ARTIFICIAL INTELLIGENCES
INTRODUCTION
The intellectual roots of AI, and the concept of intelligent machines, may be found in Greek mythology. Intelligent artifacts appear in literature since then, with real (and fraudulent) mechanical devices actually demonstrated to behave with some degree of intelligence. Some of these conceptual achievements are listed below under "Ancient History."
After modern computers became available, following World War II, it has become possible to create programs that perform difficult intellectual tasks. From these programs, general tools are constructed which have applications in a wide variety of everday problems. Some of these computational milestones are listed below under "Modern History."
ANCIENT HISTORY
Greek myths of Hephaestus and Pygmalion incorporate the idea of intelligent robots. Many other myths in antiquity involve human-like artifacts. Many mechanical toys and models were actually constructed, e.g., by Hero, Daedalus and other real persons.
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 Albert the Great reputedly among the owners.
Ramon Llull, Spanish theologian, invented machines for discovering nonmathematical truths through combinatories.
15th century
Invention of printing using moveable type. Gutenberg Bible printed (1456).
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.
Rabbi Loew of Prague is said to have invented the Golem, a clay man brought to life (1580).
17th century
Early in the century, Descartes proposed that bodies of animals are nothing more than complex machines. Many other 17th century thinkers offered 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 evisioned a universal calculus of reasoning by which arguments could be decided mechanically.
18th century
The 18th century saw a profusion of mechanical toys, including the celebrated mechanical duck of Vaucanson and von Kempelen's phony mechanical chess player, The Turk (1769).
19th century
Luddites (led by Ned Ludd) destroyed machinery in England (1811-1816).
Mary Shelley published the story of Frankenstein's monster (1818).
George Boole developed a binary algebra representing (some) "laws of thought."
Charles Babbage & Ada Byron (Lady Lovelace) worked on programmable mechanical calculating machines.
20th century - First Half
Bertrand Russell and Alfred North Whitehead published Principia Mathematica, which revolutionaized formal logic. Russell, Ludwig Wittgenstein, and Rudolf Carnap lead philosophy into logical analysis of knowledge.
Karel Capek's play "R.U.R." (Rossum's Universal Robots) opens in London (1923). - First use of the word 'robot' in English.
Warren McCulloch & Walter Pitts publish "A Logical Calculus of the Ideas Immanent in Nervous Activity" (1943), laying foundations for neural networks.
Arturo Rosenblueth, Norbert Wiener & Julian Bigelow coin the term "cybernetics" in a 1943 paper. Wiener's popular book by that name published in 1948.
Vannevar Bush published As We May Think (Atlantic Monthly, July 1945) a prescient vision of the future in which computers assist humans in many activities.
A.M. Turing published "Computing Machinery and Intelligence" (1950). - Introduction of Turing's Test as a way of operationalizing a test of intelligent behavior.
Claude Shannon published detailed analysis of chess playing as search (1950).
Isaac Asimov published his three laws of robotics (1950).
MODERN HISTORY
1956
John McCarthy coined the term "artificial intelligence" as the topic of the Dartmouth Conference, the first conference devoted to the subject.
Demonstration of the first running AI program, the Logic Theorist (LT) written by Allen Newell, J.C. Shaw and Herbert Simon (Carnegie Institute of Technology, now Carnegie Mellon University).
1957
The General Problem Solver (GPS) demonstrated by Newell, Shaw & Simon.
1952-62
Arthur Samuel (IBM) wrote the first game-playing program, for checkers, to achieve sufficient skill to challenge a world champion. Samuel's machine learning programs were responsible for the high performance of the checkers player.
1958
John McCarthy (MIT) invented the Lisp language.
Herb Gelernter & Nathan Rochester (IBM) described a theorem prover in geometry that exploits a semantic model of the domain in the form of diagrams of "typical" cases.
Teddington Conference on the Mechanization of Thought Processes was held in the UK and among the papers presented were John McCarthy's "Programs with Common Sense," Oliver Selfridge's "Pandemonium," and Marvin Minsky's "Some Methods of Heuristic Programming and Artificial Intelligence."
Late 50's & Early 60's
Margaret Masterman & colleagues at Cambridge design semantic nets for machine translation.
1961
James Slagle (PhD dissertation, MIT) wrote (in Lisp) the first symbolic integration program, SAINT, which solved calculus problems at the college freshman level.
1962
First industrial robot company, Unimation, founded.
1963
Thomas Evans' program, ANALOGY, written as part of his PhD work at MIT, demonstrated that computers can solve the same analogy problems as are given on IQ tests.
Ivan Sutherland's MIT dissertation on Sketchpad introduced the idea of interactive graphics into computing.
Edward A. Feigenbaum & Julian Feldman published Computers and Thought, the first collection of articles about artificial intelligence.
1964
Danny Bobrow's dissertation at MIT (tech.report #1 from MIT's AI group, Project MAC), shows that computers can understand natural language well enough to solve algebra word problems correctly.
Bert Raphael's MIT dissertation on the SIR program demonstrates the power of a logical representation of knowledge for question-answering systems
1965
J. Alan Robinson invented a mechanical proof procedure, the Resolution Method, which allowed programs to work efficiently with formal logic as a representation language.
Joseph Weizenbaum (MIT) built ELIZA, an interactive program that carries on a dialogue in English on any topic. It was a popular toy at AI centers on the ARPA-net when a version that "simulated" the dialogue of a psychotherapist was programmed.
1966
Ross Quillian (PhD dissertation, Carnegie Inst. of Technology; now CMU) demonstrated semantic nets.
First Machine Intelligence workshop at Edinburgh - the first of an influential annual series organized by Donald Michie and others.
Negative report on machine translation kills much work in Natural Language Processing (NLP) for many years.
1967
Dendral program (Edward Feigenbaum, Joshua Lederberg, Bruce Buchanan, Georgia Sutherland at Stanford) demonstrated to interpret mass spectra on organic chemical compounds. First successful knowledge-based program for scientific reasoning.
Joel Moses (PhD work at MIT) demonstrated the power of symbolic reasoning for integration problems in the Macsyma program. First successful knowledge-based program in mathematics.
Richard Greenblatt at MIT built a knowledge-based chess-playing program, MacHack, that was good enough to achieve a class-C rating in tournament play.
Late 60s
Doug Engelbart invented the mouse at SRI.
1968
Marvin Minsky & Seymour Papert publish Perceptrons, demonstrating limits of simple neural nets.
1969
SRI robot, Shakey, demonstrated combining locomotion, perception and problem solving.
Roger Schank (Stanford) defined conceptual dependency model for natural language understanding. Later developed (in PhD dissertations at Yale) for use in story understanding by Robert Wilensky and Wendy Lehnert, and for use in understanding memory by Janet Kolodner.
First International Joint Conference on Artificial Intelligence (IJCAI) held in Washington, D.C.
1970
Jaime Carbonell (Sr.) developed SCHOLAR, an interactive program for computer-aided instruction based on semantic nets as the representation of knowledge.
Bill Woods described Augmented Transition Networks (ATN's) as a representation for natural language understanding.
Patrick Winston's PhD program, ARCH, at MIT learned concepts from examples in the world of children's blocks.
Early 70's
Jane Robinson & Don Walker established influential Natural Language Processing group at SRI.
1971
Terry Winograd's PhD thesis (MIT) demonstrated the ability of computers to understand English sentences in a restricted world of children's blocks, in a coupling of his language understanding program, SHRDLU, with a robot arm that carried out instructions typed in English.
1972
Prolog developed by Alain Colmerauer.
1973
The Assembly Robotics group at Edinburgh University builds Freddy, the Famous Scottish Robot, capable of using vision to locate and assemble models.
1974
Ted Shortliffe's PhD dissertation on MYCIN (Stanford) demonstrated the power of rule-based systems for knowledge representation and inference in the domain of medical diagnosis and therapy. Sometimes called the first expert system.
Earl Sacerdoti developed one of the first planning programs, ABSTRIPS, and developed techniques of hierarchical planning.
1975
Marvin Minsky published his widely-read and influential article on Frames as a representation of knowledge, in which many ideas about schemas and semantic links are brought together.
The Meta-Dendral learning program produced new results in chemistry (some rules of mass spectrometry) the first scientific discoveries by a computer to be published in a refereed journal.
Mid 70's
Barbara Grosz (SRI) established limits to traditional AI approaches to discourse modeling. Subsequent work by Grosz, Bonnie Webber and Candace Sidner developed the notion of "centering", used in establishing focus of discourse and anaphoric references in NLP.
Alan Kay and Adele Goldberg (Xerox PARC) developed the Smalltalk language, establishing the power of object-oriented programming and of icon-oriented interfaces.
David Marr and MIT colleagues describe the "primal sketch" and its role in visual perception.
1976
Doug Lenat's AM program (Stanford PhD dissertation) demonstrated the discovery model (loosely-guided search for interesting conjectures).
Randall Davis demonstrated the power of meta-level reasoning in his PhD dissertation at Stanford.
Late 70's
Stanford's SUMEX-AIM resource, headed by Ed Feigenbaum and Joshua Lederberg, demonstrates the power of the ARPAnet for scientific collaboration.
1978
Tom Mitchell, at Stanford, invented the concept of Version Spaces for describing the search space of a concept formation program.
Herb Simon wins the Nobel Prize in Economics for his theory of bounded rationality, one of the cornerstones of AI known as "satisficing".
The MOLGEN program, written at Stanford by Mark Stefik and Peter Friedland, demonstrated that an object-oriented representation of knowledge can be used to plan gene-cloning experiments.
1979
Bill VanMelle's PhD dissertation at Stanford demonstrated the generality of MYCIN's representation of knowledge and style of reasoning in his EMYCIN program, the model for many commercial expert system "shells".
Jack Myers and Harry Pople at University of Pittsburgh developed INTERNIST, a knowledge-based medical diagnosis program based on Dr.Myers' clinical knowledge.
Cordell Green, David Barstow, Elaine Kant and others at Stanford demonstrated the CHI system for automatic programming.
The Stanford Cart, built by Hans Moravec, becomes the first computer-controlled, autonomous vehicle when it successfully traverses a chair-filled room and circumnavigates the Stanford AI Lab.
Drew McDermott & Jon Doyle at MIT, and John McCarthy at Stanford begin publishing work on non-monotonic logics and formal aspects of truth maintenance.
1980's
Lisp Machines developed and marketed.
First expert system shells and commercial applications.
1980
Lee Erman, Rick Hayes-Roth, Victor Lesser and Raj Reddy published the first description of the blackboard model, as the framework for the HEARSAY-II speech understanding system.
First National Conference of the American Association of Artificial Intelligence (AAAI) held at Stanford.
1981
Danny Hillis designs the connection machine, a massively parallel architecture that brings new power to AI, and to computation in general. (Later founds Thinking Machines, Inc.)
1983
John Laird & Paul Rosenbloom, working with Allen Newell, complete CMU dissertations on SOAR.
James Allen invents the Interval Calculus, the first widely used formalization of temporal events.
Mid 80's
Neural Networks become widely used with the Backpropagation algorithm (first described by Werbos in 1974).
1985
The autonomous drawing program, Aaron, created by Harold Cohen, is demonstrated at the AAAI National Conference (based on more than a decade of work, and with subsequent work showing major developments).
1987
Marvin Minsky publishes The Society of Mind, a theoretical description of the mind as a collection of cooperating agents.
1989
Dean Pomerleau at CMU creates ALVINN (An Autonomous Land Vehicle in a Neural Network), which grew into the system that drove a car coast-to-coast under computer control for all but about 50 of the 2850 miles.
1990's
Major advances in all areas of AI, with significant demonstrations in machine learning, intelligent tutoring, case-based reasoning, multi-agent planning, scheduling, uncertain reasoning, data mining, natural language understanding and translation, vision, virtual reality, games, and other topics.
Rod Brooks' COG Project at MIT, with numerous collaborators, makes significant progress in building a humanoid robot
Early 90's
TD-Gammon, a backgammon program written by Gerry Tesauro, demonstrates that reinforcement learning is powerful enough to create a championship-level game-playing program by competing favorably with world-class players.
1997
The Deep Blue chess program beats the current world chess champion, Garry Kasparov, in a widely followed match.
First official Robo-Cup soccer match featuring table-top matches with 40 teams of interacting robots and over 5000 spectators.
Late 90's
Web crawlers and other AI-based information extraction programs become essential in widespread use of the world-wide-web.
Demonstration of an Intelligent Room and Emotional Agents at MIT's AI Lab. Initiation of work on the Oxygen Architecture, which connects mobile and stationary computers in an adaptive network.
2000
Interactive robot pets (a.k.a. "smart toys") become commercially available, realizing the vision of the 18th cen. novelty toy makers.
Cynthia Breazeal at MIT publishes her dissertation on Sociable Machines, describing KISMET, a robot with a face that expresses emotions.
The Nomad robot explores remote regions of Antarctica looking for meteorite samples.
Basic Questions
Q. What is artificial intelligence?
A. 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.
Q. Yes, but what is intelligence?
A. 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.
Q. Isn't there a solid definition of intelligence that doesn't depend on relating it to human intelligence?
A. 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.
Q. Is intelligence a single thing so that one can ask a yes or no question ``Is this machine intelligent or not?''?
A. No. Intelligence involves mechanisms, and AI research has discovered how to make computers carry out some of them and not others. If doing a task requires only mechanisms that are well understood today, computer programs can give very impressive performances on these tasks. Such programs should be considered ``somewhat intelligent''.
Q. Isn't AI about simulating human intelligence?
A. 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.
Q. What about IQ? Do computer programs have IQs?
A. No. IQ is based on the rates at which intelligence develops in children. It is the ratio of the age at which a child normally makes a certain score to the child's age. The scale is extended to adults in a suitable way. IQ correlates well with various measures of success or failure in life, but making computers that can score high on IQ tests would be weakly correlated with their usefulness. For example, the ability of a child to repeat back a long sequence of digits correlates well with other intellectual abilities, perhaps because it measures how much information the child can compute with at once. However, ``digit span'' is trivial for even extremely limited computers.
However, some of the problems on IQ tests are useful challenges for AI.
Q. What about other comparisons between human and computer intelligence?
A.Arthur R. Jensen [Jen98], a leading researcher in human intelligence, suggests ``as a heuristic hypothesis'' that all normal humans have the same intellectual mechanisms and that differences in intelligence are related to ``quantitative biochemical and physiological conditions''. I see them as speed, short term memory, and the ability to form accurate and retrievable long term memories.
Whether or not Jensen is right about human intelligence, the situation in AI today is the reverse.
Computer programs have plenty of speed and memory but their abilities correspond to the intellectual mechanisms that program designers understand well enough to put in programs. Some abilities that children normally don't develop till they are teenagers may be in, and some abilities possessed by two year olds are still out. The matter is further complicated by the fact that the cognitive sciences still have not succeeded in determining exactly what the human abilities are. Very likely the organization of the intellectual mechanisms for AI can usefully be different from that in people.
Whenever people do better than computers on some task or computers use a lot of computation to do as well as people, this demonstrates that the program designers lack understanding of the intellectual mechanisms required to do the task efficiently.
Q. When did AI research start?
A. 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.
Q. Does AI aim to put the human mind into the computer?
A. 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.
Q. What is the Turing test?
A. Alan Turing's 1950 article Computing Machinery and Intelligence [Tur50] discussed conditions for considering a machine to be intelligent. He argued that if the machine could successfully pretend to be human to a knowledgeable observer then you certainly should consider it intelligent. This test would satisfy most people but not all philosophers. The observer could interact with the machine and a human by teletype (to avoid requiring that the machine imitate the appearance or voice of the person), and the human would try to persuade the observer that it was human and the machine would try to fool the observer.
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.
Daniel Dennett's book Brainchildren [Den98] has an excellent discussion of the Turing test and the various partial Turing tests that have been implemented, i.e. with restrictions on the observer's knowledge of AI and the subject matter of questioning. It turns out that some people are easily led into believing that a rather dumb program is intelligent.
Q. Does AI aim at human-level intelligence?
A. Yes. The ultimate effort is to make computer programs that can solve problems and achieve goals in the world as well as humans. However, many people involved in particular research areas are much less ambitious.
Q. How far is AI from reaching human-level intelligence? When will it happen?
A. A few people think that human-level intelligence can be achieved by writing large numbers of programs of the kind people are now writing and assembling vast knowledge bases of facts in the languages now used for expressing knowledge.
However, most AI researchers believe that new fundamental ideas are required, and therefore it cannot be predicted when human level intelligence will be achieved.
Q. Are computers the right kind of machine to be made intelligent?
A. Computers can be programmed to simulate any kind of machine.
Many researchers invented non-computer machines, hoping that they would be intelligent in different ways than the computer programs could be. However, they usually simulate their invented machines on a computer and come to doubt that the new machine is worth building. Because many billions of dollars that have been spent in making computers faster and faster, another kind of machine would have to be very fast to perform better than a program on a computer simulating the machine.
Q. Are computers fast enough to be intelligent?
A. Some people think much faster computers are required as well as new ideas. My own opinion is that the computers of 30 years ago were fast enough if only we knew how to program them. Of course, quite apart from the ambitions of AI researchers, computers will keep getting faster.
Q. What about parallel machines?
A. Machines with many processors are much faster than single processors can be. Parallelism itself presents no advantages, and parallel machines are somewhat awkward to program. When extreme speed is required, it is necessary to face this awkwardness.
experience?
A. This idea has been proposed many times, starting in the 1940s. Eventually, it will be made to work. However, AI programs haven't yet reached the level of being able to learn much of what a child learns fromQ. What about making a ``child machine'' that could improve by reading and by learning from physical experience. Nor do present programs understand language well enough to learn much by reading.
Q. Might an AI system be able to bootstrap itself to higher and higher level intelligence by thinking about AI?
A. I think yes, but we aren't yet at a level of AI at which this process can begin.
Q. What about chess?
A. Alexander Kronrod, a Russian AI researcher, said ``Chess is the Drosophila of AI.'' He was making an analogy with geneticists' use of that fruit fly to study inheritance. Playing chess requires certain intellectual mechanisms and not others. Chess programs now play at grandmaster level, but they do it with limited intellectual mechanisms compared to those used by a human chess player, substituting large amounts of computation for understanding. Once we understand these mechanisms better, we can build human-level chess programs that do far less computation than do present programs.
Unfortunately, the competitive and commercial aspects of making computers play chess have taken precedence over using chess as a scientific domain. It is as if the geneticists after 1910 had organized fruit fly races and concentrated their efforts on breeding fruit flies that could win these races.
Q. What about Go?
A. The Chinese and Japanese game of Go is also a board game in which the players take turns moving. Go exposes the weakness of our present understanding of the intellectual mechanisms involved in human game playing. Go programs are very bad players, in spite of considerable effort (not as much as for chess). The problem seems to be that a position in Go has to be divided mentally into a collection of subpositions which are first analyzed separately followed by an analysis of their interaction. Humans use this in chess also, but chess programs consider the position as a whole. Chess programs compensate for the lack of this intellectual mechanism by doing thousands or, in the case of Deep Blue, many millions of times as much computation.
Sooner or later, AI research will overcome this scandalous weakness.
Q. Don't some people say that AI is a bad idea?
A. The philosopher John Searle says that the idea of a non-biological machine being intelligent is incoherent. He proposes the Chinese room argument www-formal.stanford.edu/jmc/chinese.html The philosopher Hubert Dreyfus says that AI is impossible. The computer scientist Joseph Weizenbaum says the idea is obscene, anti-human and immoral. Various people have said that since artificial intelligence hasn't reached human level by now, it must be impossible. Still other people are disappointed that companies they invested in went bankrupt.
Q. Aren't computability theory and computational complexity the keys to AI? [Note to the layman and beginners in computer science: These are quite technical branches of mathematical logic and computer science, and the answer to the question has to be somewhat technical.]
A. No. These theories are relevant but don't address the fundamental problems of AI.
In the 1930s mathematical logicians, especially Kurt Gödel and Alan Turing, established that there did not exist algorithms that were guaranteed to solve all problems in certain important mathematical domains. Whether a sentence of first order logic is a theorem is one example, and whether a polynomial equations in several variables has integer solutions is another. Humans solve problems in these domains all the time, and this has been offered as an argument (usually with some decorations) that computers are intrinsically incapable of doing what people do. Roger Penrose claims this. However, people can't guarantee to solve arbitrary problems in these domains either. See my Review of The Emperor's New Mind by Roger Penrose. More essays and reviews defending AI research are in In the 1960s computer scientists, especially Steve Cook and Richard Karp developed the theory of NP-complete problem domains. Problems in these domains are solvable, but seem to take time exponential in the size of the problem. Which sentences of propositional calculus are satisfiable is a basic example of an NP-complete problem domain. Humans often solve problems in NP-complete domains in times much shorter than is guaranteed by the general algorithms, but can't solve them quickly in general.
What is important for AI is to have algorithms as capable as people at solving problems. The identification of subdomains for which good algorithms exist is important, but a lot of AI problem solvers are not associated with readily identified subdomains.
The theory of the difficulty of general classes of problems is called computational complexity. So far this theory hasn't interacted with AI as much as might have been hoped. Success in problem solving by humans and by AI programs seems to rely on properties of problems and problem solving methods that the neither the complexity researchers nor the AI community have been able to identify precisely.
Algorithmic complexity theory as developed by Solomonoff, Kolmogorov and Chaitin (independently of one another) is also relevant. It defines the complexity of a symbolic object as the length of the shortest program that will generate it. Proving that a candidate program is the shortest or close to the shortest is an unsolvable problem, but representing objects by short programs that generate them should sometimes be illuminating even when you can't prove that the program is the shortest.
Branches of AI
Q. What are the branches of AI?
A. Here's a list, but some branches are surely missing, because no-one has identified them yet. Some of these may be regarded as concepts or topics rather than full branches.
logical AI
What a program knows about the world in general the facts of the specific situation in which it must act, and its goals are all represented by sentences of some mathematical logical language. The program decides what to do by inferring that certain actions are appropriate for achieving its goals. The first article proposing this was [McC59]. [McC89] is a more recent summary. [McC96b] lists some of the concepts involved in logical aI. [Sha97] is an important text.
search
AI programs often examine large numbers of possibilities, e.g. moves in a chess game or inferences by a theorem proving program. Discoveries are continually made about how to do this more efficiently in various domains.
pattern recognition
When a program makes observations of some kind, it is often programmed to compare what it sees with a pattern. For example, a vision program may try to match a pattern of eyes and a nose in a scene in order to find a face. More complex patterns, e.g. in a natural language text, in a chess position, or in the history of some event are also studied. These more complex patterns require quite different methods than do the simple patterns that have been studied the most.
representation
Facts about the world have to be represented in some way. Usually languages of mathematical logic are used.
inference
From some facts, others can be inferred. Mathematical logical deduction is adequate for some purposes, but new methods of non-monotonic inference have been added to logic since the 1970s. The simplest kind of non-monotonic reasoning is default reasoning in which a conclusion is to be inferred by default, but the conclusion can be withdrawn if there is evidence to the contrary. For example, when we hear of a bird, we man infer that it can fly, but this conclusion can be reversed when we hear that it is a penguin. It is the possibility that a conclusion may have to be withdrawn that constitutes the non-monotonic character of the reasoning. Ordinary logical reasoning is monotonic in that the set of conclusions that can the drawn from a set of premises is a monotonic increasing function of the premises. Circumscription is another form of non-monotonic reasoning.
common sense knowledge and reasoning
This is the area in which AI is farthest from human-level, in spite of the fact that it has been an active research area since the 1950s. While there has been considerable progress, e.g. in developing systems of non-monotonic reasoning and theories of action, yet more new ideas are needed. The Cyc system contains a large but spotty collection of common sense facts.
learning from experience
Programs do that. The approaches to AI based on connectionism and neural nets specialize in that. There is also learning of laws expressed in logic. [Mit97] is a comprehensive undergraduate text on machine learning. Programs can only learn what facts or behaviors their formalisms can represent, and unfortunately learning systems are almost all based on very limited abilities to represent information.
planning
Planning programs start with general facts about the world (especially facts about the effects of actions), facts about the particular situation and a statement of a goal. From these, they generate a strategy for achieving the goal. In the most common cases, the strategy is just a sequence of actions.
epistemology
This is a study of the kinds of knowledge that are required for solving problems in the world.
ontology
Ontology is the study of the kinds of things that exist. In AI, the programs and sentences deal with various kinds of objects, and we study what these kinds are and what their basic properties are. Emphasis on ontology begins in the 1990s.
heuristics
A heuristic is a way of trying to discover something or an idea imbedded in a program. The term is used variously in AI. Heuristic functions are used in some approaches to search to measure how far a node in a search tree seems to be from a goal. Heuristic predicates that compare two nodes in a search tree to see if one is better than the other, i.e. constitutes an advance toward the goal, may be more useful.
Applications of AI
Q. What are the applications of AI?
A. Here are some.
game playing
You can buy machines that can play master level chess for a few hundred dollars. There is some AI in them, but they play well against people mainly through brute force computation--looking at hundreds of thousands of positions. To beat a world champion by brute force and known reliable heuristics requires being able to look at 200 million positions per second.
speech recognition
In the 1990s, computer speech recognition reached a practical level for limited purposes. Thus United Airlines has replaced its keyboard tree for flight information by a system using speech recognition of flight numbers and city names. It is quite convenient. On the the other hand, while it is possible to instruct some computers using speech, most users have gone back to the keyboard and the mouse as still more convenient.
understanding natural language
Just getting a sequence of words into a computer is not enough. Parsing sentences is not enough either. The computer has to be provided with an understanding of the domain the text is about, and this is presently possible only for very limited domains.
computer vision
The world is composed of three-dimensional objects, but the inputs to the human eye and computers' TV cameras are two dimensional. Some useful programs can work solely in two dimensions, but full computer vision requires partial three-dimensional information that is not just a set of two-dimensional views. At present there are only limited ways of representing three-dimensional information directly, and they are not as good as what humans evidently use.
expert systems
A ``knowledge engineer'' interviews experts in a certain domain and tries to embody their knowledge in a computer program for carrying out some task. How well this works depends on whether the intellectual mechanisms required for the task are within the present state of AI. When this turned out not to be so, there were many disappointing results. One of the first expert systems was MYCIN in 1974, which diagnosed bacterial infections of the blood and suggested treatments. It did better than medical students or practicing doctors, provided its limitations were observed. Namely, its ontology included bacteria, symptoms, and treatments and did not include patients, doctors, hospitals, death, recovery, and events occurring in time. Its interactions depended on a single patient being considered. Since the experts consulted by the knowledge engineers knew about patients, doctors, death, recovery, etc., it is clear that the knowledge engineers forced what the experts told them into a predetermined framework. In the present state of AI, this has to be true. The usefulness of current expert systems depends on their users having common sense.
heuristic classification
One of the most feasible kinds of expert system given the present knowledge of AI is to put some information in one of a fixed set of categories using several sources of information. An example is advising whether to accept a proposed credit card purchase. Information is available about the owner of the credit card, his record of payment and also about the item he is buying and about the establishment from which he is buying it (e.g., about whether there have been previous credit card frauds at this establishment).
More questions
Q. How is AI research done?
A. AI research has both theoretical and experimental sides. The experimental side has both basic and applied aspects.
There are two main lines of research. One is biological, based on the idea that since humans are intelligent, AI should study humans and imitate their psychology or physiology. The other is phenomenal, based on studying and formalizing common sense facts about the world and the problems that the world presents to the achievement of goals. The two approaches interact to some extent, and both should eventually succeed. It is a race, but both racers seem to be walking.
Q. What are the relations between AI and philosophy?
A. AI has many relations with philosophy, especially modern analytic philosophy. Both study mind, and both study common sense. The best best reference is [Tho03].
Q. What should I study before or while learning AI?
A. Study mathematics, especially mathematical logic. The more you learn about science in general the better. For the biological approaches to AI, study psychology and the physiology of the nervous system. Learn some programming languages--at least C, Lisp and Prolog. It is also a good idea to learn one basic machine language. Jobs are likely to depend on knowing the languages currently in fashion. In the late 1990s, these include C++ and Java.
Q. What is a good textbook on AI?
A. Artificial Intelligence by Stuart Russell and Peter Norvig, Prentice Hall is the most commonly used textbbook in 1997. The general views expressed there do not exactly correspond to those of this essay. Artificial Intelligence: A New Synthesis by Nils Nilsson, Morgan Kaufman, may be easier to read. Some people prefer Computational Intelligence by David Poole, Alan Mackworth and Randy Goebel, Oxford, 1998.
Q. What organizations and publications are concerned with AI?
A. The American Association for Artificial Intelligence (AAAI), the European Coordinating Committee for Artificial Intelligence (ECCAI) and the Society for Artificial Intelligence and Simulation of Behavior (AISB) are scientific societies concerned with AI research. The Association for Computing Machinery (ACM) has a special interest group on artificial intelligence SIGART.
The International Joint Conference on AI (IJCAI) is the main international conference. The AAAI runs a US National Conference on AI. Electronic Transactions on Artificial Intelligence, Artificial Intelligence, and Journal of Artificial Intelligence Research, and IEEE Transactions on Pattern Analysis and Machine Intelligence are four of the main journals publishing AI research papers. I have not yet found everything that should be in this paragraph.
Page of Positive Reviews lists papers that experts have found important.
Funding a Revolution: Government Support for Computing Research by a committee of the National Research covers support for AI research.
Bibliography
Den98
Daniel Dennett. Brainchildren: Essays on Designing Minds. MIT Press, 1998.
Jen98
Arthur R. Jensen. Does IQ matter? Commentary, pages 20-21, November 1998. The reference is just to Jensen's comment--one of many.
McC59
John McCarthy. Programs with Common Sense. In Mechanisation of Thought Processes, Proceedings of the Symposium of the National Physics Laboratory, pages 77-84, London, U.K., 1959. Her Majesty's Stationery Office. Reprinted in [McC90].
McC89
John McCarthy. Artificial Intelligence, Logic and Formalizing Common Sense. In Richmond Thomason, editor, Philosophical Logic and Artificial Intelligence. Klüver Academic, 1989.
McC90
John McCarthy. Formalizing Common Sense: Papers by John McCarthy. Ablex Publishing Corporation, 1990.
McC96a
John McCarthy. Defending AI research : a collection of essays and reviews. CSLI lecture notes: no. 49. Center for the Study of Language and Information, 1996. distributed by Cambridge University Press.
McC96b
John McCarthy. Concepts of Logical AI, 1996. Web only for now but may be referenced.
Mit97
Tom Mitchell. Machine Learning. McGraw-Hill, 1997.
Sha97
Murray Shanahan. Solving the Frame Problem, a mathematical investigation of the common sense law of inertia. M.I.T. Press, 1997.
Tho03
Richmond Thomason. Logic and artificial intelligence. In Edward N. Zalta, editor, The Stanford Encyclopedia of Philosophy. Fall 2003. http://plato.stanford.edu/archives/fall2003/entries/logic-ai/.
Tur50
Alan Turing. Computing machinery and intelligence. Mind, 1950.
An Overview.AI is a result of the merge of philosophy, mathematics, psychology, neurology, linguistics, computer science, and many other fields. Futhermore, the application of AI relates to almost any fields. This variety gives AI an endless potential. A relatively young science, AI has made much progress in 50 years. Though fast-growing, AI has never actually caught up with all the expectation imposed on it. There are two reasons for public's over-confidence in AI. First, AI theories are often ingenious and subtle even fictional, implying much futuristic applications. Second, AI, being incorporated with computer technology, is often expected to progress as fast as the computer technology. Conclusionally, AI is a young, energetic, and attractive science.
EXAMPLES
The branch of computer science concerned with making computers behave like humans. The term was coined in 1956 by John McCarthy at the Massachusetts Institute of Technology. Artificial intelligence includes
games playing: programming computers to play games such as chess and checkers
expert systems: programming computers to make decisions in real-life situations (for example, some expert systems help doctors diagnose diseases based on symptoms)
natural language: programming computers to understand natural human languages
neural networks: Systems that simulate intelligence by attempting to reproduce the types of physical connections that occur in animal brains
robotics: programming computers to see and hear and react to other sensory stimuli Currently, no computers exhibit full artificial intelligence (that is, are able to simulate human behavior). The greatest advances have occurred in the field of games playing. The best computer chess programs are now capable of beating humans. In May, 1997, an IBM super-computer called Deep Blue defeated world chess champion Gary Kasparov in a chess match.
In the area of robotics, computers are now widely used in assembly plants, but they are capable only of very limited tasks. Robots have great difficulty identifying objects based on appearance or feel, and they still move and handle objects clumsily.
Natural-language processing offers the greatest potential rewards because it would allow people to interact with computers without needing any specialized knowledge. You could simply walk up to a computer and talk to it. Unfortunately, programming computers to understand natural languages has proved to be more difficult than originally thought. Some rudimentary translation systems that translate from one human language to another are in existence, but they are not nearly as good as human translators. There are also voice recognition systems that can convert spoken sounds into written words, but they do not understand what they are writing; they simply take dictation. Even these systems are quite limited -- you must speak slowly and distinctly.
In the early 1980s, expert systems were believed to represent the future of artificial intelligence and of computers in general. To date, however, they have not lived up to expectations. Many expert systems help human experts in such fields as medicine and engineering, but they are very expensive to produce and are helpful only in special situations.
Today, the hottest area of artificial intelligence is neural networks, which are proving successful in a number of disciplines such as voice recognition and natural-language processing.
INTRODUCTION
The intellectual roots of AI, and the concept of intelligent machines, may be found in Greek mythology. Intelligent artifacts appear in literature since then, with real (and fraudulent) mechanical devices actually demonstrated to behave with some degree of intelligence. Some of these conceptual achievements are listed below under "Ancient History."
After modern computers became available, following World War II, it has become possible to create programs that perform difficult intellectual tasks. From these programs, general tools are constructed which have applications in a wide variety of everday problems. Some of these computational milestones are listed below under "Modern History."
ANCIENT HISTORY
Greek myths of Hephaestus and Pygmalion incorporate the idea of intelligent robots. Many other myths in antiquity involve human-like artifacts. Many mechanical toys and models were actually constructed, e.g., by Hero, Daedalus and other real persons.
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 Albert the Great reputedly among the owners.
Ramon Llull, Spanish theologian, invented machines for discovering nonmathematical truths through combinatories.
15th century
Invention of printing using moveable type. Gutenberg Bible printed (1456).
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.
Rabbi Loew of Prague is said to have invented the Golem, a clay man brought to life (1580).
17th century
Early in the century, Descartes proposed that bodies of animals are nothing more than complex machines. Many other 17th century thinkers offered 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 evisioned a universal calculus of reasoning by which arguments could be decided mechanically.
18th century
The 18th century saw a profusion of mechanical toys, including the celebrated mechanical duck of Vaucanson and von Kempelen's phony mechanical chess player, The Turk (1769).
19th century
Luddites (led by Ned Ludd) destroyed machinery in England (1811-1816).
Mary Shelley published the story of Frankenstein's monster (1818).
George Boole developed a binary algebra representing (some) "laws of thought."
Charles Babbage & Ada Byron (Lady Lovelace) worked on programmable mechanical calculating machines.
20th century - First Half
Bertrand Russell and Alfred North Whitehead published Principia Mathematica, which revolutionaized formal logic. Russell, Ludwig Wittgenstein, and Rudolf Carnap lead philosophy into logical analysis of knowledge.
Karel Capek's play "R.U.R." (Rossum's Universal Robots) opens in London (1923). - First use of the word 'robot' in English.
Warren McCulloch & Walter Pitts publish "A Logical Calculus of the Ideas Immanent in Nervous Activity" (1943), laying foundations for neural networks.
Arturo Rosenblueth, Norbert Wiener & Julian Bigelow coin the term "cybernetics" in a 1943 paper. Wiener's popular book by that name published in 1948.
Vannevar Bush published As We May Think (Atlantic Monthly, July 1945) a prescient vision of the future in which computers assist humans in many activities.
A.M. Turing published "Computing Machinery and Intelligence" (1950). - Introduction of Turing's Test as a way of operationalizing a test of intelligent behavior.
Claude Shannon published detailed analysis of chess playing as search (1950).
Isaac Asimov published his three laws of robotics (1950).
MODERN HISTORY
1956
John McCarthy coined the term "artificial intelligence" as the topic of the Dartmouth Conference, the first conference devoted to the subject.
Demonstration of the first running AI program, the Logic Theorist (LT) written by Allen Newell, J.C. Shaw and Herbert Simon (Carnegie Institute of Technology, now Carnegie Mellon University).
1957
The General Problem Solver (GPS) demonstrated by Newell, Shaw & Simon.
1952-62
Arthur Samuel (IBM) wrote the first game-playing program, for checkers, to achieve sufficient skill to challenge a world champion. Samuel's machine learning programs were responsible for the high performance of the checkers player.
1958
John McCarthy (MIT) invented the Lisp language.
Herb Gelernter & Nathan Rochester (IBM) described a theorem prover in geometry that exploits a semantic model of the domain in the form of diagrams of "typical" cases.
Teddington Conference on the Mechanization of Thought Processes was held in the UK and among the papers presented were John McCarthy's "Programs with Common Sense," Oliver Selfridge's "Pandemonium," and Marvin Minsky's "Some Methods of Heuristic Programming and Artificial Intelligence."
Late 50's & Early 60's
Margaret Masterman & colleagues at Cambridge design semantic nets for machine translation.
1961
James Slagle (PhD dissertation, MIT) wrote (in Lisp) the first symbolic integration program, SAINT, which solved calculus problems at the college freshman level.
1962
First industrial robot company, Unimation, founded.
1963
Thomas Evans' program, ANALOGY, written as part of his PhD work at MIT, demonstrated that computers can solve the same analogy problems as are given on IQ tests.
Ivan Sutherland's MIT dissertation on Sketchpad introduced the idea of interactive graphics into computing.
Edward A. Feigenbaum & Julian Feldman published Computers and Thought, the first collection of articles about artificial intelligence.
1964
Danny Bobrow's dissertation at MIT (tech.report #1 from MIT's AI group, Project MAC), shows that computers can understand natural language well enough to solve algebra word problems correctly.
Bert Raphael's MIT dissertation on the SIR program demonstrates the power of a logical representation of knowledge for question-answering systems
1965
J. Alan Robinson invented a mechanical proof procedure, the Resolution Method, which allowed programs to work efficiently with formal logic as a representation language.
Joseph Weizenbaum (MIT) built ELIZA, an interactive program that carries on a dialogue in English on any topic. It was a popular toy at AI centers on the ARPA-net when a version that "simulated" the dialogue of a psychotherapist was programmed.
1966
Ross Quillian (PhD dissertation, Carnegie Inst. of Technology; now CMU) demonstrated semantic nets.
First Machine Intelligence workshop at Edinburgh - the first of an influential annual series organized by Donald Michie and others.
Negative report on machine translation kills much work in Natural Language Processing (NLP) for many years.
1967
Dendral program (Edward Feigenbaum, Joshua Lederberg, Bruce Buchanan, Georgia Sutherland at Stanford) demonstrated to interpret mass spectra on organic chemical compounds. First successful knowledge-based program for scientific reasoning.
Joel Moses (PhD work at MIT) demonstrated the power of symbolic reasoning for integration problems in the Macsyma program. First successful knowledge-based program in mathematics.
Richard Greenblatt at MIT built a knowledge-based chess-playing program, MacHack, that was good enough to achieve a class-C rating in tournament play.
Late 60s
Doug Engelbart invented the mouse at SRI.
1968
Marvin Minsky & Seymour Papert publish Perceptrons, demonstrating limits of simple neural nets.
1969
SRI robot, Shakey, demonstrated combining locomotion, perception and problem solving.
Roger Schank (Stanford) defined conceptual dependency model for natural language understanding. Later developed (in PhD dissertations at Yale) for use in story understanding by Robert Wilensky and Wendy Lehnert, and for use in understanding memory by Janet Kolodner.
First International Joint Conference on Artificial Intelligence (IJCAI) held in Washington, D.C.
1970
Jaime Carbonell (Sr.) developed SCHOLAR, an interactive program for computer-aided instruction based on semantic nets as the representation of knowledge.
Bill Woods described Augmented Transition Networks (ATN's) as a representation for natural language understanding.
Patrick Winston's PhD program, ARCH, at MIT learned concepts from examples in the world of children's blocks.
Early 70's
Jane Robinson & Don Walker established influential Natural Language Processing group at SRI.
1971
Terry Winograd's PhD thesis (MIT) demonstrated the ability of computers to understand English sentences in a restricted world of children's blocks, in a coupling of his language understanding program, SHRDLU, with a robot arm that carried out instructions typed in English.
1972
Prolog developed by Alain Colmerauer.
1973
The Assembly Robotics group at Edinburgh University builds Freddy, the Famous Scottish Robot, capable of using vision to locate and assemble models.
1974
Ted Shortliffe's PhD dissertation on MYCIN (Stanford) demonstrated the power of rule-based systems for knowledge representation and inference in the domain of medical diagnosis and therapy. Sometimes called the first expert system.
Earl Sacerdoti developed one of the first planning programs, ABSTRIPS, and developed techniques of hierarchical planning.
1975
Marvin Minsky published his widely-read and influential article on Frames as a representation of knowledge, in which many ideas about schemas and semantic links are brought together.
The Meta-Dendral learning program produced new results in chemistry (some rules of mass spectrometry) the first scientific discoveries by a computer to be published in a refereed journal.
Mid 70's
Barbara Grosz (SRI) established limits to traditional AI approaches to discourse modeling. Subsequent work by Grosz, Bonnie Webber and Candace Sidner developed the notion of "centering", used in establishing focus of discourse and anaphoric references in NLP.
Alan Kay and Adele Goldberg (Xerox PARC) developed the Smalltalk language, establishing the power of object-oriented programming and of icon-oriented interfaces.
David Marr and MIT colleagues describe the "primal sketch" and its role in visual perception.
1976
Doug Lenat's AM program (Stanford PhD dissertation) demonstrated the discovery model (loosely-guided search for interesting conjectures).
Randall Davis demonstrated the power of meta-level reasoning in his PhD dissertation at Stanford.
Late 70's
Stanford's SUMEX-AIM resource, headed by Ed Feigenbaum and Joshua Lederberg, demonstrates the power of the ARPAnet for scientific collaboration.
1978
Tom Mitchell, at Stanford, invented the concept of Version Spaces for describing the search space of a concept formation program.
Herb Simon wins the Nobel Prize in Economics for his theory of bounded rationality, one of the cornerstones of AI known as "satisficing".
The MOLGEN program, written at Stanford by Mark Stefik and Peter Friedland, demonstrated that an object-oriented representation of knowledge can be used to plan gene-cloning experiments.
1979
Bill VanMelle's PhD dissertation at Stanford demonstrated the generality of MYCIN's representation of knowledge and style of reasoning in his EMYCIN program, the model for many commercial expert system "shells".
Jack Myers and Harry Pople at University of Pittsburgh developed INTERNIST, a knowledge-based medical diagnosis program based on Dr.Myers' clinical knowledge.
Cordell Green, David Barstow, Elaine Kant and others at Stanford demonstrated the CHI system for automatic programming.
The Stanford Cart, built by Hans Moravec, becomes the first computer-controlled, autonomous vehicle when it successfully traverses a chair-filled room and circumnavigates the Stanford AI Lab.
Drew McDermott & Jon Doyle at MIT, and John McCarthy at Stanford begin publishing work on non-monotonic logics and formal aspects of truth maintenance.
1980's
Lisp Machines developed and marketed.
First expert system shells and commercial applications.
1980
Lee Erman, Rick Hayes-Roth, Victor Lesser and Raj Reddy published the first description of the blackboard model, as the framework for the HEARSAY-II speech understanding system.
First National Conference of the American Association of Artificial Intelligence (AAAI) held at Stanford.
1981
Danny Hillis designs the connection machine, a massively parallel architecture that brings new power to AI, and to computation in general. (Later founds Thinking Machines, Inc.)
1983
John Laird & Paul Rosenbloom, working with Allen Newell, complete CMU dissertations on SOAR.
James Allen invents the Interval Calculus, the first widely used formalization of temporal events.
Mid 80's
Neural Networks become widely used with the Backpropagation algorithm (first described by Werbos in 1974).
1985
The autonomous drawing program, Aaron, created by Harold Cohen, is demonstrated at the AAAI National Conference (based on more than a decade of work, and with subsequent work showing major developments).
1987
Marvin Minsky publishes The Society of Mind, a theoretical description of the mind as a collection of cooperating agents.
1989
Dean Pomerleau at CMU creates ALVINN (An Autonomous Land Vehicle in a Neural Network), which grew into the system that drove a car coast-to-coast under computer control for all but about 50 of the 2850 miles.
1990's
Major advances in all areas of AI, with significant demonstrations in machine learning, intelligent tutoring, case-based reasoning, multi-agent planning, scheduling, uncertain reasoning, data mining, natural language understanding and translation, vision, virtual reality, games, and other topics.
Rod Brooks' COG Project at MIT, with numerous collaborators, makes significant progress in building a humanoid robot
Early 90's
TD-Gammon, a backgammon program written by Gerry Tesauro, demonstrates that reinforcement learning is powerful enough to create a championship-level game-playing program by competing favorably with world-class players.
1997
The Deep Blue chess program beats the current world chess champion, Garry Kasparov, in a widely followed match.
First official Robo-Cup soccer match featuring table-top matches with 40 teams of interacting robots and over 5000 spectators.
Late 90's
Web crawlers and other AI-based information extraction programs become essential in widespread use of the world-wide-web.
Demonstration of an Intelligent Room and Emotional Agents at MIT's AI Lab. Initiation of work on the Oxygen Architecture, which connects mobile and stationary computers in an adaptive network.
2000
Interactive robot pets (a.k.a. "smart toys") become commercially available, realizing the vision of the 18th cen. novelty toy makers.
Cynthia Breazeal at MIT publishes her dissertation on Sociable Machines, describing KISMET, a robot with a face that expresses emotions.
The Nomad robot explores remote regions of Antarctica looking for meteorite samples.
Basic Questions
Q. What is artificial intelligence?
A. 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.
Q. Yes, but what is intelligence?
A. 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.
Q. Isn't there a solid definition of intelligence that doesn't depend on relating it to human intelligence?
A. 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.
Q. Is intelligence a single thing so that one can ask a yes or no question ``Is this machine intelligent or not?''?
A. No. Intelligence involves mechanisms, and AI research has discovered how to make computers carry out some of them and not others. If doing a task requires only mechanisms that are well understood today, computer programs can give very impressive performances on these tasks. Such programs should be considered ``somewhat intelligent''.
Q. Isn't AI about simulating human intelligence?
A. 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.
Q. What about IQ? Do computer programs have IQs?
A. No. IQ is based on the rates at which intelligence develops in children. It is the ratio of the age at which a child normally makes a certain score to the child's age. The scale is extended to adults in a suitable way. IQ correlates well with various measures of success or failure in life, but making computers that can score high on IQ tests would be weakly correlated with their usefulness. For example, the ability of a child to repeat back a long sequence of digits correlates well with other intellectual abilities, perhaps because it measures how much information the child can compute with at once. However, ``digit span'' is trivial for even extremely limited computers.
However, some of the problems on IQ tests are useful challenges for AI.
Q. What about other comparisons between human and computer intelligence?
A.Arthur R. Jensen [Jen98], a leading researcher in human intelligence, suggests ``as a heuristic hypothesis'' that all normal humans have the same intellectual mechanisms and that differences in intelligence are related to ``quantitative biochemical and physiological conditions''. I see them as speed, short term memory, and the ability to form accurate and retrievable long term memories.
Whether or not Jensen is right about human intelligence, the situation in AI today is the reverse.
Computer programs have plenty of speed and memory but their abilities correspond to the intellectual mechanisms that program designers understand well enough to put in programs. Some abilities that children normally don't develop till they are teenagers may be in, and some abilities possessed by two year olds are still out. The matter is further complicated by the fact that the cognitive sciences still have not succeeded in determining exactly what the human abilities are. Very likely the organization of the intellectual mechanisms for AI can usefully be different from that in people.
Whenever people do better than computers on some task or computers use a lot of computation to do as well as people, this demonstrates that the program designers lack understanding of the intellectual mechanisms required to do the task efficiently.
Q. When did AI research start?
A. 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.
Q. Does AI aim to put the human mind into the computer?
A. 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.
Q. What is the Turing test?
A. Alan Turing's 1950 article Computing Machinery and Intelligence [Tur50] discussed conditions for considering a machine to be intelligent. He argued that if the machine could successfully pretend to be human to a knowledgeable observer then you certainly should consider it intelligent. This test would satisfy most people but not all philosophers. The observer could interact with the machine and a human by teletype (to avoid requiring that the machine imitate the appearance or voice of the person), and the human would try to persuade the observer that it was human and the machine would try to fool the observer.
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.
Daniel Dennett's book Brainchildren [Den98] has an excellent discussion of the Turing test and the various partial Turing tests that have been implemented, i.e. with restrictions on the observer's knowledge of AI and the subject matter of questioning. It turns out that some people are easily led into believing that a rather dumb program is intelligent.
Q. Does AI aim at human-level intelligence?
A. Yes. The ultimate effort is to make computer programs that can solve problems and achieve goals in the world as well as humans. However, many people involved in particular research areas are much less ambitious.
Q. How far is AI from reaching human-level intelligence? When will it happen?
A. A few people think that human-level intelligence can be achieved by writing large numbers of programs of the kind people are now writing and assembling vast knowledge bases of facts in the languages now used for expressing knowledge.
However, most AI researchers believe that new fundamental ideas are required, and therefore it cannot be predicted when human level intelligence will be achieved.
Q. Are computers the right kind of machine to be made intelligent?
A. Computers can be programmed to simulate any kind of machine.
Many researchers invented non-computer machines, hoping that they would be intelligent in different ways than the computer programs could be. However, they usually simulate their invented machines on a computer and come to doubt that the new machine is worth building. Because many billions of dollars that have been spent in making computers faster and faster, another kind of machine would have to be very fast to perform better than a program on a computer simulating the machine.
Q. Are computers fast enough to be intelligent?
A. Some people think much faster computers are required as well as new ideas. My own opinion is that the computers of 30 years ago were fast enough if only we knew how to program them. Of course, quite apart from the ambitions of AI researchers, computers will keep getting faster.
Q. What about parallel machines?
A. Machines with many processors are much faster than single processors can be. Parallelism itself presents no advantages, and parallel machines are somewhat awkward to program. When extreme speed is required, it is necessary to face this awkwardness.
experience?
A. This idea has been proposed many times, starting in the 1940s. Eventually, it will be made to work. However, AI programs haven't yet reached the level of being able to learn much of what a child learns fromQ. What about making a ``child machine'' that could improve by reading and by learning from physical experience. Nor do present programs understand language well enough to learn much by reading.
Q. Might an AI system be able to bootstrap itself to higher and higher level intelligence by thinking about AI?
A. I think yes, but we aren't yet at a level of AI at which this process can begin.
Q. What about chess?
A. Alexander Kronrod, a Russian AI researcher, said ``Chess is the Drosophila of AI.'' He was making an analogy with geneticists' use of that fruit fly to study inheritance. Playing chess requires certain intellectual mechanisms and not others. Chess programs now play at grandmaster level, but they do it with limited intellectual mechanisms compared to those used by a human chess player, substituting large amounts of computation for understanding. Once we understand these mechanisms better, we can build human-level chess programs that do far less computation than do present programs.
Unfortunately, the competitive and commercial aspects of making computers play chess have taken precedence over using chess as a scientific domain. It is as if the geneticists after 1910 had organized fruit fly races and concentrated their efforts on breeding fruit flies that could win these races.
Q. What about Go?
A. The Chinese and Japanese game of Go is also a board game in which the players take turns moving. Go exposes the weakness of our present understanding of the intellectual mechanisms involved in human game playing. Go programs are very bad players, in spite of considerable effort (not as much as for chess). The problem seems to be that a position in Go has to be divided mentally into a collection of subpositions which are first analyzed separately followed by an analysis of their interaction. Humans use this in chess also, but chess programs consider the position as a whole. Chess programs compensate for the lack of this intellectual mechanism by doing thousands or, in the case of Deep Blue, many millions of times as much computation.
Sooner or later, AI research will overcome this scandalous weakness.
Q. Don't some people say that AI is a bad idea?
A. The philosopher John Searle says that the idea of a non-biological machine being intelligent is incoherent. He proposes the Chinese room argument www-formal.stanford.edu/jmc/chinese.html The philosopher Hubert Dreyfus says that AI is impossible. The computer scientist Joseph Weizenbaum says the idea is obscene, anti-human and immoral. Various people have said that since artificial intelligence hasn't reached human level by now, it must be impossible. Still other people are disappointed that companies they invested in went bankrupt.
Q. Aren't computability theory and computational complexity the keys to AI? [Note to the layman and beginners in computer science: These are quite technical branches of mathematical logic and computer science, and the answer to the question has to be somewhat technical.]
A. No. These theories are relevant but don't address the fundamental problems of AI.
In the 1930s mathematical logicians, especially Kurt Gödel and Alan Turing, established that there did not exist algorithms that were guaranteed to solve all problems in certain important mathematical domains. Whether a sentence of first order logic is a theorem is one example, and whether a polynomial equations in several variables has integer solutions is another. Humans solve problems in these domains all the time, and this has been offered as an argument (usually with some decorations) that computers are intrinsically incapable of doing what people do. Roger Penrose claims this. However, people can't guarantee to solve arbitrary problems in these domains either. See my Review of The Emperor's New Mind by Roger Penrose. More essays and reviews defending AI research are in In the 1960s computer scientists, especially Steve Cook and Richard Karp developed the theory of NP-complete problem domains. Problems in these domains are solvable, but seem to take time exponential in the size of the problem. Which sentences of propositional calculus are satisfiable is a basic example of an NP-complete problem domain. Humans often solve problems in NP-complete domains in times much shorter than is guaranteed by the general algorithms, but can't solve them quickly in general.
What is important for AI is to have algorithms as capable as people at solving problems. The identification of subdomains for which good algorithms exist is important, but a lot of AI problem solvers are not associated with readily identified subdomains.
The theory of the difficulty of general classes of problems is called computational complexity. So far this theory hasn't interacted with AI as much as might have been hoped. Success in problem solving by humans and by AI programs seems to rely on properties of problems and problem solving methods that the neither the complexity researchers nor the AI community have been able to identify precisely.
Algorithmic complexity theory as developed by Solomonoff, Kolmogorov and Chaitin (independently of one another) is also relevant. It defines the complexity of a symbolic object as the length of the shortest program that will generate it. Proving that a candidate program is the shortest or close to the shortest is an unsolvable problem, but representing objects by short programs that generate them should sometimes be illuminating even when you can't prove that the program is the shortest.
Branches of AI
Q. What are the branches of AI?
A. Here's a list, but some branches are surely missing, because no-one has identified them yet. Some of these may be regarded as concepts or topics rather than full branches.
logical AI
What a program knows about the world in general the facts of the specific situation in which it must act, and its goals are all represented by sentences of some mathematical logical language. The program decides what to do by inferring that certain actions are appropriate for achieving its goals. The first article proposing this was [McC59]. [McC89] is a more recent summary. [McC96b] lists some of the concepts involved in logical aI. [Sha97] is an important text.
search
AI programs often examine large numbers of possibilities, e.g. moves in a chess game or inferences by a theorem proving program. Discoveries are continually made about how to do this more efficiently in various domains.
pattern recognition
When a program makes observations of some kind, it is often programmed to compare what it sees with a pattern. For example, a vision program may try to match a pattern of eyes and a nose in a scene in order to find a face. More complex patterns, e.g. in a natural language text, in a chess position, or in the history of some event are also studied. These more complex patterns require quite different methods than do the simple patterns that have been studied the most.
representation
Facts about the world have to be represented in some way. Usually languages of mathematical logic are used.
inference
From some facts, others can be inferred. Mathematical logical deduction is adequate for some purposes, but new methods of non-monotonic inference have been added to logic since the 1970s. The simplest kind of non-monotonic reasoning is default reasoning in which a conclusion is to be inferred by default, but the conclusion can be withdrawn if there is evidence to the contrary. For example, when we hear of a bird, we man infer that it can fly, but this conclusion can be reversed when we hear that it is a penguin. It is the possibility that a conclusion may have to be withdrawn that constitutes the non-monotonic character of the reasoning. Ordinary logical reasoning is monotonic in that the set of conclusions that can the drawn from a set of premises is a monotonic increasing function of the premises. Circumscription is another form of non-monotonic reasoning.
common sense knowledge and reasoning
This is the area in which AI is farthest from human-level, in spite of the fact that it has been an active research area since the 1950s. While there has been considerable progress, e.g. in developing systems of non-monotonic reasoning and theories of action, yet more new ideas are needed. The Cyc system contains a large but spotty collection of common sense facts.
learning from experience
Programs do that. The approaches to AI based on connectionism and neural nets specialize in that. There is also learning of laws expressed in logic. [Mit97] is a comprehensive undergraduate text on machine learning. Programs can only learn what facts or behaviors their formalisms can represent, and unfortunately learning systems are almost all based on very limited abilities to represent information.
planning
Planning programs start with general facts about the world (especially facts about the effects of actions), facts about the particular situation and a statement of a goal. From these, they generate a strategy for achieving the goal. In the most common cases, the strategy is just a sequence of actions.
epistemology
This is a study of the kinds of knowledge that are required for solving problems in the world.
ontology
Ontology is the study of the kinds of things that exist. In AI, the programs and sentences deal with various kinds of objects, and we study what these kinds are and what their basic properties are. Emphasis on ontology begins in the 1990s.
heuristics
A heuristic is a way of trying to discover something or an idea imbedded in a program. The term is used variously in AI. Heuristic functions are used in some approaches to search to measure how far a node in a search tree seems to be from a goal. Heuristic predicates that compare two nodes in a search tree to see if one is better than the other, i.e. constitutes an advance toward the goal, may be more useful.
Applications of AI
Q. What are the applications of AI?
A. Here are some.
game playing
You can buy machines that can play master level chess for a few hundred dollars. There is some AI in them, but they play well against people mainly through brute force computation--looking at hundreds of thousands of positions. To beat a world champion by brute force and known reliable heuristics requires being able to look at 200 million positions per second.
speech recognition
In the 1990s, computer speech recognition reached a practical level for limited purposes. Thus United Airlines has replaced its keyboard tree for flight information by a system using speech recognition of flight numbers and city names. It is quite convenient. On the the other hand, while it is possible to instruct some computers using speech, most users have gone back to the keyboard and the mouse as still more convenient.
understanding natural language
Just getting a sequence of words into a computer is not enough. Parsing sentences is not enough either. The computer has to be provided with an understanding of the domain the text is about, and this is presently possible only for very limited domains.
computer vision
The world is composed of three-dimensional objects, but the inputs to the human eye and computers' TV cameras are two dimensional. Some useful programs can work solely in two dimensions, but full computer vision requires partial three-dimensional information that is not just a set of two-dimensional views. At present there are only limited ways of representing three-dimensional information directly, and they are not as good as what humans evidently use.
expert systems
A ``knowledge engineer'' interviews experts in a certain domain and tries to embody their knowledge in a computer program for carrying out some task. How well this works depends on whether the intellectual mechanisms required for the task are within the present state of AI. When this turned out not to be so, there were many disappointing results. One of the first expert systems was MYCIN in 1974, which diagnosed bacterial infections of the blood and suggested treatments. It did better than medical students or practicing doctors, provided its limitations were observed. Namely, its ontology included bacteria, symptoms, and treatments and did not include patients, doctors, hospitals, death, recovery, and events occurring in time. Its interactions depended on a single patient being considered. Since the experts consulted by the knowledge engineers knew about patients, doctors, death, recovery, etc., it is clear that the knowledge engineers forced what the experts told them into a predetermined framework. In the present state of AI, this has to be true. The usefulness of current expert systems depends on their users having common sense.
heuristic classification
One of the most feasible kinds of expert system given the present knowledge of AI is to put some information in one of a fixed set of categories using several sources of information. An example is advising whether to accept a proposed credit card purchase. Information is available about the owner of the credit card, his record of payment and also about the item he is buying and about the establishment from which he is buying it (e.g., about whether there have been previous credit card frauds at this establishment).
More questions
Q. How is AI research done?
A. AI research has both theoretical and experimental sides. The experimental side has both basic and applied aspects.
There are two main lines of research. One is biological, based on the idea that since humans are intelligent, AI should study humans and imitate their psychology or physiology. The other is phenomenal, based on studying and formalizing common sense facts about the world and the problems that the world presents to the achievement of goals. The two approaches interact to some extent, and both should eventually succeed. It is a race, but both racers seem to be walking.
Q. What are the relations between AI and philosophy?
A. AI has many relations with philosophy, especially modern analytic philosophy. Both study mind, and both study common sense. The best best reference is [Tho03].
Q. What should I study before or while learning AI?
A. Study mathematics, especially mathematical logic. The more you learn about science in general the better. For the biological approaches to AI, study psychology and the physiology of the nervous system. Learn some programming languages--at least C, Lisp and Prolog. It is also a good idea to learn one basic machine language. Jobs are likely to depend on knowing the languages currently in fashion. In the late 1990s, these include C++ and Java.
Q. What is a good textbook on AI?
A. Artificial Intelligence by Stuart Russell and Peter Norvig, Prentice Hall is the most commonly used textbbook in 1997. The general views expressed there do not exactly correspond to those of this essay. Artificial Intelligence: A New Synthesis by Nils Nilsson, Morgan Kaufman, may be easier to read. Some people prefer Computational Intelligence by David Poole, Alan Mackworth and Randy Goebel, Oxford, 1998.
Q. What organizations and publications are concerned with AI?
A. The American Association for Artificial Intelligence (AAAI), the European Coordinating Committee for Artificial Intelligence (ECCAI) and the Society for Artificial Intelligence and Simulation of Behavior (AISB) are scientific societies concerned with AI research. The Association for Computing Machinery (ACM) has a special interest group on artificial intelligence SIGART.
The International Joint Conference on AI (IJCAI) is the main international conference. The AAAI runs a US National Conference on AI. Electronic Transactions on Artificial Intelligence, Artificial Intelligence, and Journal of Artificial Intelligence Research, and IEEE Transactions on Pattern Analysis and Machine Intelligence are four of the main journals publishing AI research papers. I have not yet found everything that should be in this paragraph.
Page of Positive Reviews lists papers that experts have found important.
Funding a Revolution: Government Support for Computing Research by a committee of the National Research covers support for AI research.
Bibliography
Den98
Daniel Dennett. Brainchildren: Essays on Designing Minds. MIT Press, 1998.
Jen98
Arthur R. Jensen. Does IQ matter? Commentary, pages 20-21, November 1998. The reference is just to Jensen's comment--one of many.
McC59
John McCarthy. Programs with Common Sense. In Mechanisation of Thought Processes, Proceedings of the Symposium of the National Physics Laboratory, pages 77-84, London, U.K., 1959. Her Majesty's Stationery Office. Reprinted in [McC90].
McC89
John McCarthy. Artificial Intelligence, Logic and Formalizing Common Sense. In Richmond Thomason, editor, Philosophical Logic and Artificial Intelligence. Klüver Academic, 1989.
McC90
John McCarthy. Formalizing Common Sense: Papers by John McCarthy. Ablex Publishing Corporation, 1990.
McC96a
John McCarthy. Defending AI research : a collection of essays and reviews. CSLI lecture notes: no. 49. Center for the Study of Language and Information, 1996. distributed by Cambridge University Press.
McC96b
John McCarthy. Concepts of Logical AI, 1996. Web only for now but may be referenced.
Mit97
Tom Mitchell. Machine Learning. McGraw-Hill, 1997.
Sha97
Murray Shanahan. Solving the Frame Problem, a mathematical investigation of the common sense law of inertia. M.I.T. Press, 1997.
Tho03
Richmond Thomason. Logic and artificial intelligence. In Edward N. Zalta, editor, The Stanford Encyclopedia of Philosophy. Fall 2003. http://plato.stanford.edu/archives/fall2003/entries/logic-ai/.
Tur50
Alan Turing. Computing machinery and intelligence. Mind, 1950.
An Overview.AI is a result of the merge of philosophy, mathematics, psychology, neurology, linguistics, computer science, and many other fields. Futhermore, the application of AI relates to almost any fields. This variety gives AI an endless potential. A relatively young science, AI has made much progress in 50 years. Though fast-growing, AI has never actually caught up with all the expectation imposed on it. There are two reasons for public's over-confidence in AI. First, AI theories are often ingenious and subtle even fictional, implying much futuristic applications. Second, AI, being incorporated with computer technology, is often expected to progress as fast as the computer technology. Conclusionally, AI is a young, energetic, and attractive science.
EXAMPLES
The branch of computer science concerned with making computers behave like humans. The term was coined in 1956 by John McCarthy at the Massachusetts Institute of Technology. Artificial intelligence includes
games playing: programming computers to play games such as chess and checkers
expert systems: programming computers to make decisions in real-life situations (for example, some expert systems help doctors diagnose diseases based on symptoms)
natural language: programming computers to understand natural human languages
neural networks: Systems that simulate intelligence by attempting to reproduce the types of physical connections that occur in animal brains
robotics: programming computers to see and hear and react to other sensory stimuli Currently, no computers exhibit full artificial intelligence (that is, are able to simulate human behavior). The greatest advances have occurred in the field of games playing. The best computer chess programs are now capable of beating humans. In May, 1997, an IBM super-computer called Deep Blue defeated world chess champion Gary Kasparov in a chess match.
In the area of robotics, computers are now widely used in assembly plants, but they are capable only of very limited tasks. Robots have great difficulty identifying objects based on appearance or feel, and they still move and handle objects clumsily.
Natural-language processing offers the greatest potential rewards because it would allow people to interact with computers without needing any specialized knowledge. You could simply walk up to a computer and talk to it. Unfortunately, programming computers to understand natural languages has proved to be more difficult than originally thought. Some rudimentary translation systems that translate from one human language to another are in existence, but they are not nearly as good as human translators. There are also voice recognition systems that can convert spoken sounds into written words, but they do not understand what they are writing; they simply take dictation. Even these systems are quite limited -- you must speak slowly and distinctly.
In the early 1980s, expert systems were believed to represent the future of artificial intelligence and of computers in general. To date, however, they have not lived up to expectations. Many expert systems help human experts in such fields as medicine and engineering, but they are very expensive to produce and are helpful only in special situations.
Today, the hottest area of artificial intelligence is neural networks, which are proving successful in a number of disciplines such as voice recognition and natural-language processing.
posted by macky the wonderer at 4:13 AM
1 Comments:
I came across your website today and noticed we have similar content. Your site is a LOT more developed and professional than mine but I am new and willing to learn! I would like to advertise your site for you and put a banner ad of some kind up for you, a bit of mutual beneficial cooperation you know. Please check out my arai motorcycle helmet site and let me know if you approve. Thank you ahead of time for your time. :)
Post a Comment
<< Home