Com1005 Machines and Intelligence

1. Com1005 Machines and Intelligence Lecturers: Dr Amanda Sharkey, Professor Phil Green

2. Lecture 1: What is Artificial Intelligence? • “Artificial Intelligence (AI) is the study of intelligent behaviour (in humans, animals and machines) and the attempt to find ways in which such behaviour could be engineered in any type of artifact” (Whitby, 2003)

3. Varying defintions • John McCarthy, “It is the science and engineering of making intelligent machines, especially intelligent computer programs.” • Herbert Simon: We call programs intelligent if they exhibit behaviours that would be regarded intelligent if they were exhibited by human beings. • Elaine Rich (1991) “AI is the study of how to make computers do things at which, at the moment, people are better.” • Astro Teller: AI is the attempt to make computers do what they do in the movies.

5. Main themes of this semester Artificial Intelligence – overview of progress • Different approaches to creating intelligent behaviour • Computationalism .... Mind as computer • Brain-like (Artificial Neural Nets) • Brain, body and world (Embodied AI) • Different goals • Understanding and simulating intelligence • Applied AI • Creating the illusion of intelligence Artificial Intelligence programming: search, planning and knowledge representation

6. Lectures • “Guru” lectures – AI research in the department • Assessment over the year: • written assignment (this semester), • Group presentation (this semester), • practical assignment (next semester) • exam for whole year (next semester).

7. Origins of AI and early history of digital computer • 1941 Germany: KonradZuse, Z3 first general purpose programmable computer

8. Colossus – 10 delivered to Bletchley Park. Designed by British engineer Tommy Flowers to break Nazi codes.

9. ENIAC (1945) electronic computer rewired by hand for each task.

10. Manchester Mark 1 computer: 1948. General purpose computer with stored programs. Punched paper tape for new job

11. AI: Where did it all begin? • 1956 Dartmouth Summer Research Project • month long ‘brain storming’ session • Attendees: John McCarthy (Father of AI, inventor of LISP). Invented term “artificial intelligence” • Also Allen Newell, Herbert Simon, Marvin Minsky, Oliver Selfridge, Claude Shannon and others • Idea that “every aspect of learning, or any other feature of intelligence can in principle be so precisely described that a machine can be made to simulate it”

12. What is Artificial Intelligence? • Attempt to understand intelligent entities • Attempt to build intelligent entities • Attempt to create the appearance of intelligence

13. Understanding intelligent entities • Can computers be intelligent? • Or is intelligence unique to humans, or to living beings? • Can we use computers to help us understand how we think?

14. John Searle: • Strong AI: an appropriately programmed computer really is a mind, can be said to understand, and has other cognitive states. • Weak AI: a computer is a valuable tool for study of mind – makes it possible to formulate and test hypotheses rigorously

16. Ray Kurzweil: “The singularity is near” • Strong AI .... Artificial intelligence that matches or exceeds human intelligence • Artificial general intelligence • Artificial narrow intelligence

17. Building intelligent machines • Designing programs to perform tasks intelligently • Should computer programs think like humans? • Should they be programmed to operate like human brains? • Should they exploit different strengths?

18. Creating the appearance of intelligence

19. What is intelligence? Can psychology help? • 1921 Journal of Educational Psychology asked 14 experts for definitions • 14 different definitions including • The ability to carry on abstract thinking (Terman) • The ability to adapt oneself adequately to relatively new situations in life (Pintner) • The capacity to acquire capacity (Woodrow) • The capacity to learn or profit by experience (Dearborn) • “few concepts in psychology have received more devoted attention and few have resisted clarification so thoroughly” (Reber, 1995)

20. Intelligence in humans • What is intelligence? • Intelligence is what is measured by intelligence tests. • IQ intelligence quotient • Based on abstract reasoning ability

21. 1915 Stanford-Binet test- • Objective measure, aims – to identify those children who need specialised, simplified education • Uses concept of mental age versus chronological age • Items in test age graded- mental age corresponds to level achieved in test. • Eg. A 4 year old should be able to complete the following: • Brother is a boy; Sister is a …… Bright child – mental age above chronological age.

22. Adult versions developed and standardized. • Weschler developed test for adults and proposed Gaussian distribution of results • 2/3 should be between 85 and 115, (100 mean) and 2.3% above 130 and below 70.

23. Single factor, or multiple intelligences? • Spearman, single factor g underlying intelligence. • Gardner – multiple intelligences • Linguistic, musical, logical-mathematical, spatial, bodily-kinesthetic, personal • But strong correlations between performance in different areas – single underlying factor of intelligence?

24. Correlations of .4 and .6 between school grades and Wechsler IQ test. • Correlation with university results lower • Correlation with job performance .51 (Hunter and Schmidt 1998)

25. Problems: • Difficult to find items which are independent of culture and education • E.g. pick odd word • cello harp drum violin guitar • Rich kids picked drum, poor kids cello • Also IQ tests don’t measure • Creativity • Motivation • Most geniuses also work very hard (e.g Mozart and practice)

26. Robert Sternberg (2003) • Triarchic theory of intelligence • 1. Analytical intelligence, the ability to complete academic, problem-solving tasks, such as those used in traditional intelligence tests. These types of tasks usually present well-defined problems that have only a single correct answer. • 2. Creative or synthetic intelligence, the ability to successfully deal with new and unusual situations by drawing on existing knowledge and skills. Individuals high in creative intelligence may give 'wrong' answers because they see things from a different perspective. • 3. Practical intelligence, the ability to adapt to everyday life by drawing on existing knowledge and skills. Practical intelligence enables an individual to understand what needs to be done in a specific setting and then do it.

27. Machines and IQ tests?

28. Intelligence and machines?The Turing Test

29. Turing test • Outside appearances used to judge intelligence • Turing, A. (1950) Computing machinery and intelligence, Mind, 59, 433-460. • I propose to consider the question ‘Can machines think?’ • When paper written, only 4 electronic computers in existence, and it was before Dartmouth conference and ‘birth’ of AI

30. Turing test • Like Victorian parlour game. I interrogator, 1 woman, 1 man • 1 interrogator, 1 person, 1 computer • Interrogator can ask any questions • Human must be truthful (trying to help the interrogator), computer can try to force wrong identification • Experiment repeated with a range of people in 2 human positions. Test passed if number of successful identifications of computer is less than 30% • Machine can lie – e.g. if asked if it has ever written a poem it could say it has, and it could pretend to get calculations wrong.

32. Predictions • 1950 Turing: “I believe that in about fifty years' time it will be possible to programme computers, with a storage capacity of about 109, to make them play the imitation game so well that an average interrogator will not have more than 70 percent chance of making the right identification after five minutes of questioning. … I believe that at the end of the century the use of words and general educated opinion will have altered so much that one will be able to speak of machines thinking without expecting to be contradicted.” • 1965, Herbert Simon predicted that “machines will be capable, within twenty years, of doing any work that a man can do”. • 1967 Marvin Minsky “within a generation… few compartments of intellect will remain outside the machine’s realm”

33. ELIZA • Weizenbaum, J. (1965) ELIZA – a computer program for the study of natural language communication between man and machine. Communications of the Association for Computing Machinery, 9(1) 36-45.

34. Therapeutic interview with a teenager • Men are all alike • IN WHAT WAY? • They’re always bugging us about something or other. • CAN YOU THINK OF A SPECIFIC EXAMPLE? • Well my boyfriend made me come here • YOUR BOYFRIEND MADE YOU COME HERE? • He says I’m depressed much of the time • I AM SORRY TO HEAR THAT YOU ARE DEPRESSED • It’s true that I am unhappy • DO YOU THINK COMING HERE WILL HELP YOU NOT TO BE UNHAPPY? • I need some help, that much seems certain. • WHAT WOULD IT MEAN TO Y0U IF YOU GOT SOME HELP?

35. Perhaps I could learn to get along with my mother. • TELL ME MORE ABOUT YOUR FAMILY • My mother takes care of me • WHO ELSE IN YOUR FAMILY TAKES CARE OF YOU? • My father • YOUR FATHER • You are like my father in some ways • WHAT RESEMBLANCE DO YOU SEE? • You are not very aggressive, but I think you don’t want me to notice that. • WHAT MAKES YOU THINK I AM NOT VERY AGGRESSIVE? • You don’t argue with me • WHY DO YOU THINK I DON’T ARGUE WITH YOU? • You are afraid of me

36. DOES IT PLEASE YOU TO BELIEVIE I AM AFRAID OF YOU? • My father is afraid of everybody • WHAT ELSE COMES TO MIND WHEN YOU THINK OF YOUR FATHER? • Bullies • DOES THAT HAVE ANYTHING TO DO WITH THE FACT THAT YOUR BOYFRIEND MADE YOU COME HERE?

37. Has the test been passed already? • Business executive and Eliza • Do you think that I can use the computer this morning? • WHY DO YOU ASK? • I might be able to make some additional Telcomp sales. • WHY AREN’T YOU SURE? • My prospective customers have never seen a demonstration of the system • DOES THIS MEAN VERY MUCH TO YOU? • Of course it does. • YOU SEEM VERY POSITIVE ABOUT THAT • This is ridiculous • WHAT DO YOU MEAN THIS IS RIDICULOUS? • Is there any reason why I can’t use the computer now? • TELL ME MORE ABOUT YOUR PROSPECTIVE CUSTOMERS

38. Parry the paranoid program.Colby, K.M., Weber, S., Hildf, F.D. (1971) Artificial Paranoia. Artificial Intelligence, 2, 1-25. • Why are you in the hospital? • I SHOULDN’T BE HERE • Who brought you here? • THE POLICE • What trouble did you have with the police? • COPS DON’T DO THEIR JOB • What do they do instead? • COPS ARREST THE WRONG PEOPLE • They arrest the right ones sometimes. • THE LAW SHOULD GET THOSE ITALIAN CROOKS • Are you Italian? • MY BACKGROUND IS BRITISH BUT I WAS BORN IN THIS COUNTRY

39. Do you know any Italian crooks? • I TRY TO AVOID THE UNDERWORLD • Are you afraid of the underworld? • THE MAFIA ARE THE WORST GANGSTERS • Are you connected with the Mafia? • YOU KNOW, THEY KNOW ME • How do they know you? • THEY SPY ON ME • Why do they? • THEY WANT TO GET FALSE EVIDENCE TO FRAME ME • What do they have against you? • THEY ARE OUT TO GET ME • What for?

40. I HAVE INCRIMINATING EVIDENCE AGAINST THE MAFIA • Are you sure of all this? • YOU DON’T BELIEVE ME? • I am doing my best….. • Silence • CAN’T YOU THINK OF A DECENT QUESTION TO ASK ME? • You seem hostile to me • SOME OF YOUR REMARKS ANNOY ME • I am sorry • FORGET THE EXCUSES, YOU GUYS ARE ALL ALIKE.

41. Loebner Test • Competition started 1991 by Hugh Loebner • Judges type questions on terminal, either human or computer replies • Judges rate responses on how human-like they appear. $2000 prize to the most human like. $100,000 prize if a program does better than one of the humans • 2010 competition: 23rd October in LA • http://www.loebner.net/Prizef/loebner-prize.html • 1997 winner: Intelligent Research Ltd, partly programmed by our NLP group

42. 2009: “Do-Much-More” won.... 2009 winner of Loebner Prize: “Do-Much-More” Designed to respond with generalities, moving conversation along in light hearted way. No specialist knowledge.

43. Judge-1: Hi, how is it going? • Do-Much-More: Hello. How are you doing? • Judge-1: Good, who are you? • Do-Much-More: Well, I am studying engineering. That ought to give you an idea about who I am. • Judge-1: No, not really? • Do-Much-More: I have a terrific yearning to know something more about experiments? • Judge-1: Where are you from? • Do-Much-More: I don't know, but don't forget I can't be in three places at once. • Judge-1: What lang~~~~city do you live in? • Do-Much-More: What city -- that's rather a tricky question. The trouble is, one city seems to be much like another. • Judge-1: Not really. Where do you live? • Do-Much-More: In the old dairy next to the river, I believe. • Judge-1: Interesting, so you are a farmer then? • Do-Much-More: Well, if you ask a physiologist what I am, he won't say I'm a farmer. He'll say I consist mostly of water -- enough to fill a large beer tun -- plus some iron, phosphorus, and lots of other prosaic ingredients. • Judge-1: Wie~~eird answer to a simple question, wouldn't yous~ say?

44. Do-Much-More: I certainly would. Just watch me! Judge-1: What will I see? Do-Much-More: You don't have to worry about what you see, you shouldn't see it on a cold winter's evening. Judge-1: What do you make of the Arctic Monkeys? Do-Much-More: Well, here's a clue: I make what a keeper in a zoo would make. Try that for size!

45. Survey of 21 experts at AGI-09 conference (Baum et al, 2010)

46. Summary • Birth of AI, and first computers • What is intelligence? – history of intelligence testing • Turing test – a particular view of what intelligence is. Takes simplifying view that if behaviour of computer is indistinguishable from human, it must be intelligent. • Early conversational programs • Loebner test