CPS 170: Artificial Intelligence

http://www.cs.duke.edu/courses/spring09/cps170/

Introduction

Instructor: Vincent Conitzer

Basic information about course

TuTh 4:25-5:40pm, LSRC D106

Text: Artificial Intelligence: A Modern Approach

Instructor: Vincent Conitzer
OH immediately after class (Tu or Th) or by appointment Ph.D. CMU 2006; third year at Duke Research on computational aspects of (micro)economics, game theory, systems with multiple intelligent agents

TA: Dmytro (Dima) Korzhyk

OH We 4-5pm, North Building 05, or by appointment 1st-year Ph.D. student at Duke Research on game theory and security

Prerequisites
Comfortable programming in language such as C (or C++) or Java Some knowledge of algorithmic concepts such as running times of algorithms

Ideally, some familiarity with probability (we will go over this from the beginning but we will cover the basics only briefly) Not scared of mathematics; ideally, some background in discrete mathematics, able to do simple mathematical proofs
If you have a nonstandard computer science background, talk to me first

Grading
Assignments: 35%
May discuss with another person (should acknowledge); writeup and code must be your own

Midterm exams: 30% Final exam: 30% Participation: 5%

What is artificial intelligence?

Popular conception driven by science ficition
Robots good at everything except emotions, empathy, appreciation of art, culture,
until later in the movie.

Perhaps more representative of human autism than of (current?) real robotics/AI

It is my belief that the existence of autism has contributed to [the theme of the intelligent but soulless automaton] in no small way. [Uta Frith, Autism] Current AI is also bad at lots of simpler stuff! There is a lot of AI work on thinking about what others are thinking

 A serious science.

Real AI

General-purpose AI like the robots of science fiction is incredibly hard

Human brain appears to have lots of special and general functions, integrated in some amazing way that we really do not understand at all (yet)

Special-purpose AI is more doable (nontrivial)

E.g., chess/poker playing programs, logistics planning, automated translation, voice recognition, web search, data mining, medical diagnosis, keeping a car on the road,

Definitions of AI
focus on action avoids philosophical issues such as is the system conscious etc.

if our system can be more rational than humans in some cases, why not?

Systems that think Systems that think like humans rationally Systems that act like humans Systems that act rationally

We will follow act rationally approach

Distinction may not be that important
acting rationally/like a human presumably requires (some sort of) thinking rationally/like a human, humans much more rational anyway in complex domains

Chinese room argument [Searle 1980]

image from http://www.unc.edu/~prinz/pictures/c-room.gif

Person who knows English but not Chinese sits in room Receives notes in Chinese Has systematic English rule book for how to write new Chinese characters based on input Chinese characters, returns his notes
Person=CPU, rule book=AI program, really also need lots of paper (storage)

Has no understanding of what they mean But from the outside, the room gives perfectly reasonable answers in Chinese! Searles argument: the room has no intelligence in it!

Turing Test
(Human) judge communicates with a human and a machine over text-only channel, Both human and machine try to act like a human,

Judge tries to tell which is which.

Numerous variants Loebner prize
image from http://en.wikipedia.org/wiki/Turing_test

Current programs nowhere close to passing this

http://www.jabberwacky.com/ http://turingtrade.org/

Turing Test on unsuspecting judges

It is possible to (temporarily) fool humans who do not realize they may be talking to a bot
ELIZA program [Weizenbaum 66] rephrases partners statements and questions (~psychotherapist)

Modern variant: AOLiza

[fragment from http://archives.cnn.com/2000/TECH/computing/08/29/aoliza.idg/]

five: just asking if richard can tell me about some good lobster diving spots thats it if this is a problem then forget iot i dont know whats going on here AOLiza: that is interesting. please continue. five: continue what just ask richard when he gets home about some lobster diving up hat way thats all if you got a problem with it then forget it,this isnt worht it i dont know whats been going on or what bev has told you but i havent done anything to anybody iam the one that got used and dumped AOLiza: tell me more about that. five: about what AOLiza: what do you think? five: therers nothing to tell she us3ed me then got tired of me said i wasnt her type anymore i was possesive and controlling AOLiza: perhaps i already know you were possesive and controlling. five: i was huh you nuts i gave her all the room she wanted never told nhewre what to do or where to go

Is Turing Test the right goal?

Aeronautical engineering texts do not define the goal of their field as making machines that fly so exactly like pigeons that they can fool even other pigeons. [Russell and Norvig]

Lessons from AI research

Clearly-defined tasks that we think require intelligence and education from humans tend to be doable for AI techniques
Playing chess, drawing logical inferences from clearly-stated facts, performing probability calculations in well-defined environments, Although, scalability can be a significant issue

Complex, messy, ambiguous tasks that are natural for humans (in some cases other animals) are much harder
Recognizing your grandmother in a crowd, drawing the right conclusion from an ungrammatical or ambiguous sentence, driving around the city,

Humans better at coming up with reasonably good solutions in complex environments

Humans better at adapting/self-evaluation/creativity (My usual strategy for chess is getting me into trouble against this person Why? What else can I do?)

Early history of AI
50s/60s: Early successes! AI can draw logical conclusions, prove some theorems, create simple plans Some initial work on neural networks Led to overhyping: researchers promised funding agencies spectacular progress, but started running into difficulties:
Ambiguity: highly funded translation programs (Russian to English) were good at syntactic manipulation but bad at disambiguation
The spirit is willing but the flesh is weak becomes The vodka is good but the meat is rotten

Scalability/complexity: early examples were very small, programs could not scale to bigger instances

Limitations of representations used

History of AI
70s, 80s: Creation of expert systems (systems specialized for one particular task based on experts knowledge), wide industry adoption

Again, overpromising
led to AI winter(s)
Funding cutbacks, bad reputation

Modern AI
More rigorous, scientific, formal/mathematical

Fewer grandiose promises

Divided into many subareas interested in particular aspects More directly connected to neighboring disciplines
Theoretical computer science, statistics, economics, operations research, biology, psychology/neuroscience, Often leads to question Is this really AI?

Some senior AI researchers are calling for reintegration of all these topics, return to more grandiose goals of AI

Some AI videos
Note: there is a lot of AI that is not quite this sexy but still very valuable!
E.g. logistics planning DARPA claims that savings from a single AI planning application during 1991 Persian Gulf crisis more than paid back for all of DARPAs investment in AI, ever. [Russell and
Norvig]

http://www.youtube.com/watch?v=1JJsBFiXGl0&feature=related http://www.youtube.com/watch?v=ICgL1OWsn58&feature=related

http://www.cs.utexas.edu/~kdresner/aim/video/fcfs-insanity.mov
http://www.youtube.com/watch?v=HacG_FWWPOw&feature=related http://videolectures.net/aaai07_littman_ai/ http://www.ai.sri.com/~nysmith/videos/SRI_AR-PA_AAAI08.avi http://www.youtube.com/watch?v=ScXX2bndGJc

This course
Focus on general AI techniques that have been useful in many applications
Will try to avoid application-specific techniques (still interesting and worthwhile!)

 Search

Topics (and examples)

Solving a Rubiks cube

Constraint satisfaction/optimization problems

Scheduling a given set of meetings (optimally)

Game playing
Playing chess

note overlap among topics

Logic, knowledge representation

Solving logic puzzles, proving theorems

Planning
Finding a schedule that will allow you to graduate (reasoning backwards from the goal)

Probability, decision theory, reasoning under uncertainty

Given some symptoms, what is the probability that a patient has a particular condition? How should we treat the patient?

(Time permitting) machine learning, reinforcement learning

Recognizing handwritten digits

AI at Duke
Ron Parr
Reasoning under uncertainty, reinforcement learning, robotics

Vince Conitzer
Systems with multiple, self-interested agents, game theory, economics

Carlo Tomasi
Computer vision, medical imaging

Alex Hartemink
Computational biology, machine learning, reasoning under uncertainty

Bruce Donald
Computational biology & chemistry

Sayan Mukherjee
Statistics

Duke Robotics, Intelligence, and Vision (DRIV) seminar (=AI seminar)