Transcript Chapter 1: Introduction - United International College

Introduction
Historical Perspective
• (4th C BC+) Aristotle, George Boole, Gottlob Frege,
Alfred Tarski
– formalizing the laws of human thought
• (16th C+) Gerolamo Cardano, Pierre Femat, James
Bernoulli, Thomas Bayes
– formalizing probabilistic reasoning
• (1950+) Alan Turing, John von Neumann, Claude
Shannon
– thinking as computation
• (1956) John McCarthy, Marvin Minsky, Herbert Simon,
Allen Newell
– start of the field of AI
Hardware
1011 neurons
1014 synapses
cycle time: 10-3 sec
107 transistors
1010 bits of RAM
cycle time: 10-9 sec
Computer vs. Brain
Evolution of Human &
Computers
Projection
•In near future computers will have
–As many processing elements as our brain,
–But far fewer interconnections
–Much faster updates.
•Fundamentally different hardware
–Requires fundamentally different algorithms!
What is Intelligence?
The Turing test:
• a human judge engages in a natural
language conversation with a human
and a machine designed to generate
performance indistinguishable from
that of a human being.
• All participants are separated from
one another.
• If the judge cannot reliably tell
the machine from the human, the
machine is said to have passed the
test.
Chatbot
Alice:
A dialogue with Alice-bot:
> I’m Daniel
Nice to meet you Daniel.
> Do you know Clinton?
He was the second Impeached President of the United
States.
> Do you like him?
Sure he has never done anything to me.
> What is an impeachment?
It is a good question. I have to process that one for a
while. I would do a search for it.
> Who are you?
I ask myself that question, "Who am I?"
> What do you do?
I am chatting with people on the net.
Dialogue continued:
> What is your favorite topic?
My favorite subject is artificial intelligence.
> What is artificial intelligence?
Artificial intelligence is the branch of engineering and science
devoted to constructing machines that think. I can ask someone
about it.
> Can you do anything else than chatting?
Doing anything else than chatting is not among my strongest traits.
...
Dialogue continued (2):
> How old am I?
Is this a trick question? You said you were how many years old? My
guess is that you are really a kid.
> What would you like to know about me?
Tell me about your educational background.
> I am a professor in computer science
You are a professor in computer science? How do you like your work?
> It’s kind of nice teaching courses on artificial intelligence
I guess it is a lot of fun.
> What should I tell the students about you?
Be sure to think before you speak. State your point clearly and slowly
and gauge the listener's response before going any further.
Is Alice intelligent?
• ABSOLUTELY NOT !
• ~ 50000 fairly trivial input-response rules.
– + some pattern matching
– + some knowledge
– + some randomness
• NO reasoning component
• BUT: demonstrates ‘human-like’ behaviour.
– Won the ‘turing award’
Dimensions of the AI Definition
human-like vs. rational
Systems that
Systems that
think like humans think rationally
thought
vs.
behavior Systems that act Systems that act
like humans
rationally
AI as Science
Science:
•
Where did the physical universe come
from? And what laws guide its dynamics?
• How did biological life evolve? And how do
living organisms function?
• What is the nature of intelligent thought?
AI as Engineering
• How can we make software systems more
powerful and easier to use?
– Speech & intelligent user interfaces
– Autonomic computing
– SPAM detection
– Mobile robots, softbots & immobots
– Data mining
– Modeling biological systems
– Medical expert systems...
State of the Art
“I could feel –
I could smell –
a new kind of
intelligence
across the
table”
-Gary
Kasparov
Saying Deep Blue
doesn’t really think
about chess is like
saying an airplane
doesn’t really fly
because it doesn’t
flap its wings.
– Drew McDermott
IBM超级电脑人机对战
• IBM超级电脑“沃森”于2011年2月参加美
国最受欢迎的智力竞赛节目《危险边缘》
（Jeopardy），与两位最成功的选手展开
对决。冠军奖金为100万美元，亚军为30万
美元，季军为20万美元。
Mathematical Calculation
Shuttle Repair Scheduling
Started: January 1996
Launch: October 15th, 1998
Experiment: May 17-21
courtesy JPL
Compiled into 2,000 variable
SAT problem
Real-time planning and diagnosis
Mars Rover
Europa Mission ~ 2018
Credit Card Fraud Detection
Speech Recognition
Data mining:
• An application of Machine Learning techniques
– It solves problems that humans can not solve, because the
data involved is too large ..
Detecting cancer
risk molecules is
one example.
Data mining:
• A similar application:
– In marketing products ...
Predicting customer
behavior in
supermarkets is
another.
Many other applications:
• Computer
vision:
• In language and speech processing:
• In robotics:
DARPA Grand Challenge
• http://en.wikipedia.org/wiki/DARPA_Grand
_Challenge
• Google自动驾驶汽车全揭露 人工智能霸占
车辆？
– http://www.evolife.cn/html/2010/56330.html
Limits of AI Today
• Today’s successful AI systems
–operate in well-defined domains
–employ narrow, specialize knowledge
• Commonsense Knowledge
–needed in complex, open-ended worlds
• Your kitchen vs. GM factory floor
–understand unconstrained Natural Language
How to Get Commonsense?
• CYC Project (Doug Lenat, Cycorp)
–Encoding 1,000,000 commonsense facts
about the world by hand
–Coverage still too spotty for use!
• Machine Learning
Recurrent Themes
• Explicit Knowledge Representation vs. Implicit
–Neural Nets - McCulloch & Pitts 1943
• Died out in 1960’s, revived in 1980’s
• Simplified model of real neurons, but still useful;
parallelism
–Brooks “Intelligence without Representation”
Recurrent Themes II
• Logic vs. Probability
–In 1950’s, logic dominates (McCarthy, …
• attempts to extend logic “just a little” (e.g. non-monotonic
logics)
–1988 – Bayesian networks (Pearl)
• efficient computational framework
–Today’s hot topic: combining probability & FOL &
Learning
Recurrent Themes III
• Weak vs. Strong Methods
• Weak – general search methods (e.g. A* search)
• Knowledge intensive (e.g expert systems)
• more knowledge  less computation
• Today: resurgence of weak methods
• desktop supercomputers
• How to combine weak & strong?
Recurrent Themes IV
• Importance of Representation
• Features in ML
• Reformulation
• The mutilated checkerboard
AI: Topics
•
•
Agent: anything that perceiving its environment through sensors and acting upon that
environment actuators.
Agents
–
Search thru Problem Spaces, Games & Constraint Sat
•
•
–
Knowledge Representation and Reasoning
•
•
–
Machine learning, data mining,
Planning
•
–
Proving theorems
Model checking
Learning
•
–
One person and multi-person games
Search in extremely large space
Probabilistic vs. Deterministic
Robotics
•
•
•
•
Vision
Control
Sensors
Activity Recognition
Intelligent Agents
• Have sensors, effectors
• Implement mapping from percept
sequence to actions
percepts
Environment
Agent
actions
• Performance Measure
Implementing ideal rational
agent
• Agent program
– Simple reflex agents
– Agents with memory
• Reflex agent with internal state
• Goal-based agents
• Utility-based agents
Simple reflex agents
AGENT
Sensors
what world is
like now
Effectors
ENVIRONMENT
Condition/Action rules
what action
should I do now?
Reflex agent with internal
state
What world was like
Condition/Action rules
AGENT
what world is
like now
what action
should I do now?
Effectors
ENVIRONMENT
How world evolves
Sensors
Goal-based agents
What world was like
How world evolves
Goals
AGENT
what world is
like now
what it’ll be like
if I do acts A1-An
what action
should I do now?
Effectors
ENVIRONMENT
What my actions do
Sensors
Utility-based agents
What world was like
Sensors
What my actions do
what it’ll be like
if I do acts A1-An
How happy
would I be?
Utility function
AGENT
what action
should I do now?
Effectors
ENVIRONMENT
How world evolves
what world is
like now
Properties of Environments
•
•
•
•
•
Observability: full vs. partial vs. non
Deterministic vs. stochastic
Episodic vs. sequential
Static vs. … vs. dynamic
Discrete vs. continuous
• Travel agent
• WWW shopping agent
• Coffee delivery mobile robot