Artificial Intelligence: Human vs. Machine Professor Adam Anthony CMSC 180 Fall 2010 Slides Provided by and Adapted From: Dr. Marie desJardins

BEFORE WE START!  DON’T FORGET ABOUT MINI-PROJECT 1!  Implementation: Make sure that whatever you submit works, even if it is incomplete!  Test Report: 2-3 pages verifying that you tested each feature of your program  How you tested it  How that test confirmed it was working correctly 2

3 Memory is at the Core (Literally)  Memory is at the core of our being (and a computer’s) ...but our memories look very different! The first magnetic core memory [ thebrain.mcgill.ca

4 Overview  What is AI? (and why is it so cool?)  AI: Past and Present  History of AI  AI Today  Computational vs. Biological Memory  The Skeptics Speak

What is AI?

6 AI: A Vision  Could an intelligent agent living on your home computer manage your , coordinate your work and social activities, help plan your vacations…… even watch your house while you take those well planned vacations?

7 RR epresent and store knowledge RR etrieve and reason about knowledge BB ehave intelligently in complex environments DD evelop interesting and useful applications II nteract with people, agents, and the environment Main Goals of AI

8 Computer Science & Engineering AI Mathematics Cognitive Science Philosophy PsychologyLinguistics Biology Economics Foundations of AI

9 Big Questions  Can machines think?  If so, how?  If not, why not?  What does this say about human beings?  What does this say about the mind?  And if we can make machines think, should we?

Architecture of an Intelligent Agent  An Intelligent Agent has:  Sensors for inputting data  Actuators for performing actions in the “world” Robotic tools in the real world Software routines in a virtual world  A level of Response: Reflex-based: only does exactly what it’s told (remote-controlled) Knowledge-based: able to follow semi-complex instructions  Baseball throwing example Goal-based: No instructions, just a goal to achieve  Makes its own instruction set  What types of things can be considered agents???  Skip to Philosophy Section Skip to Philosophy Section 10

11 Why AI?  Engineering: To get machines to do a wider variety of useful things  e.g., understand spoken natural language, recognize individual people in visual scenes, find the best travel plan for your vacation, etc.  Cognitive Science: As a way to understand how natural minds and mental phenomena work  e.g., memory, learning, language, visual perception, etc.  Philosophy: As a way to explore some basic and interesting (and important) philosophical questions  e.g., the mind body problem, what is consciousness, etc.

History of AI

13 History

14  ELIZA: A program that simulated a psychotherapist interacting with a patient and successfully passed the Turing Test  Implemented at MIT during by Joel Weizenbaum  First script was DOCTOR  Simple pattern-matching  Each pattern had an associated reply which might include bits of the input (after simple transformations (my  your)  Weizenbaum was shocked at reactions  Psychiatrists thought it had potential  People unequivocally anthropomorphized Let me tell you my problems...  Many thought it solved the natural language problem!! Early AI: Eliza

15 Eliza Speaks  I am the psychotherapist. Please, describe your problems.  i'm having a hard time with my students  Why do you say that?  well, they keep asking me for money  Maybe your plans have something to do with this.  what plans?  Is it because of your life that you say what plans?  No, it's because I didn't know what you were talking about  Is it because you didn't know what I was talking about that you came to me?  no, it's because i wanted to see if you could pass the turing test  Does it bother you that you wanted to see if I could really pass the turing test?  …..  How do you reconcile problems at school?  This is a waste of time. Good bye  My secretary will send you a bill.

16 Early AI: Parry  Kenneth Colby modeled a paranoid using the same techniques as Eliza, around 1968  PARRY has basic emotions. If it gets angry, its replies become more hostile  In a Turing test, expert psychiatrists were unable to distinguish PARRY’s ramblings from those of real paranoids

17 PARRY Speaks  Good afternoon. My name is Doctor Jones.  Pleased to meet you.  How are you?  All right I guess.  Could you tell me your name?  Pat Smith.  I though you might have some problems.  I've been upset lately.  By what?  People get on my nerves sometimes.  Tell me more.  Do you know anything about bookies? ...

18 P: I don't understand your motives. E: Why don't you understand my motives? P: Shouldn't I be? E: Why do you say that just now? P: That's just the way it is. E: Suppose that's the way it were not. P: So? E: I am not sure I understand you fully. P: You harp on the same questions too much. E: What does that similarity suggest to you? P: I don't get you. E: Do you wish that you did get I?

19 Turing Test  Three rooms contain a person, a computer, and an interrogator  The interrogator can communicate with the other two by “teleprinter” (or, say, AIM)  The interrogator tries to determine which is the person and which is the machine  The machine tries to fool the interrogator into believing that it is the person  If the machine succeeds, then we conclude that the machine can think

20 The Loebner Contest  A modern version of the Turing Test, held annually, with a $100,000 cash prize   Participants include a set of humans, a set of computers, and a set of judges  Scoring  Rank from least human to most human  Highest median rank wins $2000  If better than a human, win $100,000 (Nobody yet…)  2008 winner: ElbotElbot

21 What’s Easy and What’s Hard?  It’s been easier to mechanize many of the high-level tasks we usually associate with “intelligence” in people  e.g., symbolic integration, proving theorems, playing chess, medical diagnosis  It’s been very hard to mechanize tasks that lots of animals can do  walking around without running into things  catching prey and avoiding predators  interpreting complex sensory information (e.g., visual, aural, …)  modeling the internal states of other animals from their behavior  working as a team (e.g., with pack animals)  Is there a fundamental difference between the two categories?

AI Today

23 Who Does AI?  Academic researchers (perhaps the most Ph.D.-generating area of computer science in recent years)  Some of the top AI schools: CMU, Stanford, Berkeley, MIT, UIUC, UMd, U Alberta, UT Austin, UMBC  Government and private research labs  NASA, NRL, NIST, IBM, AT&T, SRI, ISI, MERL,...  Lots of companies!

24  A sample from the 2008 International Conference on Innovative Applications of AI:  Event management (for Olympic equestrian competition)  Language and culture instruction  Public school choice (for parents)  Turbulence prediction (for air traffic safety)  Heart wall abnormality diagnosis  Epilepsy treatment planning  Personalization of telecommunications services  Earth observation flight planning (for science data)  Crop selection (for optimal soil planning) Applications

25 Here are some example applications:  Computer vision: face recognition from a large set  Robotics: autonomous (mostly) automobile  Natural language processing: simple machine translation  Expert systems: medical diagnosis in a narrow domain  Spoken language systems: ~2000 word continuous speech  Planning and scheduling: Hubble Telescope experiments  Learning: text categorization into ~1000 topics  User modeling: Bayesian reasoning in Windows help (the infamous paper clip…)  Games: Grand Master level in chess (world champion), checkers, backgammon, etc. Breaking news (8/7/08) - MoGo beats professional Go playerMoGo beats professional Go player What Can AI Systems Do Now?

26 Robotics

27 DARPA Grand Challenge  Completely autonomous vehicles (no human guidance)  Several hundred miles over varied terrain  First challenge (2004) – 142 miles  “winner” traveled seven(!) miles  Second challenge (2005) – 131 miles  Winning team (Stanford) completed the course in under 7 hours  Three other teams completed the course in just over 7 hours  Onwards and upwards (2007)  Urban Challenge  Traffic laws, merging, traffic circles, busy intersections...  Six finishers (best time: 2.8 miles in 4+ hours)

28 Art: NEvAr  Use genetic algorithms to evolve aesthetically interesting pictures  See

29 ALife: Evolutionary Optimization  MERL: evolving ‘bots

30 Human-Computer Interaction: Sketching  Step 1: Typing  Step 2: Constrained handwriting  Step 3: Handwriting recognition  Step 4: Sketch recognition (doodling)!  MIT sketch tablet

31 Driving: Adaptive Cruise Control  Adaptive cruise control and pre- crash safety system (ACC/PCS)  Offered by dozens of makers, mostly as an option (~$1500) on high-end models  Determines appropriate speed for traffic conditions  Senses impending collisions and reacts (brakes, seatbelts)  Latest AI technology: automatic parallel parking!

32 AxonX  Smoke and fire monitoring system

33 Rocket Review  Automated SAT essay grading system

34 What Can’t AI Systems Do (Yet)?  Understand natural language robustly (e.g., read and understand articles in a newspaper)  Surf the web (or a wave)  Interpret an arbitrary visual scene  Learn a natural language  Play Go well √  Construct plans in dynamic real-time domains  Refocus attention in complex environments  Perform life-long learning

Computational vs. Biological Memory

36 How Does It Work? (Humans)  Basic idea:  Chemical traces in the neurons of the brain  Types of memory:  Primary (short-term)  Secondary (long-term)  Factors in memory quality:  Distractions  Emotional cues  Repetition

37 How Does It Work? (Computers)  Basic idea:  Store information as “bits” using physical processes (stable electronic states, capacitors, magnetic polarity,...)  One bit = “yes or no”  Types of computer storage:  Primary storage (RAM or just “memory”)  Secondary storage (hard disks)  Tertiary storage (optical jukeboxes)  Off-line storage (tape drives)  Factors in memory quality:  Power source (for RAM)  Avoiding extreme temperatures Size Speed

38 Measuring Memory  Remember that one yes/no “bit” is the basic unit  Eight (2 3 ) bits = one byte  1,024 (2 10 ) bytes = one kilobyte (1K) *  1,024K (2 20 bytes) = one megabyte (1M)  1,024K (2 30 bytes) = one gigabyte (1G)  1,024 (2 40 bytes) = one terabyte (1T)  1,024 (2 50 bytes) = one petabyte (1P)  bytes = one yottabyte (1Y?) * Note that external storage is usually measured in decimal rather than binary (1000 bytes = 1K, and so on)

39 What Was It Like Then?  The PDP-11/70s we used in college had 64K of RAM, with hard disks that held less than 1M of memory ... and we had to walk five miles, uphill, in the snow, every day! And we had to live in a cardboard box in the middle of the road!

40 What Is It Like Now?  The PDP-11/70s we used in college had 64K of RAM, with hard disks that held less than 1M of memory  The cheapest Dell Inspiron laptop has 2G of RAM and up to 80G of hard drive storage.... ...a factor of more RAM and more disk space ...and your iPod nano has 8G of blindingly fast storage ...so don’t come whining to me about how slow your computer is!

41 Moore’s Law  Computer memory (and processing speed, resolution, and just about everything else) increases exponentially

42 Showdown  Computer capacity:  Primary storage: 64GB  Secondary storage: 750GB (~10 12 )  Tertiary storage: 1PB? (10 15 )  Computer retrieval speed:  Primary: sec.  Secondary: sec.  Computing capacity: 1 petaflop (10 15 floating-point instructions per second), very special purpose  Digital  Extremely reliable  Not (usually) parallel  Human capacity:  Primary storage: 7 ± 2 “chunks”  Secondary storage: bits?? (or maybe 10 9 bits?)  Human retrieval speed:  Primary: sec  Secondary: sec  Computing capacity: possibly 100 petaflops, very general purpose  Analog  Moderately reliable  Highly parallel More at movementarian.com

43 It’s Not Just What You “Know”  Storage  Indexing  Retrieval  Inference  Semantics  Synthesis ...So far, computers are good at storage, OK at indexing and retrieval, and humans win on pretty much all of the other dimensions ...but we’re just getting started  Electronic computers were only invented 60 years ago!  Homo sapiens has had a few hundred thousand years to evolve...

The Skeptics Speak

45 Mind and Consciousness  Many philosophers have wrestled with the question:  Is Artificial Intelligence possible?  John Searle: most famous AI skeptic  Chinese Room argument  Is this really intelligence? ? !

46 What Searle Argues  People have beliefs; computers and machines don’t.  People have “intentionality”; computers and machines don’t.  Brains have “causal properties”; computers and machines don’t.  Brains have a particular biological and chemical structure; computers and machines don’t.  (Philosophers can make claims like “People have intentionality” without ever really saying what “intentionality” is, except (in effect) “the stuff that people have and computers don’t.”)

47 Let’s Introspect For a Moment... HH ave you ever learned something by rote that you didn’t really understand? WW ere you able to get a good grade on an essay where you didn’t really know what you were talking about? HH ave you ever convinced somebody you know a lot about something you really don’t? AA re you a Chinese room?? WW hat does “understanding” really mean? WW hat is intentionality? Are human beings the only entities that can ever have it? WW hat is consciousness? Why do we have it and other animals and inanimate objects don’t? (Or do they?)

48 Just You Wait...

49 Thank You!