Module Title

Topic 1:
Introduction to Artificial Intelligence

© NCC Education Limited

 Topic 1 Introduction to Artificial Intelligence - 1.2

Scope and Coverage

This topic will cover:

• What is Intelligence?
• Measuring Intelligence
• Artificial Intelligence - Definition
• Human vs. Artificial Intelligence
• The History of Artificial Intelligence
 Topic 1 Introduction to Artificial Intelligence - 1.3

Scope and Coverage

This topic will cover:

• Characteristics of AI
• Human, AI System, and Conventional System
• Limitations of AI
• Ethics of AI
• AI and its Applications
• Emerging Technologies in AI
 Topic 1 Introduction to Artificial Intelligence - 1.4

Learning Outcomes

By the end of this topic students will be able to:

• Explain the meaning of AI and its origin.
• Identify the characteristics of AI.
• Discuss the limitations and ethics of AI.
• Discuss current and future developments in the
field of AI and its applications.
 Topic 1 Introduction to Artificial Intelligence - 1.5

What is Intelligence?

The ability to learn, understand, and

make judgments or have opinions that
are based on reason
- Cambridge Dictionary

The ability to acquire and apply

knowledge and skills.
- Oxford Dictionary
 Topic 1 Introduction to Artificial Intelligence - 1.6

Measuring Intelligence

“ I propose to consider the question,

Can Machines Think? ”
- Alan Turing

• Turing did not provide definitions of machines and

thinking, he just avoided semantic arguments by
inventing a game, the Turing Imitation Game.
 Topic 1 Introduction to Artificial Intelligence - 1.7

Measuring Intelligence with Turing

Imitation Game
Phase 1 Phase 2

(Negnevitsky, 2011)

Turing believed that by the end of the 20th century it would be possible to program a
digital computer to play the imitation game. Although modern computers still cannot
pass the Turing test, it provides a basis for the verification and validation of
knowledge-based systems.
 Topic 1 Introduction to Artificial Intelligence - 1.8

Artificial Intelligence - Definition

“The science and engineering of

making intelligent machines, especially
intelligent computer programs.”
- John McCarthy, founder of AI

“The science of making machines do

things that would require intelligence
if done by men.”
- Marvin Minsky, founder of AI
 Topic 1 Introduction to Artificial Intelligence - 1.9

Artificial Intelligence - Definition

• Artificial intelligence (AI) may be defined as the

branch of computer science that is concerned with
the automation of intelligent behaviour.
• The goal of AI is to make machines do things that
would require intelligence if done by humans. We
can define intelligence as the ability to learn and
understand, to solve problems and to make
decisions.
 Topic 1 Introduction to Artificial Intelligence - 1.10

Human vs. Artificial Intelligence

Human Intelligence Artificial Intelligence

Speak & Communicate Speech Recognition

Read & Write Natural Language Processing

See & Recognise Computer Vision

Move Robotic
See Pattern Pattern Recognition

Cognitive Ability Machine Learning

 Topic 1 Introduction to Artificial Intelligence - 1.11

Human vs. Artificial Intelligence

(Chollet, 2019)

Intelligence is organized hierarchically (Chollet, 2019). AI systems can deal with task-
specific skills (e.g. chess), and are starting to be used in settings requiring broad
generalisation (e.g. self-driving cars). However, they are lacking in terms of extreme
generalization (e.g. general intelligence). Humans master all the levels of the
hierarchy.
 Topic 1 Introduction to Artificial Intelligence - 1.12

The History of Artificial Intelligence

• The emergence of AI (1943 – 1955)

– Warren McCulloch and Walter Pitts (1943) developed
the first computational model of an artificial neural
network (ANN) and proved that suitably defined
networks could learn.
– Alan Turing (1950) published a paper entitled
Computing Machinery and Intelligence to address the
question of machine intelligence. In the paper, he
introduced the Imitation Game, which later became
known as the Turing Test.
 Topic 1 Introduction to Artificial Intelligence - 1.13

The History of Artificial Intelligence

• The emergence of AI (1943 – 1955)

– Claude Shannon (1950) published a paper on chess-
playing machines entitled Programming a Computer for
Playing Chess, which demonstrated the need to use
heuristics in the search for the solution.
– Marvin Minsky and Dean Edmonds (1951) built the first
neural network machine called Stochastic Neural
Analogy Reinforcement Computer (SNARC).
 Topic 1 Introduction to Artificial Intelligence - 1.14

The History of Artificial Intelligence

• The birth of AI (1956)

– John McCarthy, Martin Minsky, Claude Shannon, and
Nathaniel Rochester organised a two-month summer
workshop at Dartmouth College. They brought together
researchers interested in the study of machine
intelligence, neural nets, and automata theory. Although
there were just ten researchers, this workshop gave
birth to a new field, where the term Artificial
Intelligence was coined.
 Topic 1 Introduction to Artificial Intelligence - 1.15

The History of Artificial Intelligence

• The rise of AI, great expectations (1957 – 1969)

– Frank Rosenblatt (1958) continued the work of
McCulloch-Pitts by devising a learning algorithm for a
single layer network called a perceptron.
– John McCarthy (1958) defined a high-level
programming language, Lisp, which is one of the oldest
and the most popular suited programming languages for
AI development.
– Arthur Samuel (1959) developed a self-learning
checkers program. He coined the term machine
learning.
 Topic 1 Introduction to Artificial Intelligence - 1.16

The History of Artificial Intelligence

• The rise of AI, great expectations (1957 – 1969)

– Allen Newell and Herbert Simon (1961) developed the
General Problem Solver (GPS) to simulate human
problem-solving or the thinking humanly approach.
– Joseph Weizenbaum (1966) built the first chatbot or
natural language program called ELIZA.
– Lotfi Zadeh (1965) introduced fuzzy logic or fuzzy set
theory. It provided a means of computing with words
and technology dealing with vague, imprecise and
uncertain knowledge and data.
 Topic 1 Introduction to Artificial Intelligence - 1.17

The History of Artificial Intelligence

• The reality of AI, unfulfilled promises (late 1960s –

early 1970s)
– The promising performance on simple problems (toy
problems) led to AI researchers overconfidence on their
future successes. However, early AI systems turned out to
fail on wider selections and on more difficult problems.
– By 1970, the euphoria about AI was gone, and most
funding for AI projects was cancelled. Many of the
problems that AI attempted to solve were too broad and
too difficult, with little or no knowledge about the
subject matter.
 Topic 1 Introduction to Artificial Intelligence - 1.18

The History of Artificial Intelligence

• The reality of AI, unfulfilled promises (late 1960s –

early 1970s)
– The National Research Council, USA, funded the
translation of Russian scientific papers after the launch
of the first satellite (Sputnik). It was soon found that
translation requires a general understanding of the
subject and was too difficult. In 1966, all translation
projects funded by the US government were cancelled.
– GPS failed to solve complex problems. The amount of
computer time and memory required to solve real-world
problems led to the project being abandoned.
 Topic 1 Introduction to Artificial Intelligence - 1.19

The History of Artificial Intelligence

• The reality of AI, unfulfilled promises (late 1960s –

early 1970s)
– In 1971, the British government reduced support for AI
research. Sir James Lighthill reported to the British
Science Research Council on the current state of AI. He
concluded that “In no part of the field have the
discoveries made so far produced the major impact that
was then promised”.
– The term AI winter refers to a period of reduced funding
and interest in the development of AI.
 Topic 1 Introduction to Artificial Intelligence - 1.20

The History of Artificial Intelligence

• The technology of expert systems, key to success

(1970s – mid-1980s)
– When weak methods failed (general-purpose search),
researchers realised an alternative way to deliver
practical results is to use more powerful, domain-
specific knowledge that allows larger reasoning steps,
and solve typical cases in narrow areas of expertise.
– Expert systems first appeared in the 1970s and again in
the 1980s as an attempt to reduce the computational
requirements by using the knowledge of experts.
 Topic 1 Introduction to Artificial Intelligence - 1.21

The History of Artificial Intelligence

• The technology of expert systems, key to success

(1970s – mid-1980s)
– A number of successful expert system applications in
different areas. Edward Feigenbaum, Bruce Buchanan,
and Joshua Lederberg (1965) developed the first expert
system, DENDRAL, to analyse the molecular structure
of Martian soil.
– MYCIN (1972) was developed to diagnose blood
infections, and PROSPECTOR (1974) built for mineral
exploration.
 Topic 1 Introduction to Artificial Intelligence - 1.22

The History of Artificial Intelligence

• The technology of expert systems, key to success

(1970s – mid-1980s)
– However, problems with expert systems include:
• hard to create and maintain
• restricted to a narrow domain of expertise
• cannot relate heuristic knowledge for deeper
understanding
• no ability to learn from their experience
 Topic 1 Introduction to Artificial Intelligence - 1.23

The History of Artificial Intelligence

• The return of neural networks (mid-1980s - present)
– By the late 60s, the basic ideas and concepts for neural
computing had already been formulated. However, only in
the mid-80s did the solution emerge. The major reason for
the delay was lack of computer technology to model and
experiment with artificial neural networks.
– Bryson and Ho (1969) first introduced the
backpropagation learning algorithm which has been
reinvented by several groups. Rumelhart and McClelland
(1986) resurrected these connectionist models in
Parallel Distributed Processing which caused great
excitement.
 Topic 1 Introduction to Artificial Intelligence - 1.24

The History of Artificial Intelligence

• The return of neural networks (mid-1980s - present)

– Yann LeCun (1980s) discovered the Convolutional
Neural Networks (CNN).
– Hopfield (1982) introduced recurrent neural networks
with feedback called Hopfield networks.
– Kohonen (1982) introduced self-organising maps,
unsupervised learning models based on competitive
learning.
– Barto, Sutton and Anderson (1983) published their work
on reinforcement learning and its application in control.
 Topic 1 Introduction to Artificial Intelligence - 1.25

The History of Artificial Intelligence

• AI adopts the scientific method (1987 - present)

– Recently there has been a revolution in AI content and
methodology of work. It is now more common to build
on existing theories than to propose brand-new ones, to
base claims on rigorous theorems or hard experimental
evidence rather than on intuition, and to show relevance
to real-world applications rather than toy examples.
– Speech technology and the related field of handwritten
character recognition are already making the transition
to widespread industrial and consumer applications.
 Topic 1 Introduction to Artificial Intelligence - 1.26

The History of Artificial Intelligence

• AI adopts the scientific method (1987 - present)

– Similar revolutions have occurred in robotics, computer
vision, and knowledge representation. A better
understanding of the problems and their complexity
properties, combined with increased mathematical
sophistication, has led to robust methods.
– The increasing availability of very large data sources
also emphasis that AI algorithms need to make more
sense about the data and be less worried about what
algorithm to apply.
 Topic 1 Introduction to Artificial Intelligence - 1.27

Checkpoint Summary

• Intelligence is the ability to learn and understand, to

solve problems and to make decisions.
• Artificial intelligence is a branch of computer science to
make machines do things that would require
intelligence if done by humans.
• The history of AI has had periods of success, misplaced
optimism, and resulting cutbacks in enthusiasm and
funding. There have also been periods of introducing
new creative approaches and systematically refining the
best ones.
 Topic 1 Introduction to Artificial Intelligence - 1.28

Characteristics of AI
Reasoning The ability to solve problems through
logical deduction.
Knowledge The ability to present knowledge about
representation the world
Planning The ability to set and achieve goals.
Communication The ability to understand spoken and
written language.
Perception The ability to infer things about the
world via sounds, images, and other
sensory inputs.
Human, AI system, and
Topic 1 Introduction to Artificial Intelligence - 1.29

conventional system
Human AI system Conventional system

Use knowledge in Process knowledge using Process data and use

the form of rules of symbolic, probabilistic, or algorithms, a series of well-
thumb or heuristics logical reasoning to solve defined operations, to solve
to solve problems. problems in a narrow domain. general numerical problems.
In human brain, Provide a clear separation of Do not separate knowledge
knowledge exists in knowledge from its from the control structure or
a compiled form. processing. processing.
Capable of Trace the reasoning during a Do not explain how a
explaining a problem-solving and explain particular result was
reasoning and how a particular conclusion obtained and why input data
providing the was reached and why specific was needed.
details. data was needed.
Human, AI system, and
Topic 1 Introduction to Artificial Intelligence - 1.30

conventional system
Human AI system Conventional system

Use inexact reasoning Allow inexact reasoning and Work only on problems
and can deal with can deal with incomplete and where data is complete
incomplete and uncertain uncertain data. and exact.
information.
Can make mistakes Can make mistakes when Provide no solution, or a
when information is data is incomplete. wrong one, when data is
incomplete. incomplete.
Enhance the quality of Enhance the quality of Enhance the quality of
problem solving via problem solving by adding problem solving by
years of learning and new knowledge or adjusting changing the program
practical training. This old ones in the knowledge code, which affects both
process is slow, base. When new knowledge the knowledge and its
inefficient and is acquired, changes are processing, making
expensive. easy to accomplish. changes difficult.
 Topic 1 Introduction to Artificial Intelligence - 1.31

Limitations of AI

• Lack of common sense and intuition

• The need for large, labeled data sets
• The explainability problem
• Generalisability of learning
• Bias in data and algorithm
• High costs of creation
• Skills shortage
 Topic 1 Introduction to Artificial Intelligence - 1.32

Ethics of AI

• Human agency and oversight: AI systems should enable

equitable societies by supporting human agency and
fundamental rights, and not decrease, limit or misguide
human autonomy.
• Robustness and safety: Trustworthy AI requires
algorithms to be secure, reliable and robust enough to deal
with errors or inconsistencies during all life cycle phases of
AI systems.
• Privacy and data governance: Citizens should have full
control over their own data, while data concerning them will
not be used to harm or discriminate against them.
 Topic 1 Introduction to Artificial Intelligence - 1.33

Ethics of AI

• Transparency: The traceability of AI systems should be

ensured.
• Diversity, non-discrimination and fairness: AI systems
should consider the whole range of human abilities, skills
and requirements, and ensure accessibility.
• Societal and environmental well-being: AI systems
should be used to enhance positive social change and
enhance sustainability and ecological responsibility.
• Accountability: Mechanisms should be put in place to
ensure responsibility and accountability for AI systems and
their outcomes.
 Topic 1 Introduction to Artificial Intelligence - 1.34

AI and its Applications

[Source]
 Topic 1 Introduction to Artificial Intelligence - 1.35

AI and its Applications

Generic Fields

• Artificial creativity • Handwriting recognition

• Computer vision, virtual • Natural language processing,
reality, and image processing translation, and chatterbots
• Diagnosis • Nonlinear control and
• Face recognition robotics
• Game artificial intelligence, • Optical character recognition
computer game bot, game • Speech recognition
theory, and strategic planning
 Topic 1 Introduction to Artificial Intelligence - 1.36

AI and its Applications

Specific Fields

• Artificial life • Intelligent agent and

• Automated reasoning intelligent control
• Automation • Knowledge representation
• Biologically Inspired • Litigation
Computing • Robotics: behaviour-based
• Concept mining robotics, cognition,
• Data mining cybernetics, developmental
robotics (epigenetic), and
• Email spam filtering evolutionary robotics
• Hybrid intelligent system • Semantic web
 Topic 1 Introduction to Artificial Intelligence - 1.37

Emerging Technologies in AI

• Singularity
• Social robot
• Brain-Computer Interface
• Smart Cities
• Robot Sensing
• Augmented Reality
• Behaviour Understanding
 Topic 1 Introduction to Artificial Intelligence - 1.38

Summary

• The five AI characteristics identified are reasoning,

knowledge representation, planning, communication
and perception.
• Unlike conventional programs, AI systems can deal with
incomplete and uncertain data and permit inexact
reasoning. However, like their humans, AI systems can
make mistakes when information is incomplete or fuzzy.
• There are still plenty of limitations and ethical issues of
AI that require considerable effort to overcome,
particularly when running on real-world use cases.
 Topic 1 Introduction to Artificial Intelligence - 1.39

Summary

• AI has advanced more rapidly in the past decade

because of greater use of the scientific method in
experimenting with and comparing approaches.
• Recent progress in understanding the theoretical basis
for intelligence has gone hand in hand with
improvements in real systems' capabilities. The
subfields of AI have become more integrated, and AI
has found common ground with other disciplines.
 Topic 1 Introduction to Artificial Intelligence - 1.40

References
• Russell, S., & Norvig, P. (2016). Artificial
Intelligence: A Modern Approach: Pearson
• Negnevitsky, M. (2011). Artificial Intelligence: A
Guide to Intelligent Systems: Pearson Education
Limited.
• Luger, G. F. (2011). Artificial Intelligence:
Structures and Strategies for Complex Problem
Solving: Pearson Education.
• Chollet, F. (2019). On the measure of intelligence.
https://arxiv.org/abs/1911.01547.
Topic 1 – Introduction to Artificial Intelligence

Any Questions?