ARTIFICIAL

INTELLIGENCE
 OBJECTIVES

Understand the Origins of AI and its evolution

Define, Understand and Create basic concepts of simple search

algorithms in AI

Define, Understand and Create basic concepts on Machine learning

algorithms breadth first search and depth first search algorithm
 WHAT IS ARTIFICIAL
INTELLIGENCE AND HOW
DOES IT WORK?
 THE HISTORY AND
EVOLUTION OF ARTIFICIAL
INTELLIGENCE
 4th Century BC - Syllogistic logic was being introduced —
one of the earliest contributions by Aristotle, a system of
reasoning that laid the groundwork for deductive thinking,
a key element in AI.

SYLLOGISTIC LOGIC

PHILOSOPHICAL
FOUNDATIONS
OF AI

16th Century - René Descartes contributed by proposing

that animal bodies are complex machines, influencing the
later mechanistic views of intelligence.
 12th century - "Talking heads" devices were reportedly
invented by thinkers like Roger Bacon and Albert the Great,
attempting to mimic human speech.

12TH-15TH
CENTURY
INVENTIONS
15th century - clockmakers began designing
mechanical animals, and stories like Rabbi Loew’s golem
—a clay figure brought to life—reflected early human
fascination with creating lifelike entities.
 17th century - a rise in mechanical and computational
advancements.
- Blaise Pascal created the first digital calculating
machine, while Thomas Hobbes (1651) introduced mechanical
theories of thought, laying the groundwork for considering the
mind as a machine.
16TH-17TH
CENTURY
INNOVATIONS

Leibniz further improved Pascal’s machine, focusing on

more complex calculations, including multiplication and
division.
18th century - Joseph-Marie Jacquard invented the first
programmable device—the Jacquard loom— a key
advancement in automation.

18TH CENTURY:
AUTOMATION
AND LITERATURE

Mary Shelley’s novel Frankenstein (1818) explored themes of

artificial life and the ethical dilemmas of creating sentient
beings.
 Modern AI foundations were laid in the 1940s and 50s

1943: Warren McCulloch and Walter Pitts published a key

THE RISE OF paper on neural networks.

MODERN AI 1950: Alan Turing introduced the Turing Test to measure

machine intelligence
(1940S-1950S)
1950: Isaac Asimov introduced the Three Laws of Robotics
as ethical guidelines for AI development
 1956: John McCarthy coined the term "Artificial Intelligence"
THE BIRTH OF AI during the Dartmouth Summer Research Project.

AS A FIELD This marked the birth of AI as a formal field of study

(1956) The conference laid the foundation for decades of AI

research and development

AI research faced setbacks during the "AI Winter" of the

1960s and 70s.
AI WINTERS
AND Slowed progress due to poor results in machine translation
and limitations in perceptron models (1969).
CHALLENGES 1975: Breakthrough with the development of the
backpropagation algorithm, enabling more efficient
neural network training.
1980s: Rise of expert systems, knowledge-based AI
systems.

Lisp became a key programming language for AI

research.

EXPERT
SYSTEMS AND
NEURAL
NETWORKS
Early 1990s: Expert systems failed to meet expectations, (1980S-1990S)
leading to disillusionment with AI.

Late 1990s: Neural networks fell out of favor; support

vector machines gained prominence.
 2000s: Resurgence of AI driven by the rise of data science.
Machine learning algorithms applied across industries to
solve real-world problems
AI IN THE
2010s: Deep learning became a major focus. Deep neural 2000S-2010S
networks achieved state-of-the-art results in tasks like
image recognition, natural language processing, and
machine translation

Late 2010s - early 2020s: Models like GPT (Generative Pre-

trained Transformers) gained attention for generating
THE MODERN
human-like text and performing diverse tasks.
ERA: 2020S
Pushed the boundaries of AI capabilities across industries. AND BEYOND
Today, AI is integrated into many areas, from virtual
assistants to autonomous vehicles, shaping the future of
technology.
 Two main approaches to AGI:
ARTIFICIAL
SYMBOLIC AI
GENERAL
-uses logic and reasoning to solve problems.
INTELLIGENCE
-excels in tasks requiring structured reasoning but struggles
with unstructured or unambiguous problems.
“Artificial General Intelligence” or AGI
refers to AI systems that are capable of
-less adaptable
performing any intellectual task that a
human could do.

It aims to develop AI systems that can

NEURAL NETWORK-BASED AI
learn and adapt like humans for a wide
-inspired by the human brain.
range of tasks.

-more flexible and adaptive but can be less reliable and

harder to interpret.
 AI has evolved from symbolic AI to advanced methods like
deep learning and reinforcement learning
THE CURRENT
The field is rapidly developing, with new advances
STATE OF AI expanding AI capabilities quickly
 ARTIFICIAL
Key characteristics of ANI: NARROW
INTELLIGENCE
Fixed Purpose

-systems are limited to the tasks they were designed for and
(ANI)
cannot transfer skills to other domains.

Artificial Narrow Intelligence (ANI) refers

Lack of Adaptability to AI systems designed to perform a
specific task or solve a particular
-ANI systems cannot handle new or unexpected situations.
problem, without the ability to learn or
adapt beyond that scope. Examples
include IBM’s Deep Blue chess-playing
Non-Self-Learning program and self-driving cars.

-do not improve or learn independently.

AGI aims to surpass ANI by creating adaptable, learning
systems

Capable of handling diverse tasks and improving over ARTIFICIAL

time
GENERAL
AGI could revolutionize fields like healthcare (e.g.,
diagnoses, treatments, drug development) INTELLIGENCE
Current AI lacks human-like traits (e.g., common sense, (AGI)
creativity)

Many experts believe AGI is still years away from reality

ARTIFICIAL ASI refers to AI that is more intelligent than any human
being, and that is capable of improving its own
SUPER capabilities over time.

INTELLIGENCE Some experts worry that ASI could pose serious risks to
humanity, while others believe that it could be used for
tremendous good.
(ASI)
There are no fully realized ASI systems in existence today.
However, there are some systems that are starting to
approach the capabilities that would be considered ASI.

For example, there are some language models, like GPT-4,

that are able to generate text that is very close to human-
level quality.
GENERATIVE AI
Text generation: GPT-3, which can generate human-like
Generative AI refers to AI systems that text for tasks like chatbots and content creation.
create new content or data from scratch
rather than just analyzing existing data
like other types of AI.. Image generation: DALL-E 2, capable of producing
images from textual descriptions.

Music generation: AI systems that can compose original

music.
 AI SEARCH
ALGORITHMS
WHAT ARE
TRADITIONAL VS AI-POWERED
TRADITIONAL AI-POWERED
AI SEARCH ASPECT SEARCH SEARCH
ALGORITHMS Search Method Keyword-based Context-based

An AI search algorithm
is the method for Provides links to Direct answers or
Results
web pages summaries
understanding natural
language queries and
Users click links and
finding relevant results User Interaction Conversational
browse pages
by evaluating indexed
data and documents. Treats each query Remembers context
Context Handling from previous
separately interactions

Requires browsing Provides concise,

Efficiency through multiple links instant answers
 TYPES OF AI SEARCH ALGORITHMS

MACHINE LEARNING- NATURAL LANGUAGE DEEP LEARNING AND

BASED ALGORITHMS PROCESSING (NLP) NEURAL NETWORKS
Machine learning (ML) NLP allows machines to Deep learning is a subset
algorithms enable systems understand, interpret, of ML that uses artificial
to learn from data and and generate human neural networks, modeled
improve their performance language. after the human brain, to
over time without being solve complex problems.
explicitly programmed.
FUNDAMENTAL TYPES OF SEARCH
ALGORITHMS
 CHALLENGES IN AI SEARCH

COMPUTATION INFORMED SEARCH LIMITATIONS TO THE

COMPLEXITY ALGORITHM CAN ONLY BE PROBLEM-SOLVING
AS GOOD AS THE SCOPE
HEURISTICS
HOW AI SEARCH
ENHANCES USER
EXPERIENCE
Personalized Search
Predictive Capabilities
Improved Accuracy and Relevance
APPLICATIONS AND
EXAMPLES

Informational retrieval
Recommendations
Speech recognition
Medical diagnosis
Pathfinding
POTENTIAL DEVELOPMENTS

Improved natural More sophisticated Enhanced multi-

language personalization modal search
understanding techniques capabilities
 ETHICAL CONSIDERATIONS
AND DATA PRIVACY

Maintaining
Ensuring user data Addressing biases in
transparency in how
protection AI algorithms
search results are
generated
BASIC CONCEPT Query Initiation
SAMPLE
Heuristic Evaluation
(Greedy Search)

Optimization
(A*)

Ranking

Display
WHAT IS MACHINE
LEARNING (ML)?
is a branch of AI and computer
science that focuses on the using data
and algorithms to enable AI to imitate
the way that humans learn, gradually
improving its accuracy.
TYPES OF
ML
Supervised learning
Unsupervised learning
Reinforcement learning
SUPERVISED LEARNING
Supervised Learning is akin to a teacher
guiding a student. In this approach, the
model is provided with a labeled dataset,
where each example has input features and
corresponding target outputs. The goal is for
the model to learn a mapping from inputs to
outputs.
UNSUPERVISED LEARNING
Unsupervised Learning is more like letting a
child explore a room full of toys without any
instructions. Here, the model is given an
unlabeled dataset and is tasked with finding
patterns or structures within the data on its
own.
REINFORCEMENT LEARNING
Reinforcement Learning is like training a dog.
The model interacts with an environment and
receives feedback in the form of rewards or
penalties based on its actions. It learns to
take actions that maximize its cumulative
reward over time.
BREADTH-FIRST SEARCH
ALGORITHM
Breadth First Traversal or Breadth First
Search is a recursive algorithm for
searching all the vertices of a graph or
tree data structure.

It begins at a specific node and explores all

neighboring nodes at the current level
before moving on to nodes at the next level
BREADTH-FIRST SEARCH ALGORITHM

The algorithm works as follows:

1. Start by putting any one of the graph's vertices at the
back of a queue.
2. Take the front item of the queue and add it to the visited
list.
3. Create a list of that vertex's adjacent nodes. Add the
ones which aren't in the visited list to the back of the
queue.
4. Keep repeating steps 2 and 3 until the queue is empty.
EXAMPLE OF HOW BFS ALGORITHM WORKS

Visited 1

Queue 5 2
EXAMPLE OF HOW BFS ALGORITHM WORKS

Visited 1 5

Queue 2 3
EXAMPLE OF HOW BFS ALGORITHM WORKS

Visited 1 5 2

Queue 3 4
EXAMPLE OF HOW BFS ALGORITHM WORKS

Visited 1 5 2 3

Queue 4 7
EXAMPLE OF HOW BFS ALGORITHM WORKS

Visited 1 5 2 3 4

Queue 7 6
EXAMPLE OF HOW BFS ALGORITHM WORKS

Visited 1 5 2 3 4 7

Queue 6
EXAMPLE OF HOW BFS ALGORITHM WORKS

Visited 1 5 2 3 4 7 6

Queue
APPLICATIONS OF BREADTH-FIRST SEARCH
ALGORITHM

Peer-to-Peer Networks: In Peer-to-Peer Networks like

BitTorrent, Breadth First Search is used to find all neighbor
nodes.
GPS Navigation systems: Breadth First Search is used to
find all neighboring locations.
Image processing: BFS can be used to flood-fill an image
with a particular color or to find connected components
of pixels.
DEPTH-FIRST SEARCH
ALGORITHM
The depth-first search or DFS algorithm traverses or explores
data structures, such as trees and graphs. The algorithm
starts at the root node (in the case of a graph, you can use
any random node as the root node) and examines each
branch as far as possible before backtracking. When a
dead-end occurs in any iteration, the Depth First Search
(DFS) method traverses a network in a deathward motion
and uses a stack data structure to remember to acquire the
next vertex to start a search.
STEPS OF THE DFS ALGORITHM:

Start at the root (or any starting node).

Mark the node as visited to avoid revisiting it.
Explore an unvisited adjacent node, push it onto the stack
(or recursive call), and repeat this process.
Backtrack when you hit a dead-end (i.e., no unvisited
adjacent nodes), popping nodes off the stack (or
returning from recursion).
Continue until all nodes have been visited.
EXAMPLE OF DFS ALGORITHM:

Initialize the stack.

EXAMPLE OF DFS ALGORITHM:

Mark S as visited and put it

onto the stack. Explore any
unvisited adjacent node
from S. We have three
nodes and we can pick any
of them. For this example,
we shall take the node in an
alphabetical order.
EXAMPLE OF DFS ALGORITHM:

Mark A as visited and put it

onto the stack. Explore any
unvisited adjacent node
from A. Both S and D are
adjacent to A but we are
concerned for unvisited
nodes only.
EXAMPLE OF DFS ALGORITHM:

Visit D and mark it as visited

and put onto the stack.
Here, we have B and C
nodes, which are adjacent
to D and both are unvisited.
However, we shall again
choose in an alphabetical
order.
EXAMPLE OF DFS ALGORITHM:

We choose B, mark it as
visited and put onto the
stack. Here B does not have
any unvisited adjacent
node. So, we pop B from the
stack.
EXAMPLE OF DFS ALGORITHM:

We check the stack top for

return to the previous node
and check if it has any
unvisited nodes. Here, we
find D to be on the top of the
stack.
EXAMPLE OF DFS ALGORITHM:
Only unvisited adjacent
node is from D is C now. So
we visit C, mark it as visited
and put it onto the stack.
As C does not have any unvisited
adjacent node so we keep
popping the stack until we find a
node that has an unvisited
adjacent node. In this case,
there's none and we keep
popping until the stack is empty.
EXAMPLE OF DFS ALGORITHM:
As C does not have any unvisited
adjacent node so we keep
popping the stack until we find a
node that has an unvisited
adjacent node. In this case,
there's none and we keep
popping until the stack is empty.
APPLICATION OF DFS ALGORITHM:
Maze Solving: DFS can be used to explore a maze. It keeps
going down one path as far as possible before
backtracking when a dead end is reached.

Topological Sorting: DFS is used in topological sorting of

directed graphs to determine a linear ordering of
vertices.

Finding Connected Components: In undirected graphs,

DFS helps identify all connected components by
exploring all reachable nodes from a given starting node.
 ARTIFICIAL INTELLIGENCE
END OF PRESENTATION