![The Turing Test
More on Turing test
g
Objective: The machine needs to fool the machine
[INT] I heard that a striped rhinoceros flow on the
Mississippi in a pink balloon this morning. What do
you think about?
[COMP] That sound rather ridiculous to me
[INT] Really? My uncle did this one... Why this sound
ridiculous?
[COMP] Option 1: Rhinoceros don't have stripes
don t
[COMP] Option 2: Rhinoceros can't fly
Try
Tr to change ON for UNDER the Mississipi
Is this unfair for the computer?
[INT] What’s the result of 324 x 678?
[COMP] This is too difficult. I’m not a calculator!
Needs to seem more foolish than it actually is (has to lie!)
Artificial Intelligence Machine Learning Slide 16](https://image.slidesharecdn.com/lecture1-ai1-introductionai-090922155953-phpapp02/85/Lecture1-AI1-Introduction-to-artificial-intelligence-16-320.jpg)
Ad
More Related Content
What's hot (20)
Similar to Lecture1 AI1 Introduction to artificial intelligence (20)
Ad
More from Albert Orriols-Puig (20)
Ad
Recently uploaded (20)
Lecture1 AI1 Introduction to artificial intelligence
- 1. Introduction Artificial Intelligence Lecture 1 Albert Orriols i Puig http://www.albertorriols.net htt // lb t i l t [email protected] Artificial Intelligence – Machine Learning g g Enginyeria i Arquitectura La Salle Universitat Ramon Llull
- 2. Today’s Agenda Brainstorming from y g your “postits” p Some Definitions Prehistory and History of AI Where are we headed? Artificial Intelligence Machine Learning Slide 2
- 3. Brainstorming What’s AI? A A … Do you know of some real-world applications? A A … Artificial Intelligence Machine Learning Slide 3
- 4. What’s Intelligence? Intelligence (dictionary) g ( y) capacity for learning, reasoning, understanding, and similar forms o mental ac o s of e a activity; ap ude in grasping truths, y; aptitude g asp g u s, relationships, facts, meanings, etc. In particular, we cou d say pa cu a , e could say: Ability to act as human beings Solve problems Think rationally Artificial intelligence … Building a machine that is (or seems to be at the eyes of the beholder) intelligent Artificial Intelligence Machine Learning Slide 4
- 5. Can You Be More Formal? What is artificial intelligence? g 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 co o confine itself to methods that a e b o og ca y obse ab e e se o e ods a are biologically observable. Yes, but what is intelligence? Intelligence i th computational part of the ability t achieve goals i I t lli is the t ti l t f th bilit to hi l in the world. Varying kinds and degrees of intelligence occur in people, many animals and some machines. Isn't there a solid 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. d t d f th h i f i t lli d t th See the complete interview at: http://www-formal.stanford.edu/jmc/whatisai/node1.html Artificial Intelligence Machine Learning Slide 5
- 6. What’s Involved in Intelligence? Ability to interact with the real world to perceive, understand, and act e.g., speech recognition and understanding Searching the best solution Reasoning and Planning modeling the external world, given input solving new problems, planning, and making decisions ability to deal with unexpected problems, uncertainties Learning and Adaptation we are continuously learning and adapting y g p g our internal models are always being “updated” e.g., a baby learning to categorize and recognize animals g, y g g g Artificial Intelligence Machine Learning Slide 6
- 7. AI Is Not Alone at Home Crossbreeding of a lot of fields g Philosophy Logic, methods of reasoning, mind as physical system, foundations of learning language rationality learning, language, rationality. Mathematics Formal representation and proof, algorithms, computation, (un)decidability, (in)tractability Statistics Modeling uncertainty, learning from data Economics Utility, decision theory, rational economic agents Neuroscience Neurons as information processing units Psychology / Neuro How do people behave, perceive, process cognitive Science information, information represent knowledge Computer Engineering Building fast computers Control Theory y Design systems that maximize an objective function g y j over time Linguistics Knowledge representation, grammars Artificial Intelligence Machine Learning Slide 7
- 8. Prehistory of AI Through history, people though of mythic “artificial” g y, p p g y robots golden robots of Hephaestus and Pygmalion s Galatea Pygmalion's alchemical means of placing mind into matter More specific, tangible advances 5th century B.C. Aristotle invented syllogistic logic, the first formal deductive reasoning system. 13th century. Talking heads were said to have been created (Roger Bacon and Alb t th G d Albert the Great). t) Ramon Lull, Spanish theologian, invented machines for discovering nonmathematical truths through combinatory. combinatory Artificial Intelligence Machine Learning Slide 8
- 9. Prehistory of AI More specific, tangible advances (cont.) p , g ( ) 15th century Invention of printing using moveable type. Gutenberg Bible type printed (1456). 15th 16th 15th-16th century Clocks, the first modern measuring machines, were first produced using lathes. 16th century Clockmakers extended their craft to creating mechanical animals and other novelties. Artificial Intelligence Machine Learning Slide 9
- 10. Prehistory of AI More specific, tangible advances (cont.) p , g ( ) 17th century - The revolution of thinking about thinking Descartes proposed that bodies of animals are nothing more than complex machines (strong AI). Variations and elaborations of Cartesian mechanism. Hobbes published The Leviathan, containing a material and combinatorial theory of thinking. Pascal created the first mechanical digital calculating machine (1642). Leibniz improved Pascal's machine to do multiplication & division ( (1673) and envisioned a universal calculus of reasoning by which ) g y arguments could be decided mechanically. Artificial Intelligence Machine Learning Slide 10
- 11. Prehistory of AI More specific, tangible advances (cont.) p , g ( ) 18th century – Mechanical toys Vaucanson’s Duck Von Kempelen’s phony mechanical chess player Artificial Intelligence Machine Learning Slide 11
- 12. Prehistory of AI More specific, tangible advances (cont.) p , g ( ) 19th century – Frankenstein’s birth George Boole developed a binary algebra representing (some) "laws of thought," published in The Laws of Thought. Charles Babbage and Ada Byron (Lady Lovelace) worked on g y ( y ) programmable mechanical calculating machines. Mary Shelley published the story of Frankenstein's monster (1818). Crossing the century bridge Behaviorism was expounded by psychologist Edward Lee Thorndike in "Animal Intelligence." Artificial Intelligence Machine Learning Slide 12
- 13. Pre-birth of AI Beginning of the 20th century g g y Russell and Whitehead published Principia Mathematica. Capek s Capek's play “Rossum's Universal Robots” produced in 1921 (London Rossum s Robots opening, 1923). First use of the word 'robot' in English. McCulloch and Pitts publish "A Logical Calculus of the Ideas Immanent in A Nervous Activity" (1943), laying foundations for neural networks. Rosenblueth, Wiener and Bigelow coin the term cybernetics (1943). Bush published As We May Think (1945) a prescient vision of the future in which computers assist humans in many activities. Artificial Intelligence Machine Learning Slide 13
- 14. The 3 Key Ingredients The first key ingredient: The computer and the program y g p p g ENIAC (1945). The first electronic digital computer EDVAC (1949) Th first stored program computer (1949). The fi t t d t Artificial Intelligence Machine Learning Slide 14
- 15. The 3 Key Ingredients The second key ingredient: The TURING TEST. y g (Human) judge communicates with a human and a machine o e e o y channel. over text-only c a e Both human and machine try to act like a human Judge tries to tell hi h is hi h J d t i t t ll which i which. Numerous variants. Loebner prize. Cu e t programs o e e close Current p og a s nowhere c ose to passing this http://www.jabberwacky.com/ http://turingtrade.org/ Artificial Intelligence Machine Learning Slide 15
- 16. The Turing Test More on Turing test g Objective: The machine needs to fool the machine [INT] I heard that a striped rhinoceros flow on the Mississippi in a pink balloon this morning. What do you think about? [COMP] That sound rather ridiculous to me [INT] Really? My uncle did this one... Why this sound ridiculous? [COMP] Option 1: Rhinoceros don't have stripes don t [COMP] Option 2: Rhinoceros can't fly Try Tr to change ON for UNDER the Mississipi Is this unfair for the computer? [INT] What’s the result of 324 x 678? [COMP] This is too difficult. I’m not a calculator! Needs to seem more foolish than it actually is (has to lie!) Artificial Intelligence Machine Learning Slide 16
- 17. The 3 Key Ingredients The third key ingredient: THE DARMONT CONFERENCE. y g People working on building intelligent machines. J. McCarthy, M. L. Minsky, N. Rochester, and C.E. Shannon. Shannon August 31, 1955. "We propose that a 2 month 31 1955 We month, 10 man study of artificial intelligence be carried out during the summer of 1956 at Dartmouth College in Hanover, New Hampshire. The study is to proceed on the basis of the conjecture 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." i l t it " Artificial Intelligence Machine Learning Slide 17
- 18. Brief History of AI The Golden years ( y (1956 – 1974) ) ‘1960s Strong funding of AI centers Building intelligent automata Searching in complex search spaces First AI programs that work Samuel’s checker program (which learns) S l’ h k ( hi h l ) Newell and Simon’s Logic Theorist Gelernter’s geometry engine G l t ’ t i Robinson’s complete algorithm for logical reasoning First programming languages for AI McCarthy - Lisp (1958) Artificial Intelligence Machine Learning Slide 18
- 19. Brief History of AI The Golden years ( y (1956 – 1974) ) And the first chatterbots: (1966). ELIZA (1966) It carried out very realistic conversations. It searched for key words in the conversation and asked y information about that Artificial Intelligence Machine Learning Slide 19
- 20. Brief History of AI The Winter: After expansion, there’s always a contraction p , y First doubts on the feasibility of all the approach Problems: P bl Limited computer power Combinatorial C bi t i l explosion (exponential time) l i ( ti l ti ) Commonsense knowledge and reasoning Moravec’s paradox M ’ d The Chinese room argument undermined the goal of building intelligent machines END OF FUNDING Artificial Intelligence Machine Learning Slide 20
- 21. Brief History of AI The Chinese room argument (Searle, 1980) g ( , ) Person who knows English but not C ese sits Chinese s s in room oo Receives notes in Chinese Has H systematic English rule b k f t ti E li h l book for how to write new Chinese characters based on input Chinese c a acte s, returns his notes o put C ese characters, etu s s otes 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 i Chinese! in Chi ! Searle’s argument: the room has no intelligence in it! Artificial Intelligence Machine Learning Slide 21
- 22. Brief History of AI But in parallel… expert systems rise and grow p p y g MYCIN(1972): Diagnosed infection blood diseases. diseases It had a set of about 600 rules and started to ask questions. In some cases, better than human experts. cases experts XCON (1980): Production-rule-based system that assisted the ordering of a P d ti l b d t th t i t d th d i f type of computers systems by automatically selecting the computer systems components based on the customers requirements. Saving $40 million dollars to the company. 2500 rules and processed 80000 orders with 95%-98% accuracy. The gain in money was because it reduced the need to give free components when the technicians made errors, by speeding errors the assembly process and by increasing customer satisfaction Artificial Intelligence Machine Learning Slide 22
- 23. Brief History of AI But in parallel… expert systems rise and grow p p y g PROSPECTOR (1981) A computer-based consultation system for mineral exploration. Recommending exploratory drilling g p y g And many others. Search the web for more! New funding due to this success AI groups were formed in many large companies to develop expert systems. t t 1986 sales of AI-based hardware and software were $425 million. illi Artificial Intelligence Machine Learning Slide 23
- 24. Brief History of AI Q Quick pace in the ‘90s p NCSA releases the first web browser, Mosaic Deep Bl b t G D Blue beats Gary K Kasparov Robotic soccer players in RoboCup Sony corporation introduced the robotic dog AIBO Remote age t auto o ous y d e deep space 1 e ote agent autonomously drive Even moving faster in the 00’s iRobot introduces the vacuum cleaning robot Roomba DARPA grand challenge (we’ll see it in a minute) A Touareg R5 won the challenge Artificial Intelligence Machine Learning Slide 24
- 25. Some Cool Applications Three cool applications among hundreds pp g Deep Blue DARPA G d Ch ll Grand Challenge Robotics Cog Loebner Prize Roomba oo ba Rob-Cup ASIMO Data mining Stock Market Medical Diagnosis Artificial Intelligence Machine Learning Slide 25
- 26. Deep Blue Origins at CMU It was a massively parallel, RS/6000 SP Thin P2SC-based system with 30-nodes Deep Blue took Gary Kasparov to the cleaners Artificial Intelligence Machine Learning Slide 26
- 27. DARPA Grand Challenge Grand Challenge Cash prizes ($1 to $2 million) offered to first robots to complete a long course completely unassisted Stimulates research in vision robotics planning machine vision, robotics, planning, learning, reasoning, etc 2004 Grand Challenge: 150 mile route in Nevada desert Furthest any robot went was about 7 miles … but hardest terrain was at the beginning of the course 2005 Grand Challenge: G d Ch ll 132 mile race Narrow t N tunnels, winding mountain passes, etc l i di t i t Stanford 1st, CMU 2nd, both finished in about 6 hours Artificial Intelligence Machine Learning Slide 27
- 28. DARPA Grand Challenge http://cs.stanford.edu/group/roadrunner/ p g p Artificial Intelligence Machine Learning Slide 28
- 29. DARPA Grand Challenge The challenge: a driverless car competes for wining the g p g race 150 mile off-road robot race across the Mojave desert Natural and manmade hazards No driver, no remote control N di t t l No dynamic passing Fastest vehicle wins the race (and 2 million dollar prize) Artificial Intelligence Machine Learning Slide 29
- 30. DARPA Grand Challenge The architecture Artificial Intelligence Machine Learning Slide 30
- 31. Robotics - Cog Humanoid intelligence requires humanoid interactions g q with the world Artificial Intelligence Machine Learning Slide 31
- 32. Loebner Prize Prizes the chatterbots considered to be the most human-like The Th contest begun in 1990 t tb i $25,000 is offered for the first chatterbot that judges cannot j g distinguish from a real human and that can convince judges that the human is the computer program $100,000 is the reward for the first chatterbot that judges cannot distinguish from a real human in a Turing test that includes deciphering and understanding text, visual, and auditory input Artificial Intelligence Machine Learning Slide 32
- 33. Roomba Go around “smartly” to clean up a house y p Artificial Intelligence Machine Learning Slide 33
- 34. RobCup First official Rob-Cup soccer match (1997) p ( ) Artificial Intelligence Machine Learning Slide 34
- 35. ASIMO Advanced Step in Innovative Mobility y Able of Moving Interacting with human beings Help people Artificial Intelligence Machine Learning Slide 35
- 36. Data Mining Explosion Data mining: Extract novel, useful, and interesting g , , g information from data Why so a big deal? Companies are generating lots of data about the business They want to process these data and obtain useful information Why no Wh now, not before? Computers have a lot of power nowadays Artificial Intelligence Machine Learning Slide 36
- 37. Modeling the Stock Market Modeling market traders g LETS project: Evolving artificial traders for successful market trading (Sonia Sc u e bu g et a , 2007) ad g (So a Schulenburg al, 00 ) Evolutionary economics: Create trend followers and value investors Let them interact Evolve a population of strategies Artificial Intelligence Machine Learning Slide 37
- 38. Medical Diagnosis Data mining An important application domain of artificial intelligence John H. Holmes Epidemiologic study by means of LCSs Hidden relationships among variables discovered by LCSs Xavier Llorà et al. Better than Human Capability in Diagnosing Prostate Cancer Using Infrared Spectroscopic imaging Artificial Intelligence Machine Learning Slide 38
- 39. But… Slow it down! There are no castles in the sky All these applications rely on: Search & Optimization Knowledge representation Learning Planning These are the four topics that we’ll see in this course. And we will start for the beginning Artificial Intelligence Machine Learning Slide 39
- 40. Detailed Outline AI1 2. Solving search problems 1. Introduction to search problems 2. Blind search 3. Informed/heuristic search 4. Adversary search (first project) 5. Constraint satisfaction problems 3. 3 Knowledge representation 1. Introduction to knowledge representation 2. 2 Knowledge representation based on logics 3. Knowledge and uncertainty 4. Fuzzy Logics F L i 4. Lisp Artificial Intelligence Machine Learning Slide 40
- 41. Detailed Outline AI2 5. Machine learning 1. Introduction to machine learning 2. Supervised learning 1. Decision trees, Instance-based learning, Bayesian decision theory, Support vector machines and Neural networks 3. 3 Unsupervised learning – association rules 4. Unsupervised learning – clustering 5. Reinforcement learning g 6. New challenges in data mining 6. Planning 1. Introduction to planning p g 2. STRIPS 3. Search through the state world g 4. Search through the plan world Artificial Intelligence Machine Learning Slide 41
- 42. Introduction Artificial Intelligence Lecture 1 Albert Orriols i Puig http://www.albertorriols.net htt // lb t i l t [email protected] Artificial Intelligence – Machine Learning g g Enginyeria i Arquitectura La Salle Universitat Ramon Llull