Artificial Intelligence

Lecture 1 - Overview of
Artificial Intelligence

Lecturer: Vu Trong Sinh

1
Quiz
Which of the following applications is AI?

A. A website traffic monitoring dashboard that automatically sends alerts when

there are over 1,000 concurrent visits.

B. A security camera app that automatically highlights moments when someone

enters through the door.

C. A Facebook fan page that automatically sends greetings when someone sends
a message.

D. The Canva app that can generate illustrations for input text content.

2
Content
1.1. Intelligence & the Turing test

1.2. AI basic concepts

Definition

Milestones

1.3. Subfields and related fields in AI

1.4. Job opportunities in AI

3
1.1. Intelligence & the Turing test
What is intelligence?

Distinguish between AI and non AI systems

Computers can be intelligent like human?

How to measure intelligence?

4
What is intelligence?
Take an IQ test https://test.iqtestacademy.org/

- Why the IQ test can measure intelligence?

- What kinds of questions are used in an IQ test, why?
- Is the IQ test enough for measuring intelligence?

5
Human intelligence
Intelligence, in the human context, is defined as a person’s ability to learn, to
deal with new situations, to understand abstract concepts, and use
knowledge to manipulate one’s environment (Sternberg, 2017)

https://www.sciencedirect.com/topics/psychology/human-intelligence

In more general terms, intelligence is defined as the ability to perceive and

process data, transform data into information and ultimately knowledge, and use
this knowledge towards goal-directed behavior

6
Human intelligence
K.R. Chwodhary, 2020 - Fundamentals of AI

» Learning: Having the ability to obtain and process new information.

» Reasoning: Being able to manipulate information in various ways.

» Understanding: Considering the result of information manipulation.

» Grasping truths: Determining the validity of the manipulated information.

» Seeing relationships: Divining how validated data interacts with other data.

» Considering meanings: Applying truths to particular situations in a manner consistent with their
relationship.

» Separating fact from belief: Determining whether the data is adequately supported by provable
sources that can be demonstrated to be consistently valid.

7
AI vs non-AI
Learning Capability:

learn from data and improve over time

Automation:

operate autonomously and adapt to

new environments

Recognition and Understanding:

recognize patterns, objects, and

understand natural language to some extent

8
Discussion 1
- What is your favourite AI application

- Which company develops

- What is the impact of that app, to the social and to you?

9
 Face ID
Siri virtual assistant Photos application

10
e-Commerce recommendations

Camera fines for traffic violations

11
 Auto-pilot
cars

12

Auto-caption and
translation
The Turing test
Discussion:

Computers can be intelligent like human?

ChatGPT/Gemini is at human intelligence level?

13
The Turing test
“Computing Machinery and Intelligence” - Alan Turing, 1950

- Empirical test for machine intelligence → measure the performance of an

intelligent machine against humans, for its intelligent behavior

This test includes:

- An interrogator (referee)
- A machine to be evaluated
- A human to compare
- Some connecting devices
(room, keyboard, screen,
speaker, …)

14
Measuring intelligence
Discussion:

- OpenAI GPT-4 vs. Google Gemini, which one is more intelligent?

- From Academic research paper
- From Expert views (blogs, forums, communities, …)
- Your own experiences

15
Measuring intelligence
Benchmark in each domain

16
1.2. AI basic concepts
John McCarthy (1955): “the science and engineering of making
intelligent machines”

Encyclopedia Britannica: “AI is the ability of digital computers or

computer controlled robots to solve problems that are normally
associated with the higher intellectual processing capabilities of
humans”

K. R. Chwodhary (2020): Artificial Intelligence (AI) is a branch of

Computer Science, which is mainly concerned with automation of
Intelligent behavior
17
AI Definitions

Intelligence = Perceive + Analyze + React

Perceive: receive the input from environment

Analyze: think, predict, compute, … to find the solution
React: make decisions, make actions to the environment

K. R. Chwodhary (2020): Artificial Intelligence (AI) is a branch of

Computer Science, which is mainly concerned with automation of
Intelligent behavior
18
AI Scope
Strong AI (General AI): Do anything human can Weak AI (Narrow AI): accomplish specific tasks
do rather than possess full cognitive abilities like
the human brain
→ Robotics
→ Many AI applications in our daily lives
E.g. Doraemon, Iron Man, … not popular yet

Examples?

19
Goals of AI
The scientific goal of AI is to determine theories about knowledge representation,
learning, rule-based systems, and search that explain various sorts of intelligence.

The engineering goal of AI is to acquire ability in the machine so that it can solve
the real-life problems, regardless of the criteria for an abstract intelligence

20
Why AI winter?
Why rebirth?
How about the future?

Milestones in AI history
21
An example from machine translation evolution

From 1950s-1980s:
Rule-based approaches + Expert systems

Word-by-word translation based on a large dictionary

22
An example from machine translation evolution

From 1980s-2000s:
Statistical machine translation (SMT)

Parallel sentence pairs corpus, Hidden Markov Model,

Alignment, Statistical model, Language model

23
An example from machine translation evolution

From 2000s-now:
Neural machine translation (NMT)

Recurrent neural networks, Attention mechanism,

Transformers, Word2Vec

24
1.3. Subfields and related fields in AI

25
Sub-fields of AI
Signal Processing:

Sound, images, videos

Scientific measurements

Sensors signal

26
Subfields of AI
Natural Language Processing
Word Tokenization, Spell check and Correction, Information Retrieval,
Information Extraction,
Text Classification, Question-answering, Information summarization, Machine
Translation.
Computer Vision
Image Segmentation, Image Classification, Face Detection & Recognition, …
Video classification, Video Thumbnail Selection, Video Segmentation

27
Chatbot workflow

Try your own chatbot - Siri/Google Assistant

28
Subfields of AI
● Planning
● Expert Systems
● Fuzzy Systems
● NeuroScience

(Chwodhary - 1.7, pages 38)

29
Subfields of AI
A major technology driving the recent rise of AI, our focus of study in this course
Machine Learning: Field of study that gives computers the ability to learn without being
explicitly programmed (Arthur Samuel - 1959)
- Non-ML: we have to tell the machine what to do for each situation
- Double-click on a folder → Open it
- Type some address in the browser → Go to that website
- Tap on Check out button in the Shopee app → Lose money

→ Computers/Cellphones/Tablet/… don’t need to “think”, they just follow what they

were programmed before
Deep Learning: a class of Machine Learning techniques/algorithms that based on
artificial neural networks → simulate our brains processing

30
Subfields of AI
How about Data Science?

Extract
knowledge and
insight of data

31
1.4. Job opportunities in AI
Artificial intelligence (AI) engineer
Machine learning engineer
Data engineer
Data scientist
AI research scientist
Robotics engineer
Computer vision engineer
NLP engineer
https://www.coursera.org/articles/artificial-intelligence-jobs

32
1.4. Job opportunities in AI
Discussion:

- What kinds of job you are looking for in the field of AI

- What requirements (knowledge, skills, …) you need to get that job

33
Homework
Find AI communities and influential people on social network to follow

Install Anaconda, VS Code, Jupyter Notebook

Download ebooks

34