioning of autonomous (self-driving)
vehicles. Key applications include:

● Perception: AI uses computer vision to process data from sensors (cameras, radar,
LiDAR) to understand the environment.

● Decision Making: AI helps vehicles make real-time decisions such as when to turn,
brake, or accelerate.

● Navigation: AI algorithms use GPS and mapping systems to determine the best
routes.

● Safety: AI predicts pedestrian and vehicle behavior, reducing human error and
increasing safety.

● Companies like Tesla, Waymo, and Uber are investing heavily in AI for self-driving
technology.

2. Give any three important applications of AI in daily life.

1. Smart Assistants: AI-powered assistants like Siri, Alexa, and Google Assistant help
with reminders, voice searches, and home automation.

2. Social Media: AI curates personalized feeds, recommends content, and filters harmful
or inappropriate material.

3. Online Shopping: AI helps in recommendation systems by suggesting products

based on past searches and preferences.

3. List down the three most important jobs at risk due to AI.
1. Customer Service Representatives – Due to the use of AI chatbots.

2. Drivers (Transport & Delivery) – Due to autonomous vehicles and drones.

3. Manufacturing Line Workers – Due to robotics and automation systems.

4. Explain any two sources of AI bias.
1. Data Bias:

○ AI is trained on historical data which may contain human prejudices.

○ If the dataset is incomplete or unbalanced, the AI may give biased results.

○ Example: Amazon’s recruitment system favored male candidates due to male-

dominated training data.

2. People Bias (Developer Bias):

○ AI reflects the assumptions and perspectives of its developers.

○ Developers may unintentionally introduce bias by focusing only on specific

goals and ignoring broader ethical concerns.

○ Example: Design choices that don’t consider diversity can lead to exclusion of
certain groups.