AI Applications
Unit-2
AI applications
include Generative or
Understandin
advanced web Creative
g Human
search engines tools
Speech
Automated
Recommendati Self-Driving
Decision-Making ,
on Systems Cars
and competing at
the highest level in
strategic game
AI applications include
advanced web search
engines:
Artificial Intelligence plays a crucial role in the way modern web search
engines work. Search engines like Google, Bing, and DuckDuckGo use AI to
provide faster, smarter, and more relevant search results, making the user
experience seamless and efficient.
AI is used in advanced web search engines like Google Search to
deliver faster, more accurate, and personalized results based on what users
are really looking for—not just matching keywords.
1. Understanding User Intent
(Natural Language Processing -
NLP)
AI allows search engines to understand what a user means, not just what they
type.
Example:
❖ If a user searches "places to eat near me", Google understands this is a query
about restaurants based on location.
❖ Even if you make typos or write incomplete queries, AI still gives relevant
results.
2. AI Algorithms (e.g., RankBrain)
RankBrain is Google's AI-based algorithm that learns from:
❖ User behavior (clicks, bounce rate)
❖ Time spent on pages
❖ Link patterns
It adjusts search result rankings over time based on what users find useful.
3. Voice Search and Conversational
AI
❖ AI powers voice assistants like Google Assistant to understand and
respond to spoken queries.
❖ Voice queries are more natural and longer, and AI helps process them
effectively.
❖ Example: “What’s the weather like tomorrow in Vizag?” – AI understands
location, context, and time.
4. Auto-complete & Smart
Suggestions
As you type, Google provides real-time suggestions powered by AI.
These are based on:
❖ Popular searches
❖ Your own search history
❖ Trending topics
5. Knowledge Graph
Google uses AI to build a web of facts to provide instant answers.
For example, if you search “Who is the Prime Minister of India?”, it shows a
direct answer instead of just links.
6. Contextual Search Results
AI considers location, device, language, time of day, and user preferences to
deliver personalized results.
Example: Searching “coffee shops” in Vizag vs. in Hyderabad will give different
results.
7. Visual Search
With tools like Google Lens, AI can analyze an image and return related search
results.
Example: Take a picture of a plant, and AI identifies it and provides care tips.
Conclusion
AI has made search engines smarter, faster, and user-centric. Instead of
just matching keywords, AI allows search engines to think, learn, and adapt,
creating a more interactive and intelligent search experience for users
around the world.
Recommendation Systems
A Recommendation System is an AI-powered tool that suggests
products, videos, movies, or content to users based on their past behavior,
preferences, and patterns.
It helps businesses increase user engagement, time spent, and
sales/conversions by showing users exactly what they’re likely to want next.
How AI Works in Recommendation
Systems
AI uses techniques like:
❖ Machine Learning – to analyze patterns in user behavior.
❖ Collaborative Filtering – finds similarities between users.
❖ Content-Based Filtering – finds similarities between items.
❖ Deep Learning – to understand complex user interests over
time.
Examples of Recommendation
Systems in Action
1. YouTube:
Recommends videos based on:
● Watch history
● Likes/dislikes
● Subscriptions
● Time spent watching a video
Goal: Keep the user on the platform longer by showing videos they'll
enjoy.
2. Amazon
Suggests products under:
● “Recommended for You”
● “Customers who bought this also bought”
Uses:
● Purchase history
● Browsing behavior
● Products in cart or wish list
Goal: Increase product discovery and sales.
3. Netflix:
Recommends movies/series based on:
● Watch history
● Ratings given
● Genres you watch often
Uses complex AI models to personalize the home screen.
Goal: Keep the user subscribed and entertained.
Benefits of AI-Powered
Recommendation Systems
Benefit Impact on Business & User
Personalized experience Users feel understood
Boost in sales & retention Higher customer loyalty
Better content discovery Users explore more options
Saves time Faster decision-making
Conclusion:
Recommendation systems are one of the most powerful
applications of AI in digital platforms. They improve user satisfaction,
increase engagement, and drive revenue by delivering personalized
suggestions at the right time.
These systems are a key reason why users stay longer on
YouTube, buy more on Amazon, and binge-watch on Netflix.
Understanding Human
Speech (e.g., Siri, Alexa)
Artificial Intelligence allows digital assistants like Siri (Apple) and Alexa
(Amazon) to understand, process, and respond to human speech using
Natural Language Processing (NLP) and Speech Recognition.
How It Works:
1. Speech Recognition:
AI converts your spoken words into text.
Example: When you say “What’s the weather today?”, the system
understands and converts that to a text query.
2. Natural Language Processing (NLP):
● AI interprets the meaning and intent behind your words.
● It understands context, slang, accents, and different sentence
structures.
3. Response Generation:
● AI finds the best answer or action.
● Then it uses Text-to-Speech (TTS) to speak the answer back to
you.
Real-Life Uses:
Assistant Functions
Siri Set alarms, send messages, search the web, give directions
Alexa Control smart home devices, play music, answer questions,
shop online
Benefits of AI in Speech
Understanding:
● Hands-free convenience
● Faster access to information
● Useful for visually impaired users
● Supports multi-language interaction
Conclusion:
AI enables devices like Siri and Alexa to understand and
communicate with humans in natural language, making technology
more accessible, user-friendly, and interactive.
Self-Driving Cars (e.g.,
Waymo)
Self-driving cars (also called autonomous vehicles) use Artificial
Intelligence to drive without human input by sensing the environment,
making decisions, and navigating safely.
Waymo (by Google’s parent company Alphabet) is one of the leading
companies in this field.
How AI Powers Self-Driving Cars
1. Perception (Seeing the World)
Uses cameras, LiDAR, radar, and sensors to detect:
● Road signs
● Lanes
● Pedestrians
● Other vehicles
AI processes this data in real time like a human driver would see
with their eyes.
2. Decision Making:
AI algorithms decide:
● When to stop or turn
● How to change lanes
● How to avoid obstacles or slow-moving vehicles
Uses machine learning and deep learning to improve decisions over
time.
3. Navigation:
● Uses GPS, maps, and AI to plan the safest and fastest route.
● Constantly adjusts based on traffic and road conditions.
Example: Waymo
Waymo cars have no steering wheel or pedals in some models.
Operates fully autonomous taxis in cities like Phoenix, Arizona.
Benefits of AI in Self-Driving Cars:
Benefit Impact
Fewer accidents Eliminates human error
Time-saving Optimized traffic flow and route planning
Eco-friendly Smooth driving reduces emissions
Accessibility Helps people who can't drive (elderly,
disabled)
Conclusion:
Self-driving cars like Waymo use AI to sense, think, and act like a
human driver—making transportation safer, smarter, and more
efficient. They represent the future of AI-powered mobility.
Generative or Creative
Tools (e.g., ChatGPT, AI
Art Tools)
Generative AI tools are computer programs that can create new things, just
like humans do.
These tools can write text, draw images, make music, or even write
computer code, without copying from anywhere.
They are trained to understand patterns and then create something original
based on those patterns.
Examples of Generative AI Tools:
Tool What It Creates
ChatGPT Answers, stories, emails, explanations
(text)
Mid Journey Drawings and digital artwork
Sora (by OpenAI) Short video clips
Amper, AIVA Music
How Do These Tools Work (In
Simple Words)?
1. Learning from Data:
These tools are trained using millions of examples.
For example:
Chat GPT was trained on books, websites, and articles.
AI art tools were trained on paintings, photographs, and digital
designs.
2. Understanding Patterns:
AI learns what good sentences, images, or music look like by
studying many examples.
3. Generating New Content:
When you ask it a question (like “Draw a cat in a spacesuit”), it
doesn’t copy from memory — it creates a new image or sentence
based on the patterns it learned.
3. Navigation:
● Uses GPS, maps, and AI to plan the safest and fastest route.
● Constantly adjusts based on traffic and road conditions.
Example: Waymo
Waymo cars have no steering wheel or pedals in some models.
Operates fully autonomous taxis in cities like Phoenix, Arizona.
Example 1: ChatGPT (Text
Generation Tool)
What it does:
❖ Answers questions (like Google)
❖ Writes stories, blogs, poems, essays
❖ Helps in summarizing long content
❖ Assists in writing emails, resumes, or study material
Example 2: AI Art Tools (Image
Generation)
Popular Tools:
❖ DALL·E by OpenAI
❖ Midjourney
❖ Stable Diffusion
What they do:
❖ Create images from text instructions
❖ (e.g., "a cat sitting on the moon in Van Gogh style")
Conclusion:
Generative AI tools like ChatGPT and AI art apps are changing
how we work, create, and communicate. They help people generate
high-quality content quickly and easily—from writing emails to
drawing digital art—all by learning from data and imitating creativity.
Automated Decision-
Making
Automated Decision-Making (ADM) refers to the process where a computer
system or AI model makes a decision without human intervention.
These decisions are made based on:
● Large amounts of data
● Algorithms (step-by-step instructions)
● Machine learning models (that can learn from data and improve over
time)
The key idea is that once an AI system is trained, it can take inputs (data) and
produce an output (a decision or prediction) automatically.
How Automated Decision-Making
Works
1. Data Collection:
AI systems collect data from various sources.
Example:
● User behavior on websites
● Transaction records
● Medical reports
● Images, videos, text, etc.
This data is the "fuel" for the AI system.
2. Data Processing & Pattern Recognition
AI uses algorithms to analyze the data and look for patterns or
relationships.
Example:
● If people who click on Product A often buy Product B, the AI
learns this association.
● If a patient’s symptoms and reports match past cases of a
disease, AI learns how to recognize it.
3. Decision Logic or Machine Learning Models
Based on the patterns, the AI is programmed or trained to take
decisions.
There are two ways:
Rule-based logic: “If condition A is true, do B.”
Machine Learning: AI learns from past examples and makes
decisions even in new situations.
4. Decision Execution
Once the AI is confident, it executes the decision automatically,
without waiting for a human.
Examples:
● Approving a loan
● Recommending a movie
● Rejecting spam emails
● Diagnosing a disease (suggestion to doctor)
5. Feedback Loop
In more advanced systems, AI keeps learning from new outcomes
and user feedback to improve future decisions.
This is what makes it "intelligent."
Why is Automated Decision-
Making Important?
1. Speed – AI can make decisions in milliseconds, much faster
than humans.
2. Scalability – It can process and decide on thousands or millions
of cases at once.
3. Cost Reduction – It reduces the need for manual staff and
repetitive work.
4. 24/7 Operation – Unlike humans, AI can work non-stop, 24/7.
5. Consistency – AI gives consistent decisions without fatigue or
emotional bias (if trained correctly).
Challenges and Ethical Concerns:
1. Bias in Data – If historical data is biased, AI will also make
biased decisions.
a. Example: If past hiring data favored men, AI may also favor
men.
2. Lack of Transparency – Many AI decisions are made in a
“black box” — hard to understand or explain.
3. Privacy Issues – Automated decisions often use personal data.
4. Accountability – Who is responsible if AI makes a mistake? The
developer, company, or machine?
5. Loss of Jobs – In some sectors, automation may replace human
Real-World Case Studies
Amazon’s Recommendation Engine
Amazon uses automated decision-making to decide:
● What products to show you
● Which ads to display
● When to send you promotional emails
It looks at your browsing history, purchases, and compares you with
similar users.
Real-World Case Studies
Google Ads (Smart Bidding)
Google uses AI to automatically decide
● Which user sees which ad
● How much to bid for each ad placement
This ensures maximum return on investment without human
guesswork.
Real-World Case Studies
Credit Score Analysis
Banks use AI to automatically approve or reject loan applications
based on:
● Credit score
● Salary
● Past behavior
● Risk factors
This is faster, scalable, and reduces manual error — but may be
unfair if the data used is flawed.
Conclusion:
AI-powered automated decision-making is like teaching a
machine to think and decide on its own. It looks at data, learns from
it, and then makes decisions instantly. While it helps save time and
money, it must be used responsibly to avoid mistakes, unfairness, or
lack of control.
AI Competing at the
Highest Level in Strategic
Game Systems
AI systems are now able to outperform even world champions in complex
games that require strategy, planning, and prediction. This is a major proof
of AI’s advanced thinking abilities.
Examples of Strategic Games:
● Chess
● Go
● Shogi (Japanese chess)
How Does AI Do This?
1. Deep Learning + Reinforcement Learning
AI uses techniques like Deep Reinforcement Learning, where it learns
by playing millions of games against itself, improving over time
without needing human data.
2. Self-Play
AI plays with itself over and over again to discover the best
strategies.
Example: AlphaZero (by DeepMind) mastered chess and Go in just a
3. Monte Carlo Tree Search (MCTS)
AI explores possible future moves and evaluates which one gives the
best outcome — like thinking many steps ahead in a tree of choices.
4. Big Data & Fast Computation
AI can process thousands of possible positions in a few seconds,
much faster than any human.
🏆 Notable Success Stories:
Deep Blue vs Garry Kasparov (1997)
IBM’s Deep Blue defeated world chess champion Kasparov. It
marked the first time a reigning champion lost to a computer in a
match under standard conditions.
AlphaGo vs Lee Sedol (2016)
Google DeepMind’s AlphaGo beat Lee Sedol, one of the best Go
players in the world — a major achievement since Go is far more
complex than chess.
Conclusion:
● Shows that AI can handle very complex, dynamic environments.
● Demonstrates planning, adaptation, and creativity — qualities
once thought to be uniquely human.
● The same technology can be used in finance, robotics, logistics,
and warfare strategy.
