CS 5/7320

Artificial
Intelligence

Introduction

AIMA Chapter 1

Slides by Michael Hahsler

based on slides by Svetlana
Lazepnik with figures and cover
art from the AIMA textbook.

This work is licensed under a Creative Commons

Attribution-ShareAlike 4.0 International License.
What History AI
?
is AI? of AI Today

AI Ethics & Safety

What is AI?

ASIMO (Advanced Step in Innovative Mobility) is

a humanoid robot created by Honda in 2000
What is it the Goal of AI?
“Have machines solve problems that a challenging for humans.”

• Narrow AI focuses on intelligent agents to solve a specific subproblem.

• An artificial general intelligence (AGI) is a hypothetical intelligent agent

which can understand or learn any intellectual task that human beings
or other animals can. [Wikipedia entry on AGI]

How do we achieve this?

Create an agent that

thinks like acts like a thinks acts

a human? human? rationally? rationally?
Thinking Like a Human
The brain as an How to understand
information processing cognition as a AI consciousness
machine. computational process?
• Requires scientific • Introspection: try to • What does it mean
theories of how the think about how we that a machine is
brain works. think. conscient/sentient?
• Predict the behavior of • How can we tell?
human subjects.
Note: The brain does • Image the brain,
not work like artificial examine neurological (What do we do?)
neural networks from data
ML!

Cognitive Sciences
Acting Like a Human
• Alan Turing (1950) "Computing machinery and intelligence“
• The Turing Test tries to define what acting like a human means

• What capabilities would a computer need to have to pass the Turing Test?
• Natural language processing
• Knowledge representation
• Automated reasoning
• Machine learning

• Turing predicted that by the year 2000, machines would be able to fool 30% of human judges for
five minutes.
• ChatGPT in 2023 is probably doing a lot better than that!
Turing Test: Criticism
What are some potential problems with
the Turing Test? Chinese Room Argument
• Some human behavior is not intelligent.
• Some intelligent behavior may not be human.
• Human observers may be easy to fool.
• A lot depends on expectations.
• Anthropomorphic fallacy (humans tend to
humanize things)
• Imitate intelligence without intelligence.
E.g., the early chatbots ELIZA (1964)
simulates a conversation using pattern
matching.
Thought experiment by John
Searle (1980): Imitate
intelligence using rules.
Is passing the Turing test a good
scientific goal? What about modern chatbots
• Engineering perspective: Not a good way to like ChatGPT?
solve practical problems.
• We can create useful intelligent agents without
trying to imitate humans.
Thinking Rationally
• Idealized or “right” way of thinking.
• Logic: Patterns of argument that always yield correct conclusions when supplied
with correct premises
• “Socrates is a man; all men are mortal; therefore, Socrates is mortal.”
• Beginning with Aristotle (385 BC), philosophers and mathematicians have
attempted to formalize the rules of logical thought.
• Logic-based approach to AI: Describe problem in formal logical notation and
apply general deduction procedures to solve it.
• Problems with the logic-based approach to AI
• Describing real-world problems and knowledge in logical notation is hard.
• Computational complexity of finding the solution.
• A lot of intelligent or “rational” behavior in an uncertain world cannot be
defined by simple rules.

What about the logical implication

𝑠𝑡𝑢𝑑𝑦 ℎ𝑎𝑟𝑑 ֜ 𝐴 𝑖𝑛 𝐴𝐼

Should it be
𝑠𝑡𝑢𝑑𝑦 ℎ𝑎𝑟𝑑 𝐴𝑁𝐷 𝑏𝑒 𝑙𝑢𝑐𝑘𝑦 ֜ 𝐴 𝑖𝑛 𝐴𝐼
Acting Rationally

A rational agent acts to achieve the best expected

outcome:

• Goals are application-dependent and are expressed in

terms of the utility of outcomes.
• Being rational means acting to maximizing your expected
utility. Expectation means that different outcomes are
possible (probabilities).
• In practice, utility optimization is subject to the agent’s
knowledge and computational constraints (bounded
rationality or bounded optimality).
Acting Rationally

Advantages of the “expected utility maximization”

formulation

• Generality: an optimization that goes beyond explicit

reasoning with rules.
• Practicality: can be adapted to many real-world problems.
• Amenable to good scientific and engineering
methodology including simulation and experimentation.
• Only concerns the decisions/actions that are made, not
the cognitive process behind them. Avoids philosophy and
psychology in favor of a clearly defined objective.
What type of AI do we cover in this course?

Create a narrow AI agent that

thinks like acts like a thinks acts
a human? human? rationally? rationally.

That is, use machines to solve a specific hard

problem that traditionally would have been
thought to require human intelligence.
What are the Components of an
Intelligent Agent?
Intelligent agents need to
• Communicate with the Developer
environment
• Represent knowledge,
reason and plan to
achieve a desired
outcome

Optional
• Learn
Agent interacting with the environment
[Artificial Intelligence: A Modern Approach, Editions 1-3]
Machine Learning vs. Artificial Intelligence

AI
Designing an intelligent agent

• Vision
•
•
Sensors
Motion ML
• NLP Learning from examples
• Knowledge instead of being programmed
• Supervised learning
representation
• Unsupervised learning
• Planning • RL ML
• Goals
Deep
Learning ML

ML

Agent interacting with the environment

[Artificial Intelligence: A Modern Approach, Editions 1-3]
The History
of AI
 1974-1980
First AI Winter

1987-1993
Second AI Winter

1989: Universal
approximation theorem
for neural networks.
2010

2022
2017
2015

Transformer Generative AI
architecture models:
and large • DALL-E
language • ChatGPT, Bard
models LLMs • …
Deep Learning Revolution
(learning layered artificial neural
networks) starts fueled by NVIDIA
GPUs. enables leaps in image
Now Google processing and speech recognition.

Source: https://qbi.uq.edu.au/brain/intelligent-machines/history-artificial-intelligence + additions

What accounts for
recent successes in
AI?

• Faster computers and

specialized hardware
(GPUs).
• Lots of data (the Internet,
text, sensors) and storage
(cloud)
• Dominance of machine
learning.
• New optimization methods
(deep learning).
AI is harder than originally
thought

Herbert Simon, 1957

“It is not my aim to surprise or shock you--- but …

there are now in the world machines that think,
that learn and that create. Moreover, their
ability to do these things is going to increase
rapidly until---in a visible future---the range of
problems they can handle will be coextensive
with the range to which human mind has been
applied. More precisely: within 10 years a
computer would be chess champion, and an
important new mathematical theorem would
be proved by a computer.”

Simon’s prediction came true --- but 40 years

later instead of 10
From Blocks World to Modern Object Recognition
Roberts (1963) Now

This is a lot harder!

But we can do it now….
 “Moravec’s Paradox”
Hans Moravec (1988): “It is comparatively
easy to make computers exhibit adult level
performance on intelligence tests or
playing checkers, and difficult or
impossible to give them the skills of a
one-year-old when it comes to perception
and mobility.”

A teenager can learn how to drive in a few

hours with very little input, but we still
have no truly self-driving car.

https://www.newsweek.com/googles-new-two-legged-robot-
future-warfare-429831
The AI Effect:
AI gets no respect?

• As soon as a machine gets good at performing some

task, the task is no longer considered to require much
intelligence
• Calculating ability used to be prized – not anymore
• Chess was thought to require high intelligence
• Now, massively parallel computers essentially use
brute force search to beat grand masters
• Learning once thought uniquely human
• Ada Lovelace (1842): “The Analytical Engine has no
pretensions to originate anything. It can do
whatever we know how to order it to perform.”
• Now machine learning is a well-developed
discipline
• Similar picture with animal intelligence… “Even a
monkey can do this!”
AI Today
Vision and Image
Processing

• OCR: read license plates,

handwriting recognition
(e.g., mail sorting).
• Face detection: now
standard for smart phone DALL-E
cameras.
• Vehicle safety systems
• Visual search
• Image generation

All these technologies can

now operate now at
superhuman performance.
 Natural Language
Processing
• Text-to-speech
• Speech-to-text to detect voice
commands
• Machine translation
• Text generation (Q/A systems)

All these technologies can operate now with

close to or even superhuman performance.

Humans use language to reason. Does that

mean AI that can create good language can
reason?
Language understanding is still elusive!
Robotics

• Mars rovers
• Autonomous vehicles
• DARPA Grand
Challenge
• Google self-driving
cars
• Autonomous
helicopters and drones
• Robot soccer
• RoboCup
• Personal robotics
• Humanoid robots
• Robotic pets
• Personal
assistants?
Question Answering: IBM Watson
• Listens to spoken • http://www.research.ibm.com/deepqa/
language. • NY Times article
• Speaks. • Trivia demo
• Finds questions to • YouTube video
factual answers. • IBM Watson wins on Jeopardy (February
2011)
Math, Games and Puzzles
• 1996: A computer program written by researchers at Argonne
National Laboratory proved a mathematical conjecture (Robbins
conjecture) unsolved for decades
• NY Times story: “[The proof] would have been called creative if
a human had thought of it”
• 1996/97: IBM’s Deep Blue defeated the reigning world chess
champion Garry Kasparov in 1997
• 1996: Kasparov Beats Deep Blue
“I could feel --- I could smell --- a new kind
of intelligence across the table.”
• 1997: Deep Blue Beats Kasparov
“Deep Blue hasn't proven anything.”
• 2007: Checkers was “solved” --- a computer system that never loses
was developed. Science article
• 2017+: AlphaZero learns chess, shogi and go by playing itself. Science
article
• 2019: MuZero learns to play Atari computer games.

AI exhibits superhuman performance on almost all games.

Logistics, Scheduling and Planning

During the 1991 Gulf War, US

forces deployed an AI logistics In 2004, NASA introduced the
planning and scheduling program MAPGEN system to plan the daily
that involved up to 50,000 operations for the Mars
vehicles, cargo, and people Exploration Rovers

May 1999 2018

1991 2004

NASA’s Remote Agent software Optimized Prime: How AI And

operated the Deep Space 1 Anticipation Power Amazon's 1-
spacecraft during two experiments Hour Deliveries
in May 1999
 • Search engines
• Chat Agents / Automated helpdesks
• Recommendation systems
Information • Spam filtering
agents • Medical diagnosis systems
• Fraud detection
• Automated stock trading
Self-driving Cars
and Safety

SAE Automation Levels

• Level 1 - Driver Assistance (“hands on”)
• Level 2 - Partial Automation (“hands off”)
• Level 3 - Conditional Automation
• Level 4 - High Automation
• Level 5 - Full Automation (“steering wheel
optional”)

Components
• Sensing
• Maps
• Path planning
• Controlling the vehicle

Why is this so hard?

Large Language Models (LLMs)

…
Source: https://chat.openai.com/
AI Ethics & Safety
A new Frontier for Fairness and Freedom
 European
Union
Has regulations
since 2016 included
in the General Data
Protection
Regulation (GDPR)
Art. 22 GDPR – Automated
individual decision-making,
including

California’s CCPA
was not modeled
after the GDPR
 Australia: AI Ethics Framework for Industry
A set of voluntary AI Ethics Principles (2019)

https://consult.industry.gov.au/strategic-policy/artificial-intelligence-ethics-
framework/supporting_documents/ArtificialIntelligenceethicsframeworkdiscussionpaper.pdf
In the US
 European Union
Study (2019)

https://www.europarl.europa.eu/thinktank/en/document.html?reference=EPRS_STU(2019)624262
 https://www.ai.gov

US National AI Initiative Act of 2020

“The National AI Initiative Act of 2020 became law on January 1, 2021, providing for a coordinated program
across the entire Federal government to accelerate AI research and application for the Nation’s economic
prosperity and national security. The mission of the National AI Initiative is to ensure continued U.S. leadership
in AI research and development, lead the world in the development and use of trustworthy AI in the public and
private sectors, and prepare the present and future U.S. workforce for the integration of AI systems across all
sectors of the economy and society.”
May also in the future provide regulations
Source: https://ai.google/static/documents/recommendations-for-regulating-ai.pdf, 2021
Algorithmic Bias and Fairness
“Algorithmic bias describes systematic and repeatable errors in a
computer system that create unfair outcomes, such as privileging one
arbitrary group of users over others” Wikipedia

• Social and institutional norms influence design and

Pre-existing training data choices.

bias • For example: Evaluate job applicants for a job which is

historically almost exclusively held by males.

• Limitations of a program or computational power.

Technical bias • For example: instead of a random sample, the
program uses the first n data points.

• Use of algorithms for new data without checking for

Emergent bias bias (e.g., existing correlations in the data).
• Use of an algorithm for an unanticipated application.
 AI Safety
“Prevent accidents, misuse, or other
harmful consequences of AI.”

• Robustness: Black swan vs.

adversarial robustness
• Monitoring AI
• What about liability?
• Goal/reward alignment
• Reward hacking
• AGI and instrumental convergence
Alignment?
Goals
User
Objectives
and
Rules
Intelligent
Agent
Alignment? Action
program
train

Goals
select
Data
Developers / owners

Credit: Terminator 3: Rise of the Machines. Warner Bros.

Outlook
AI is a technology that is on
the verge of significant
leaps…

• New technologies always

had a profound
impacted on the way we
live and work (e.g.,
electricity, the internet,
mobile communication).
• We can expect
unprecedented gains in
productivity from better
narrow AI.

• This course will

introduce simple
techniques to create
intelligent agents.