Introduction

to AI
By: Ashok Basnet
 • Overview
• Definition of AI
• AI and Related Fields
• Brief History of AI
Course • Applications of AI

Outline •
•
Importance of AI
Definition of Knowledge and Learning
• Importance of Knowledge and
Learning
Any Idea on
AI?
Definition of AI
• AI stands for Artificial Intelligence
• It is a branch of computer science that deals with the development of intelligent machines
that can perform tasks that typically require human intelligence.
• Definition: AI (Artificial Intelligence) can be defined as the simulation of human
intelligence processes by computer systems.
• This involves the development of algorithms and models that enable machines to perform
tasks that would normally require human intelligence, such as visual perception, speech
recognition, decision-making, and language translation.
• AI systems can also learn and improve over time, using algorithms and data to identify
patterns and make predictions.
• John McCarthy: John McCarthy is credited with coining the term "Artificial Intelligence" in 1956. He
defined AI as "the science and engineering of making intelligent machines."

• Marvin Minsky: Marvin Minsky was a pioneer in the field of AI and co-founder of the MIT AI
Laboratory. He defined AI as "the science of making machines do things that would require
intelligence if done by humans."

• Ray Kurzweil: Ray Kurzweil is a futurist and inventor who has written extensively on AI. He defines AI
as "the engineering of making intelligent machines, especially intelligent computer programs."

• Stuart Russell: Stuart Russell is a leading AI researcher and professor at UC Berkeley. He defines AI as
"the development of computer systems that can perform tasks that would normally require human
intelligence, such as visual perception, speech recognition, decision-making, and language
translation."
Think Humanly: The Cognitive Modelling Approach

• The exciting new effort to make computers think… machines with minds, in the full and
literal sense. (Haugeland, 1985)
• “The automation of activities that we associate with human thinking, activities such as
decision-making, problem solving, learning…”(Bellman, 1978)
• This approach aims to model human cognition and use this model to create intelligent
machines.
• Cognitive psychology and neuroscience research are used to understand how humans
think, reason, and solve problems.
• Needs understanding of how human thinks?
• Example: General Problem Solver-GPS (Newell and Simon, 1961)
Act Humanly: The Turing Test Approach

• The art of creating machines that perform functions that require intelligence when
performed by people.(Kurzwail, 1990)
• The study of how to make computers do things at which, at the moment, people are
better.(Rich and Knight, 1991)
• Based on Turing Test (Alan Turing, 1950)
• Test based on indistinguishability from undeniably intelligent entities (human).
• The computer passes a test if a human interrogator, after posing some written
questions, can’t tell whether the responses were made by a human or not.
Think Rationally: The “Laws of Thought” Approach
• The study of mental faculties through the use of computational models. (Charniak and
McDermott, 1985)
• The study of the computations that make it possible to perceive, reason and act.
(Winston, 1992)
• Based on Rational Thinking (Right Thinking)
• Irrefutable Reasoning Process
• Syllogisms providing patterns for argument structures that always yielded correct
conclusions
• Logic: the Laws of thought
• Logicist tradition within AI hopes to build on such programs to create intelligent
systems
Act Rationally: The Rational Agent Approach
• Computational Intelligence is the study of the design of intelligent agents. (Poole et al.,
1998)
• AI… is concerned with intelligent behaviour in artifacts. (Nilsson, 1998)
• Based on Intelligent Agents
• Agents are the things that act.
• Computer agents are expected to have other attributes that distinguish them from
mere “programs”, such as operating under autonomous control, perceiving their
environment, persisting over a prolonged time period, adapting to change, and being
capable of taking on another’s goals.
• Rational agents are those who act so as to achieve the best outcome or, the best
expected outcome when there is uncertainty.
Types of AI
• There are three types of AI:
• Artificial Narrow Intelligence (ANI)
• Artificial General Intelligence (AGI)
• Artificial Super Intelligence (ASI)
• ANI refers to AI that is designed for a specific task, such as speech recognition or
image classification.
• AGI refers to AI that is capable of performing any intellectual task that a human can.
• ASI refers to hypothetical AI that surpasses human intelligence in every possible way.
AI and Related Fields
• Computer Science:
• AI is rooted in computer science, which provides the hardware and software
infrastructure that enables AI systems to operate.
• Mathematics:
• Mathematics plays a central role in AI, as it provides the foundation for algorithms and
models used in machine learning and other AI techniques.
• Statistics:
• Statistics is used extensively in AI to analyze data and make predictions about future
outcomes.
• Logic:
• Logic is used in AI to represent and reason about knowledge and decision-making.
• Neuroscience:
• AI researchers draw inspiration from neuroscience to model the behavior of the brain
and develop algorithms for learning and decision-making.
AI and Related Fields
• Philosophy:
• AI raises many philosophical questions, such as the nature of consciousness,
intelligence, and free will.
• Philosophical inquiry helps to shape the goals and ethical considerations of AI
research.
• Linguistics:
• Natural language processing, a key area of AI, draws heavily on the field of linguistics
to understand and generate human language.
• Psychology:
• AI researchers also draw inspiration from psychology to model human cognition and
behavior, and to design intelligent systems that can interact with humans in natural
and intuitive ways.
Brief History of AI
The term "artificial intelligence" was first used
Warren McCulloch and Walter Pitts created the by John McCarthy, Marvin Minsky, Nathaniel
first artificial neuron model, which became the Rochester, and Claude Shannon at the
basis for future neural networks. Dartmouth C

1943 1956

1950

Alan Turing published a paper titled "Computing

Machinery and Intelligence," in which he
proposed the Turing Test.
Brief History of AI

Shakey, the first mobile robot capable of

John McCarthy invented the Lisp reasoning about its own actions, was
programming language, which became the developed at the Stanford Research
main language for AI research. Institute.

1958 1969

1964 1975

Joseph Weizenbaum created ELIZA, a The first expert system, MYCIN, was
computer program that could simulate a developed at Stanford University to
conversation with a human and is diagnose blood infections.
considered one of the earliest examples of
natural language processing (NLP).
Brief History of AI

The first autonomous vehicle, the Navlab, was IBM's Watson defeated human champions in the
developed by Carnegie Mellon University. game show Jeopardy!

1981 2011

1997

IBM's Deep Blue defeated chess champion Garry

Kasparov, marking the first time a computer beat
a human world champion in a game of chess.
Brief History of AI
Google's DeepMind developed a
neural network called AlexNet, which OpenAI's GPT-2, a language model
achieved a significant breakthrough in capable of generating human-like
image recognition. text, was released.

2012 2018

2016 2020

AlphaGo, a program developed by GPT-3, a more advanced language

DeepMind, defeated the world model than GPT-2, was released by
champion in the board game Go. OpenAI.
Brief History of AI: Success Stories
Chess-playing computers: In 1997, IBM's Deep Blue computer defeated the reigning world
chess champion, Garry Kasparov, in a six-game match.

Speech recognition: Voice assistants like Siri, Alexa, and Google Assistant have become
ubiquitous in recent years.

Autonomous vehicles: Companies like Tesla and Waymo have made significant progress in this
area.

Medical diagnosis: AI algorithms can analyze medical images, such as CT scans and X-rays, to
identify potential issues and help doctors make more accurate diagnoses.

Fraud detection: AI can analyze large datasets to identify patterns and anomalies that might
indicate fraudulent activity, and this can help prevent financial losses and protect consumers.
Heros of AI
25
26
27
Applications of AI
AI is used in various industries,
including healthcare, finance, and
transportation.

In healthcare, AI is used for

medical image analysis and drug
discovery.

In finance, AI is used for fraud

detection and risk assessment.

In transportation, AI is used for

self-driving cars and traffic
management.
Importance of AI
Automation: AI is capable of automating repetitive and mundane tasks,
allowing humans to focus on more complex and creative work.

Decision-making: AI can process large amounts of data and make

complex decisions based on that data.

Personalization: AI can analyze data about individual preferences and

behaviors to personalize products and services.

Improved healthcare: AI can help doctors and medical professionals make

more accurate diagnoses and develop more effective treatment plans.
Importance of AI
Increased safety: AI can be used to improve safety in a variety of industries,
including transportation, manufacturing, and construction.

Advancements in science and research: AI can help scientists and researchers

process large amounts of data and make new discoveries. This can accelerate
scientific progress and lead to new breakthroughs in a variety of fields.

Addressing global challenges: AI can be used to address some of the world's

most pressing challenges, such as climate change, poverty, and disease.
 Challenges of AI
• One of the biggest challenges of AI is ethical concerns,
such as bias and privacy.
• There are also concerns about the impact of AI on jobs
and the economy.
• Another challenge is the lack of transparency in how
AI systems make decisions, which can make it difficult
to understand and trust them.
 • AI has the potential to transform various
industries and improve our lives in many
ways.
• However, it also poses significant
challenges that need to be addressed.
Conclusion • It is important to continue to develop AI
in a responsible and ethical manner to
ensure its benefits are maximized and
its risks are minimized.
 Data, Information, Knowledge
and Wisdom
• Data, information, knowledge, and wisdom are hierarchical concepts that
represent different levels of understanding or processing of information.
• Data:
• Data is a collection of facts or observations that are typically in a raw and unorganized
form.
• For example, a list of numbers or a spreadsheet containing sales figures for a particular
month is data.
• Information:
• Information is data that has been processed, organized, and presented in a meaningful
way.
• For example, if we take the sales figures from the previous example and organize them
into a bar graph or a table, we now have information that can be used to understand
how sales have changed over time.
 Data, Information, Knowledge
and Wisdom
• Knowledge:
• Knowledge is information that has been assimilated and understood by a person.
• It involves connecting different pieces of information to form a deeper understanding.
• For example, if a sales manager examines the sales figures and identifies a trend that
sales tend to increase during certain times of the year, they have gained knowledge
about the sales patterns of the company.
• Wisdom:
• Wisdom is the highest level of understanding, and it involves using knowledge to
make judgments and decisions.
• It is a more holistic and subjective understanding of a situation, based on experience,
intuition, and reflection.
• For example, a wise sales manager might use their knowledge of the company's sales
patterns to make strategic decisions about when to launch new products or
promotions.
Learning
• In AI, learning refers to the process of improving a machine's performance
on a particular task by providing it with examples or feedback.
• There are several types of learning in AI, including supervised learning,
unsupervised learning, and reinforcement learning.
• In supervised learning, the machine is provided with labeled examples of
inputs and outputs, and it learns to make predictions based on those
examples.
• In unsupervised learning, the machine is not given labeled examples, and
it must find patterns or structures in the data on its own.
• In reinforcement learning, the machine learns by receiving feedback in the
form of rewards or punishments, based on the actions it takes.
Thank you