INTRODUCTION TO

ARTIFICIAL
INTELLIGENCE

Nguyễn Ngọc Thảo – Nguyễn Hải Minh

{nnthao, nhminh}@fit.hcmus.edu.vn
Outline
• What is Artificial Intelligence (AI)?
• The foundations of AI
• A brief history of AI
• What are we going to learn?

2
Which one is a real image?

3
Which one is a real image?

4
Is it a real video?

5
 Amazon Go: A store of the future

6
Source: YouTube video
 Humanoid robots: Atlas Robot

7
Source: YouTube video
 Tesla’s Autopilot: Automated driver

8
Source: YouTube video
You are just
getting
started!
9
What is AI?
Intelligence vs. Artificial Intelligence
Intelligence includes the capacity for logic, understanding,
learning, reasoning, creativity, and problem solving, etc.

Artificial intelligence (AI) attempts not just to understand

but also to build intelligent entities.
11
The field of Artificial Intelligence
• AI is one of the newest fields in science and engineering.
• Work started in earnest soon after World War II
• The name was coined at a conference at Dartmouth College in
1956.

John McCarthy Marvin Minsky Allen Newell Arthur Samuel Herbert Simon
(1927 – 2011) (1927 – 2016) (1927 – 1992) (1901 – 1990) (1916 – 2001)

12
The goals of Artificial Intelligence
• AI research builds intelligent entities that simulate humans
in different aspects.

✓ Thinking: learning, planning, and

refining knowledge

✓ Perception: see, hear, feel, etc.

✓ Communication in natural languages

✓ Manipulation and moving objects

• AI studies the intelligent part concerned with human and

represents those actions using computers.
13
What is Artificial Intelligence?

14
What is Artificial Intelligence?
Thought processes and reasoning

Systems that Systems that

Rationality
think think
Humans

like humans rationally

Systems that Systems that

act act
like humans rationally

Behavior 15
Systems that act like humans
• The Turing Test approach (Alan Turing, 1950)

A computer passes the test if a human interrogator, after posing several written
questions, cannot tell whether the responses come from a person or a computer.

16
Systems that act like humans
• The Turing test only checks whether the computer behaves
like a human being, but not whether it behaves intelligently.

ntelligent
nintelligent
Turing ehavior
human
test humans
ehavior
don t do

17
 A better Turing Test?
• AI researchers have devoted little effort to pass the test.
• It is more important to study the underlying principles of
intelligence than to duplicate an exemplar.

18
Image credit: KNLIT
Systems that think like humans
• General Problem Solver – GPS (Newell and Simon, 1961)
• Not merely solve problems correctly
• Compare the trace of its reasoning steps to traces of human subjects
while solving the same problems

• Cognitive Science
• Computer models from AI precise and testable
theories of
• Experimental techniques from psychology the human mind

• These approaches are now distinct from AI

• Share the available theories but do not explain anything resembling
human intelligence
• All share a principal direction

19
Systems that think rationally
• “Right thinking” = irrefutable reasoning processes
• E.g., Aristotle’s syllogisms provided argument patterns that always
yielded correct conclusions when given correct premises.

All men are mortal. x.man(x) mortal(x)

Socrates is a man. man(Socrates)
Therefore,
Socrates is mortal. mortal(Socrates)

• There are obstacles applied to any attempt to build

computational reasoning systems
• Not all intelligence is mediated by logic behavior
• Solving a problem “in principle” is different from doing in practice
20
Systems that act rationally
• The rational agent approach
• Rational behavior = “doing the right thing”.
• “Right thing”: what is expected to maximize goal achievement given
the available information
• An agent is just something that perceives and then acts
𝒇: 𝓟 → 𝓐
• A rational agent acts to achieve the best outcome or, when
there is uncertainty, the best expected outcome.
• Include thinking, inference as a part of being rational agent
• Include more: action without thinking, e.g., reflexes

21
Systems that act rationally
• A behavior is either a reflex action or an intelligent one.
• A reflex action can be rational or not, while an intelligent
action is usually rational.
• An intelligent behavior is usually obtained via a learning process.

A man withdraws his fingers from a hot stove. Two people cross the street
at the crosswalk.
22
Systems that act rationally
• More general than the “laws of thought” approach
• Correct inference is not all of rationality.
• In some situations, there is no provably correct thing to do, but
something must still be done.
• Amenable to scientific development than those based on
human behavior or human thought

23
Pros and Cons of AI
More powerful and more useful computers
New and improved interfaces
Solve new problems
Better handling of information
Relieve information overload
Conversion of information into knowledge

Increased costs
Difficulty with software development - slow and expensive
Few experienced programmers
24
Foundations
of AI
Research fields related to AI

Control theory Mathematics

and
cybernetics Philosophy

Linguistics Neuroscience

Economics Computer Psychology

Engineering
26
Research fields related to AI
Field Description
Logic, methods of reasoning, mind as physical
Philosophy system, foundations of learning, language,
rationality.
Formal representation and proof, algorithms,
Mathematics computation, (un)decidability, (in)tractability,
probability.

Economics Utility, decision theory, rational economic agents

Neuroscience Neurons as information processing units.

Psychology/ How do people behave, perceive, process
Cognitive Science information, represent knowledge.
Computer Engineering Building fast computers
Design systems that maximize an objective
Control Theory
function over time
Linguistic Knowledge representation, grammar 27
Research fields related to AI

28
A brief
history of AI
A brief history of AI
• 1940-1950: Early days
• 1943: McCulloch & Pitts: Boolean circuit model of brain
• 1950: Turing's “Computing Machinery and ntelligence”
• 1950—70: Excitement: Look, Ma, no hands!
• 1950s: Early AI programs, including Samuel's checkers program, Newell & Simon's
Logic Theorist, Gelernter's Geometry Engine
• 1956: Dartmouth meeting: “Artificial ntelligence” adopted
• 1965: Robinson's complete algorithm for logical reasoning
• 1970—90: Knowledge-based approaches
• 1969—79: Early development of knowledge-based systems
• 1980—88: Expert systems industry booms
• 1988—93: Expert systems industry busts: “A Winter”

30
A brief history of AI
• 1990—: Statistical approaches
• Resurgence of probability, focus on uncertainty
• General increase in technical depth
• Agents and learning systems… “A Spring”?
• 2000—: Where are we now?

Source: Stack Exchange

31
AI Innovations: Deep Blue – AlphaGo

AlphaGo vs. Lee Sedol

(03/2016)

Deep Blue vs. Kasparov

(02/1996 and 05/1997)

32
 The complexity of Chess and GO

33
Source: YouTube video
 AI Innovations: OpenAI Five

34
Source: OpenAI Five
What are we
going to learn?
Problem solving by search
• Search strategies for single-agent environments
• Adversarial search
• Constraint satisfaction problems

36
Knowledge and reasoning

• Propositional logic and predicate logic

• Inference techniques: forward chaining, backward chaining,
and resolution
• Uncertain knowledge and reasoning
37
 Machine learning

Image credit: 38
YouTube video
39