Transcript What is AI? - UB Computer Science and Engineering

What is AI?
Copyright © 2008 by Helene G. Kershner
What is AI?
AI = Artificial Intelligence
Artificial Intelligence (AI) is the area of computer
science focusing on creating machines that can
engage on behaviors that humans consider
intelligent. http://library.thinkquest.org/2705/
Copyright © 2008 by Helene G. Kershner
What is AI?

Turing Test

A machine would be considered intelligent it could not be
distinguished from a person during a “blind” comparison
test. Turing Test



A person (“judge”) would ask questions via a computer
terminal of two “entities.” One entity would be a human, the
other a machine/computer.
“If the judge regularly failed to correctly distinguish the
computer from the human, then the computer was said to have
passed the test.” © István S. N. Berkeley Ph.D. 1997.
“The Turing test is a one-sided test. A machine that
passes the test should certainly be considered intelligent,
but a machine could still be considered intelligent
without knowing enough about humans to imitate a
human.” http://www-formal.stanford.edu/jmc/whatisai/node1.html
Copyright © 2009 by Helene G. Kershner
What is AI? – John McCarthy

Computer Scientist & Cognitive Scientist


A cognitive scientist studies the nature of intelligence
Combines psychology, philosophy, neuroscience, linguistics,
anthropology, computer science, sociology and biology
http://en.wikipedia.org/wiki/Cognitive_science



He coined the word Artificial Intelligence
Received the Turing Award in 1971 for his work in AI.
Turing award is given by ACM to “an individual
selected for contributions of a technical nature made
to the computing community. The contributions
should be of lasting and major technical importance
to the computer field". Often recognized as the
“Nobel Prize of computing” http://en.wikipedia.org/wiki/Turing_Award
Copyright © 2009 by Helene G. Kershner
What is AI? -- John McCarthy
What is artificial intelligence?
AI “it is the science and engineering of making intelligent
machines, especially intelligent computer programs.

It is related to the similar task of using computers to understand
human intelligence, but

AI does not have to confine itself to methods that are
biologically observable.“http://www-formal.stanford.edu/jmc/whatisai/node1.html
So if it looks like it is intelligent, and acts like it is intelligent even if
it does so in ways different from human it could be considered
intelligent.
Sort of “I’ll know it when I see it.”

Copyright © 2009 by Helene G. Kershner
What is AI? – John McCarthy
Can we even really define intelligence?


Intelligence is the “computational part of the ability to achieve
goals in the world.
Varying kinds and degrees of intelligence occur in people, many
animals and some machines.”
But what is intelligence really?
Is there a definition of intelligence that doesn’t depend
on relating it to human intelligence?



“ Not yet.
The problem is that we cannot yet characterize in general what
kinds of computational procedures we want to call intelligent.
We understand some of the mechanisms of intelligence and not
others.”http://www-formal.stanford.edu/jmc/whatisai/node1.html
Copyright © 2009 by Helene G. Kershner
What is AI? – John McCarthy
“Isn't AI about simulating human intelligence?




Sometimes but not always or even usually.
On the one hand, we can learn something about how to make
machines solve problems by observing other people or just by
observing our own methods.
On the other hand, most work in AI involves studying the
problems the world presents to intelligence rather than studying
people or animals.
AI researchers are free to use methods that are not observed in
people or that involve much more computing than people can
do. “
http://www-formal.stanford.edu/jmc/whatisai/node1.html
Copyright © 2009 by Helene G. Kershner
What is AI? -- John McCarthy
“When did AI research start?
 After WWII, a number of people independently
started to work on intelligent machines.
 The English mathematician Alan Turing may have
been the first. He gave a lecture on it in 1947.
 He also may have been the first to decide that AI was
best researched by programming computers rather
than by building machines.
 By the late 1950s, there were many researchers on
AI, and most of them were basing their work on
programming computers.
http://www-formal.stanford.edu/jmc/whatisai/node1.html
Copyright © 2009 by Helene G. Kershner
What is AI?



In the 1950s, scientists thought artificially intelligent
machines were just around the corner.
Clearly R2D2 & C3PO have yet to arrive.
Computers are very good at following rules


These rules can include “rules of thumb” called heuristics
Good at playing games, Chess, GO



So good in fact that in 1997, Deep Blue a computer chess
program won a series of chess games against the human world
champion of chess Gary Kasparov
Largely used brute force
But, Games are not the only things that have rules!
Copyright © 2009 by Helene G. Kershner
What is AI?

What else has rules?



Grammar
Spelling
Some Medical diagnosis






Drug interactions
Blood tests
Certain common illnesses
Spam filters
Web searches
Programs to catch “cheaters” in school such as “TurnItIn”
and “Moss”
Copyright © 2008 by Helene G. Kershner
What is AI? – Expert Systems

Expert Systems
 Software that attempts to duplicate the ways in
which professionals make decisions are called
expert systems.
 An expert is a specialist in a specific, often narrow
field who has gained an incredible amount of
knowledge about this narrow field



A surgeon who specializes in sports medicine
National expert in the US Tax Code
A group of oncologists specializing in drug side effects
Copyright © 2008 by Helene G. Kershner
Expert Systems








EasyDiagnosis – provides online analysis of important medical
symptoms in a user-friendly format. http://easydiagnosis.com/
CADUCEUS (expert system) - Blood-borne infectious bacteria
Dendral – Helps organic chemists identify unknown molecules
Dipmeter Advisor - Analysis of data gathered during oil exploration
Mycin - Diagnose infectious blood diseases and recommend antibiotics
(by Stanford University)
XCON formerly R1 (for eXpert CONfigurer)- Order processing
SHINE Real-time Expert System – An expert system for writing expert
systems developed by NASA
STD Wizard - Expert system for recommending medical screening tests

http://www.stdwizard.org/
*http://en.wikipedia.org/wiki/Expert_system
Copyright © 2009 by Helene G. Kershner
AI – Expert System
Program
Knowledge
base of rules
Inference
Natural
Expert
Language
Interface
Engine
Database
of facts
User
http://www.cs.cf.ac.uk/Dave/AI1/mycin.html
Copyright © 2009 by Helene G. Kershner
AI – Expert System

Example from Mycin




When the system is in Consultation mode the user interacts
with the system by entering data in English.
Based on the data provided the inference engine
(IF/Then/Else rules) derives an answer to the question
posed by the user.
The role of the inference engine is to search for facts that
match the conditions set up by the questions and answers.
The kinds of questions and answers that can be used are
limited to the realm of the system.
http://www.cs.cf.ac.uk/Dave/AI1/mycin.html
Copyright © 2009 by Helene G. Kershner
AI – Expert System

Mycin – a sample
System: What is the patient's name?
User: George Washington
System: What is the patient's age?
User: 254 years
System: Is the patient male or female?
User: Male
System: What are the symptoms?
User: Headache dizziness upset stomach high fever
System: When did these symptoms first appear?
User: December 13,1799
System: I conclude that the patient has a viral infection , my
recommendation is that he should take two aspirin, drink plenty of
fluid, get lots of rest and call me in the morning.
Fact: George Washington died the next day of a viral infection,
no antibiotics
Copyright © 2009 by Helene G. Kershner
What is AI?

What about common sense?



Computers are TERRIBLE at this!
Humans are awesome at this.
Common sense requires “rules of thumb” (Heuristics)
for ordinary things.





Talking and chewing gum at the same time
What is time?
Where we are in “space” in relation to other objects
The nature of physical object
 How not to crush a light bulb
 How objects interact
How people interact
Copyright © 2008 by Helene G. Kershner
What is AI? – Common Sense Reasoning

CYC and Doug Lenat
 EnCYClopedia of everyday rules of how the world
works
 Applications to everything





Military
Medical
Transportation system
 Air traffic control
Incredibly complex and difficult
Ongoing project by CYCorp
Copyright © 2008 by Helene G. Kershner
What is AI?
“Does AI aim to put the human mind into the
computer?
 Some researchers say they have that objective.
 But maybe they are using the phrase
metaphorically.
 The human mind has a lot of peculiarities, and
I'm not sure anyone is serious about imitating all
of them.” http://www-formal.stanford.edu/jmc/whatisai/node1.html
Copyright © 2009 by Helene G. Kershner
AI – Why is this so Hard?
Human recognition software and artificial intelligence
http://www.youtube.com/watch?v=OiqJ-lpnD-U
Does AI mean we mimic how humans do things OR is AI a
computer duplicating a task humans do only doing it using a
different technique?

Robot vacuum




AI as science


vacuums the floor
doesn’t perform the task in the way a human would do it
does it matter?
Involves -- Cognitive Science – how the human brain works
AI as Engineering –

solves a problem, but may not use a human solution and doesn’t
care.
Copyright © 2009 by Helene G. Kershner
AI – Why is this so Hard?

Some AI scientists are trying to understand
the “nature” of intelligence.


Can be viewed as understanding the science of
intelligence
Other AI scientists are trying to “mimic”
intelligence.

Can be viewed as engineering or designing a
machine that does what humans do but probably in
a different way.
Copyright © 2009 by Helene G. Kershner