Artificial_intelligence

{{Redirect|AI|other uses of "AI" and "Artificial intelligence"}}
[[Image:p11 kasparov breakout.jpg|thumb|right|280px|[[Garry Kasparov]] playing against [[IBM Deep Blue|Deep Blue]], the first machine to win a chess match against a reigning world champion.]]

The modern definition of '''artificial intelligence''' (or '''AI''') is "the study and design of [[intelligent agents]]" where an intelligent agent is a system that perceives its environment and takes actions which maximizes its chances of success.<ref>
Textbooks that define AI this way include
{{Harvnb|Poole|Mackworth|Goebel|1998|loc=[http://www.cs.ubc.ca/spider/poole/ci/ch1.pdf p. 1]}},
{{Harvnb|Nilsson|1998}},
and {{Harvnb|Russell|Norvig|2003|loc=[http://aima.cs.berkeley.edu/preface.html preface]}} (who prefer the term "rational agent") and write "The whole-agent view is now widely accepted in the field" {{Harv|Russell|Norvig|2003|p=55}}
</ref>
[[John McCarthy (computer scientist)|John McCarthy]], who coined the term in 1956,<ref>
Although there is some controversy on this point (see {{Harvnb|Crevier|1993|p=50}}), [[John McCarthy|McCarthy]] states unequivocally "I came up with the term" in a c|net interview. (See [http://news.com.com/Getting+machines+to+think+like+us/2008-11394_3-6090207.html Getting Machines to Think Like Us].)
</ref>
defines it as "the science and engineering of making intelligent machines."<ref>
See [[John McCarthy (computer scientist)| John McCarthy]], [http://www-formal.stanford.edu/jmc/whatisai/whatisai.html What is Artificial Intelligence?]
</ref>
Other names for the field have been proposed, such as [[computational intelligence]],<ref name=PMGp1>
{{Harvnb|Poole|Mackworth|Goebel|1998|loc=[http://www.cs.ubc.ca/spider/poole/ci/ch1.pdf p. 1]}}
</ref>
[[synthetic intelligence]]<ref name=PMGp1/>
or computational rationality.<ref>
{{Harvnb|Russell|Norvig|2003|p=17}}
</ref>

The term '''artificial intelligence''' is also used to describe a ''property'' of machines or programs: the [[intelligence (trait)|intelligence]] that the system demonstrates. Among the traits that researchers hope machines will exhibit are [[:#Deduction, reasoning, problem solving|reasoning]], [[#Knowledge representation|knowledge]], [[#Planning|planning]], [[#Learning|learning]], [[#Natural language processing|communication]], [[#Perception|perception]] and the ability to [[#Motion and manipulation|move]] and manipulate objects.<ref name=I>
This list of intelligent traits is based on the topics covered by the major AI textbooks, including:
{{Harvnb|Russell|Norvig|2003}},
{{Harvnb|Luger|Stubblefield|2004}},
{{Harvnb|Poole|Mackworth|Goebel|1998}} and
{{Harvnb|Nilsson|1998}}.
</ref>
[[#General intelligence|General intelligence]] (or "[[strong AI]]") has not yet been achieved and is a long-term goal of AI research.<ref name=GI>
General intelligence ([[strong AI]]) is discussed by popular introductions to AI, such as:
{{Harvnb|Kurzweil|1999}},
{{Harvnb|Kurzweil|2005}},
{{Harvnb|Hawkins|Blakeslee|2004}}
</ref>

AI research uses tools and insights from many fields, including [[computer science]], [[psychology]], [[philosophy]], [[neuroscience]], [[cognitive science]], [[computational linguistics|linguistics]], [[operations research]], [[computational economics|economics]], [[control theory]], [[probability]], [[optimization (mathematics)|optimization]] and [[logic]].<ref>{{Harvnb|Russell|Norvig|2003|pp=5-16}}</ref>
AI research also overlaps with tasks such as [[robotics]], [[control system]]s, [[automated planning and scheduling|scheduling]], [[data mining]], [[logistics]], [[speech recognition]], [[facial recognition system|facial recognition]] and many others.<ref>See [http://www.aaai.org/AITopics/html/applications.html AI Topics: applications]</ref>

{{portal}}
==Perspectives on AI==
[[Image:PygmalianGalatea.jpg|thumb|right|213px|''Pygmalion and Galatea'' ([[1890]]) by [[Jean-Léon Gérôme]] ([[1824]]–[[1904]])]]
===History of AI===
{{Main|history of artificial intelligence|timeline of artificial intelligence}}

[[Samuel Butler]] first raised the possibility of "mechanical consciousness" in an article signed with the [[Pen name|''nom de plume]] [[Cellarius]] and headed "Darwin among the Machines", which appeared in the [[Christchurch]], [[New Zealand]], newspaper ''[[The Press]]'' on [[13 June]] [[1863]].
<ref>PREFACE TO THE REVISED EDITION, [[Project Gutenberg]] eBook [http://www.gutenberg.org/etext/1906 '''Erewhon'''], by Samuel Butler. Release Date: March 20, 2005.</ref>
Butler envisioned [[mechanical]] [[consciousness]] emerging by means of ''[[Charles Darwin|Darwinian]]'' ''[[Evolution]]'', specifically by ''[[Natural selection]]'', as a form of ''natural'', not ''artificial'', intelligence.

{{Harvtxt|McCorduck|2004}}, however, writes "Artificial intelligence in one form or another is an idea that has pervaded Western intellectual history, a dream in urgent need of being realized." She continues:
<blockquote>"Work toward that end has been a splendid effort, the variety of its form as wondrous as anything humans have conceived; its practitioners as lively a group of poets, dreamers, holy men, rascals, and assorted eccentrics as one could hope to find—not a dullard among them. Its visionaries have lifted our spirits and made us transcend our own species, its poets have told us things about ourselves we never suspected, and its fast talkers have set everybody's teeth on edge."<ref>
{{Harvnb|McCorduck|2004|p=xviii}}
</ref>
</blockquote>
Beginning with the myth of [[Pygmalian]] and [[Galatea]], we have imagined making copies of ourselves, with [[cult image|sacred statues]], [[alchemical]] beings and charming clockwork [[automaton]]s.<ref>
{{Harvnb|McCorduck|2004|p=3−35}}
</ref> Yet we also have a fear that our creations may turn on us, as in [[Golem|The Golem of Prague]] and [[Frankenstein]].

In the middle of the 20th century, a handful of scientists explored a new approach to an ancient dream, based on their discoveries in [[neurology]], a new mathematical theory of [[information]], an understanding of control and stability called [[cybernetic]]s, and above all, by the invention of the [[digital computer]], a machine based on the abstract essence of mathematical reasoning.<ref name=CYBER>
Among the researchers who laid the foundations of the [[theory of computation]], [[cybernetic]]s, [[information theory]] and [[neural networks]] were [[Claude Shannon]], [[Norbert Weiner]], [[Warren McCullough]], [[Walter Pitts]], [[Donald Hebb]], [[Donald McKay]], [[Alan Turing]] and [[John Von Neumann]].
{{Harvnb|McCorduck|2004|pp=51-107}}
{{Harvnb|Crevier|1993|pp=27-32}},
{{Harvnb|Russell|Norvig|2003|pp=15,940}},
{{Harvnb|Moravec|1988|p=3}}.
</ref>

In the summer of 1956, at a conference on the campus of [[Dartmouth College]], the field of AI research was born.<ref>
{{Harvnb|Crevier|1993|pp=47-49}},
{{Harvnb|Russell|Norvig|2003|p=17}}
</ref>
Those who attended would become the leaders of AI research for many decades, especially [[John McCarthy (computer scientist)|John McCarthy]], [[Marvin Minsky]], [[Allen Newell]] and [[Herbert Simon]], who founded AI laboratories at [[MIT]], [[CMU]] and [[Stanford]]. They and their students wrote programs that were, to most people, simply astonishing:<ref>
Russell and Norvig write "it was astonishing whenever a computer did anything kind of smartish." {{Harvnb|Russell|Norvig|2003|p=18}}
</ref>
computers were solving word problems in algebra, proving logical theorems and speaking English.<ref>
{{Harvnb|Crevier|1993|pp=52-107}}, {{Harvnb|Moravec|1988|p=9}} and {{Harvnb|Russell|Norvig|2003|p=18-21}}. The programs described are [[Daniel Bobrow]]'s [[STUDENT (computer program)|STUDENT]], [[Allen Newell|Newell]] and [[Herbert Simon|Simon]]'s [[Logic Theorist]] and [[Terry Winograd]]'s [[SHRDLU]].
</ref>
By the middle 60s their research was heavily funded by [[DARPA]]<ref>
{{Harvnb|Crevier|1993|pp=64-65}}
</ref>
and they were optimistic about the future of the new field:
*1965, [[H. A. Simon]]: "[M]achines will be capable, within twenty years, of doing any work a man can do"<ref>
{{Harvnb|Simon|1965|p=96}} quoted in {{Harvnb|Crevier|1993|p=109}}
</ref>
*1967, [[Marvin Minsky]]: "Within a generation ... the problem of creating 'artificial intelligence' will substantially be solved."<ref>{{Harvnb|Minsky|1967|p=2}} quoted in {{Harvnb|Crevier|1993|p=109}}</ref>
These predictions, and many like them, would not come true. They had failed to recognize the difficulty of some of the problems they faced.<ref>
See [[History of artificial intelligence#The problems|History of artificial intelligence — the problems]].
</ref>
In 1974, in response to the criticism of England's [[Sir James Lighthill]] and ongoing pressure from Congress to fund more productive projects, [[DARPA]] cut off all undirected, exploratory research in AI. This was the first [[AI Winter]].<ref>
{{Harvnb|Crevier|1993|pp=115-117}},
{{Harvnb|Russell|Norvig|2003|p=22}},
{{Harvnb|NRC|1999}} under "Shift to Applied Research Increases Investment." and also see Howe, J. [http://www.dai.ed.ac.uk/AI_at_Edinburgh_perspective.html ''"Artificial Intelligence at Edinburgh University : a Perspective"'']
</ref>

In the early 80s, the field was revived by the commercial success of [[expert systems]] and by 1985 the market for AI had reached more than a billion dollars.<ref>
{{harvnb|Crevier|1993|pp=161-162,197-203}} and and {{Harvnb|Russell|Norvig|2003|p=24}}
</ref>
[[Marvin Minsky|Minsky]] and others warned the community that enthusiasm for AI had spiraled out of control and that disappointment was sure to follow.<ref>
{{Harvnb|Crevier|1993|p=203}}
</ref>
[[Marvin Minsky|Minsky]] was right. Beginning with the collapse of the [[Lisp Machine]] market in 1987, AI once again fell into disrepute, and a second, more lasting [[AI Winter]] began.<ref>
{{Harvnb|Crevier|1993|pp=209-210}}
</ref>

In the 90s and early 21st century AI achieved its greatest successes, albeit somewhat behind the scenes. Artificial intelligence was adopted throughout the technology industry, providing the heavy lifting for [[logistics]], [[data mining]], [[medical diagnosis]] and many other areas.<ref>
{{Harvnb|Russell|Norvig|p=28}},
{{Harvnb|NRC|1999}} under "Artificial Intelligence in the 90s"
</ref>
The success was due to several factors: the incredible power of computers today (see [[Moore's law]]), a greater emphasis on solving specific subproblems, the creation of new ties between AI and other fields working on similar problems, and above all a new commitment by researchers to solid mathematical methods and rigorous scientific standards.<ref>{{Harvnb|Russell|Norvig|pp=25-26}}</ref>

===Philosophy of AI===
{{portalpar|Mind and Brain}}
{{main|philosophy of artificial intelligence}}

[[Image:User-FastFission-brain.gif|thumb|213px|left|Can the brain be simulated? Does this prove machines can think?]]
The [[philosophy of artificial intelligence]] considers the question "Can machines think?" [[Alan Turing]], in his classic 1950 paper, [[Computing Machinery and Intelligence]], was the first to try to answer it. In the years since, several answers have been given:<ref>
All of these positions are mentioned in standard discussions of the subject, such as {{Harvnb|Russell|Norvig|2003|pp=947-960}} and {{Harvnb|Fearn|2007|pp=38-55}}
</ref>
*[[Turing Test|Turing's "polite convention"]]: ''If a machine acts as intelligently as a human being, then it is as intelligent as a human being.''<ref>
This is a paraphrase of the essential point of the [[Turing test]]. {{Harvnb|Turing|1950}}, {{Harvnb|Haugeland|1985|pp=6-9}}, {{Harvnb|Crevier|1993|p=24}}, {{Harvnb|Russell|Norvig|2003|pp=2-3 and 948}}
</ref>
* The [[Dartmouth Conferences|Dartmouth proposal]]: ''Every aspect of learning or any other feature of intelligence can be so precisely described that a machine can be made to simulate it.'' This assertion was printed in the program for the [[Dartmouth Conferences|Dartmouth Conference]] of 1956, and represents the position of most working AI researchers.<ref>
{{Harvnb|McCarthy|Minsky|Rochester|Shannon|1955}} See also {{Harvnb|Crevier|1993|p=28}}
</ref>
* [[Alan Newell|Newell]] and [[Herbert Simon|Simon]]'s [[physical symbol system|physical symbol system hypothesis]]: ''A physical symbol system has the necessary and sufficient means of general intelligent action.'' This statement claims that essence of intelligence is symbol manipulation.<ref>
{{Harvnb|Newell|Simon|1963}} and {{Harvnb|Russell|Norvig|2003|p=18}}
</ref>
* The [[artificial brain]] argument: ''The brain can be simulated.'' This argument combines the idea that a [[Turing complete]] machine can simulate any process, with the [[materialist]] idea that the [[mind]] is the result of a physical process in the [[brain]].<ref>
{{Harvnb|Kurzweil|2005|p=262}}. Also see {{Harvnb|Russell|Norvig|p=957}} and {{Harvnb|Crevier|1993|pp=271 and 279}}. The most extreme form of this argument (the brain replacement scenario) was put forward by [[Clark Glymour]] in the mid-70s and was touched on by [[Zenon Pylyshyn]] and [[John Searle]] in 1980. It is now associated with [[Hans Moravec]] and [[Ray Kurzweil]].
</ref>
* [[Gödel's incompleteness theorem]]: ''There are statements that no physical symbol system can prove.'' [[Roger Penrose]] is among those who claim that Gödel's theorem limits what machines can do.<ref>
This is a paraphrase of the most important implication of Gödel's theorems, according {{Harvtxt|Hofstadter|1979}}. See also {{Harvnb|Russell|Norvig|2003|p=949}}, {{Harvnb| Gödel|1931}}, {{Harvnb|Church|1936}}, {{Harvnb|Kleene|1935}}, {{Harvnb|Turing|1937}}, {{Harvnb|Turing|1950}} under “(2) The Mathematical Objection”
</ref>
* [[Hubert Dreyfus|Dreyfus]]' "[[What Computers Can't Do|psychological assumption]]": ''The mind can be viewed as a device operating on bits of information according to formal rules.'' Dreyfus refuted this statement by showing that human expertise depends on unconscious instinct rather than conscious symbol manipulation and on having a "feel" for the situation rather explicit symbolic knowledge.<ref>
{{Harvnb|Dreyfus|1992|p=156}}. See also {{Harvnb|Dreyfus|Dreyfus|1986}}, {{Harvnb|Russell|Norvig|2003|pp=950-952}}, {{Harvnb|Crevier|1993|120-132}} and {{Harvnb|Hearn|2007|pp=50-51}}
</ref>
* [[John Searle|Searle]]'s "strong AI position": ''A physical symbol system can have a [[mind]] and [[consciousness|mental states]].'' Searle refuted this with his [[Chinese room]] argument, which asks us to look ''inside'' the computer and try to find where the "mind" might be.<ref name=SWAI>{{Harvnb|Searle|1980}}. See also {{Harvtxt|Russell|Norvig|2003|p=947}}: "The assertion that machines could possibly act intelligently (or, perhaps better, act as if they were intelligent) is called the 'weak AI' hypothesis by philosophers, and the assertion that machines that do so are actually thinking (as opposed to simulating thinking) is called the 'strong AI' hypothesis," although Searle's arguments, such as the [[Chinese Room]], apply only to [[physical symbol system]]s, not to machines in general (he would consider the brain a machine). Also, notice that the positions as Searle states them don't make any commitment to how ''much'' intelligence the system has: it is one thing to say a machine can act intelligently, it is another to say it can act as intelligently as a human being.</ref>

===AI in fiction===
{{main|Artificial intelligence in fiction}}

[[Image:Hal-9000.jpg|thumb|213px|[[HAL 9000]]'s iconic camera eye.]]

In modern [[science fiction]], AI are not necessarily limited by the fundamental problems of [[perception]], [[knowledge representation]], [[common sense reasoning]], or [[learning]]. This allows speculation on the technology's potential impact on humanity, meditations on [[metaphysics]] or the nature of [[awareness]], and the use of novel plot devices. AI has appeared in fiction as a servant ([[R2D2]]), a comrade ([[Data (Star Trek)|Lt. Commander Data]]), a technology expanding human ability ([[Ghost in the Shell]]), a conqueror ([[With Folded Hands]]), an exterminator ([[Terminator (series)|Terminator]], [[Battlestar Galactica (re-imagining)|Battlestar Galactica]]), a manager ([[Portal (video game)]]).
Some realistic potential consequences of AI investigated in fiction are decreased labor demand, the enhancement of human ability or experience, and a need for redefinition of human identity and basic values (or a threat to existing identity and values). (See the [[Uncanny Valley]] hypothesis.) One area of speculation focuses on potential disaster. (See [[artificial intelligence in fiction#AI and Society| AI and Society in fiction]])

Though in fiction AI are often aware and capable of feeling, the phenomena that allow these experiences are not understood, and as such, there is no theoretical basis for their synthesis. Current theories provide for machines that can replicate or surpass all external human behavior, but not necessarily human experience.

An AI can play any role traditionally assigned to humans in a [[narrative]], such as that of [[protagonist]] ([[Bicentennial Man (film)|Bicentennial Man]]), [[antagonist]] ([[Terminator (character concept)|Terminator]], [[HAL 9000]]), faithful companion ([[R2D2]]), or comic relief ([[C3PO]]). (See [[artificial intelligence in fiction#Sentient AI|Sentient AI in fiction]].)

Many portrayals of AI in [[science fiction]] deal either with person-like or [[sentient]] AI, but the technology of AI appears in many other forms. (See [[artificial intelligence in fiction#Non-Sentient AI| non-sentient AI in fiction]].)

The inevitability of the integration of AI into human society is also argued by some science/futurist writers such as [[Kevin Warwick]] and [[Hans Moravec]] and the manga [[Ghost in the Shell]]

==AI research==

===Problems of AI===
While there is no universally accepted definition of intelligence,<ref>
"We cannot yet characterize in general what kinds of computational procedures we want to call intelligent." [[John McCarthy (computer scientist)|John McCarthy]], [http://www-formal.stanford.edu/jmc/whatisai/node1.html Basic Questions]
</ref>
AI researchers have studied several traits that are considered essential.<ref name=I/>

====Deduction, reasoning, problem solving ====
Early AI researchers developed algorithms that imitated the process of conscious, step-by-step reasoning that human beings use when they solve puzzles, play board games, or make logical deductions.<ref>
Problem solving, puzzle solving, game playing and deduction:
{{Harvnb|Russell|Norvig|2003|loc=chpt. 3-9}},
{{Harvnb|Poole|Mackworth|Goebel|1998|chpt. 2,3,7,9}},
{{Harvnb|Luger|Stubblefield|2004|loc=chpt. 3,4,6,8}},
{{Harvnb|Nilsson|loc=chpt. 7-12}}.
</ref>
By the late 80s and 90s, AI research had also developed highly successful methods for dealing with [[uncertainty|uncertain]] or incomplete information, employing concepts from [[probability]] and [[economics]].<ref>
Uncertain reasoning:
{{Harvnb|Russell|Norvig|2003|pp=452-644}},
{{Harvnb|Poole|Mackworth|Goebel|1998|pp=345-395}},
{{Harvnb|Luger|Stubblefield|2004|pp=333-381}},
{{Harvnb|Nilsson|1998|loc=chpt. 19}}
</ref>

For difficult problems, most of these algorithms can require enormous computational resources — most experience a "[[combinatorial explosion]]": the amount of memory or computer time required becomes astronomical when the problem goes beyond a certain size. The search for more efficient problem solving algorithms is a high priority for AI research.<ref>
[[Intractable|Intractability and efficiency]] and the [[combinatorial explosion]]:
{{Harvnb|Russell|Norvig|2003|pp=9, 21-22}}
</ref>

It is not clear, however, that conscious human reasoning is any more efficient when faced with a difficult abstract problem. [[Cognitive science|Cognitive scientists]] have demonstrated that human beings solve most of their problems using [[unconscious]] reasoning, rather than the conscious, step-by-step deduction that early AI research was able to model.<ref>
Several famous examples: {{Harvtxt|Wason|1966}} showed that people do poorly on completely abstract problems, but if the problem is restated to allowed the use of intuitive [[social intelligence]], performance dramatically improves. (See [[Wason selection task]]) {{Harvtxt|Tversky|Slovic|Kahnemann|1982}} have shown that people are terrible at elementary problems that involve uncertain reasoning. (See [[list of cognitive biases]] for several examples). {{Harvtxt|Lakoff|Nunez|2000}} have controversially argued that even our skills at mathematics depend on knowledge and skills that come from "the body", i.e. sensorimotor and perceptual skills. (See [[Where Mathematics Comes From]])
</ref>
[[Embodied cognitive science]] argues that unconscious [[sensorimotor]] skills are essential to our problem solving abilities. It is hoped that sub-symbolic methods, like [[computational intelligence]] and [[situated]] AI, will be able to model these instinctive skills. The problem of unconscious problem solving, which forms part of our [[commonsense reasoning]], is largely unsolved.

====Knowledge representation====
{{Main|knowledge representation|commonsense knowledge}}

[[Knowledge representation]]<ref>
[[Knowledge representation]]:
{{Harvnb|ACM|1998|loc=I.2.4}},
{{Harvnb|Russell|Norvig|2003|pp=320-363}},
{{Harvnb|Poole|Mackworth|Goebel|1998|pp=23-46, 69-81, 169-196, 235-277, 281-298, 319-345}}
{{Harvnb|Luger|Stubblefield|2004|pp=227-243}},
{{Harvnb|Nilsson|1998|loc=chpt. 18}}
</ref>
and [[knowledge engineering]]<ref>
[[Knowledge engineering]]:
{{Harvnb|Russell|Norvig|2003|pp=260-266}},
{{Harvnb|Poole|Mackworth|Goebel|1998|pp=199-233}},
{{Harvnb|Nilsson|1998|loc=chpt. ~17.1-17.4}}
</ref>
are central to AI research. Many of the problems machines are expected to solve will require extensive knowledge about the world. Among the things that AI needs to represent are: objects, properties, categories and relations between objects;<ref name=DL>
Representing categories and relations: [[Semantic network]]s, [[description logic]]s, [[inheritance (computer science)|inheritance]], including [[Frame (artificial intelligence)|frame]]s and [[scripts (artificial intelligence)|scripts]]):
{{Harvnb|Russell|Norvig|2003|pp=349-354}},
{{Harvnb|Poole|Mackworth|Goebel|1998|pp=174-177}},
{{Harvnb|Luger|Stubblefield|2004|pp=248-258}},
{{Harvnb|Nilsson|1998|loc=chpt. 18.3}}
</ref>
situations, events, states and time;<ref name=SC>
Representing events and time: [[Situation calculus]], [[event calculus]], [[fluent calculus]] (including solving the [[frame problem]]):
{{Harvnb|Russell|Norvig|2003|pp=328-341}},
{{Harvnb|Poole|Mackworth|Goebel|1998|pp=281-298}},
{{Harvnb|Nilsson|1998|loc=chpt. 18.2}}
</ref>
causes and effects;<ref name=CC>
[[Causality#causal calculus|Causal calculus]]:
{{Harvnb|Poole|Mackworth|Goebel|1998|pp=335-337}}
</ref>
knowledge about knowledge (what we know about what other people know);<ref name=BC>
Representing knowledge about knowledge: [[Belief calculus]], [[modal logic]]s:
{{Harvnb|Russell|Norvig|2003|pp=341-344}},
{{Harvnb|Poole|Mackworth|Goebel|1998|pp=275-277}}
</ref>
and many other, less well researched domains. A complete representation of "what exists" is an [[ontology (computer science)|ontology]]<ref>
[[Ontology (computer science)|Ontology]]:
{{Harvnb|Russell|Norvig|2003|pp=320-328}}
</ref>
(borrowing a word from traditional [[philosophy]]). Ontological engineering is the science of finding a general representation that can handle all of human knowledge.

Among the most difficult problems in knowledge representation are:
* ''Default reasoning and the [[qualification problem]]'': Many of the things people know take the form of "working assumptions." For example, if a bird comes up in conversation, people typically picture a animal that is fist sized, sings, and flies. None of these things are true about birds in general. [[John McCarthy (computer scientist)|John McCarthy]] identified this problem in 1969<ref>{{Harvnb|McCarthy|Hayes|1969}}</ref> as the qualification problem: for any commonsense rule that AI researchers care to represent, there tend to be a huge number of exceptions. Almost nothing is simply true or false in the way that abstract logic requires. AI research has explored a number of solutions to this problem.<ref name=NML>
Default reasoning and [[default logic]], [[non-monotonic logic]]s, [[circumscription]], [[closed world assumption]], [[abduction]] (Poole ''et al.'' places abduction under "default reasoning". Luger ''et al.'' places this under "uncertain reasoning"):
{{Harvnb|Russell|Norvig|2003|pp=354-360}},
{{Harvnb|Poole|Mackworth|Goebel|1998|pp=248-256, 323-335}}
{{Harvnb|Luger|Stubblefield|2004|pp=335-363}},
{{Harvnb|Nilsson|1998|loc=~18.3.3}}
</ref>
* ''Unconscious knowledge'': Much of what people know isn't represented as "facts" or "statements" that they could actually say out loud. They take the form of intuitions or tendencies and are represented in the brain unconsciously and sub-symbolically. This unconscious knowledge informs, supports and provides a context for our conscious knowledge. As with the related problem of unconscious reasoning, it is hoped that [[situated]] AI or [[computational intelligence]] will provide ways to represent this kind of knowledge.
* ''The breadth of [[common sense knowledge]]'': The number of atomic facts that the average person knows is astronomical. Research projects that attempt to build a complete knowledge base of [[commonsense knowledge]], such as [[Cyc]], require enormous amounts of tedious step-by-step ontological engineering — they must be built, by hand, one complicated concept at a time.<ref>
{{Harvnb|Crevier|1993|pp=113-114}},
{{Harvnb|Moravec|1988|p=13}},
{{Harvnb|Lenat|1989}} (Introduction),
{{Harvnb|Russell|Norvig|2003|p=21}}
</ref>

====Planning====
{{Main|automated planning and scheduling}}

Intelligent agents must be able set goals and achieve them.<ref>
[[automated planning and scheduling|Planning]]:
{{Harvnb|ACM|1998|loc=~I.2.8}},
{{Harvnb|Russell|Norvig|2003|pp= 375-459}},
{{Harvnb|Poole|Mackworth|Goebel|1998|pp=281-316}},
{{Harvnb|Luger|Stubblefield|2004|pp=314-329}},
{{Harvnb|Nilsson|1998|loc=chpt. 10.1-2, 22}}
</ref>
They need a way to visualize the future: they must have a representation of the state of the world and be able to make predictions about how their actions will change it. There are several types of planning problems:
* Classical planning problems assume that the agent is the only thing acting on the world, and that the agent can be certain what the consequences of it's actions may be.<ref>
[[Classical planning]]:
{{Harvnb|Russell|Norvig|2003|pp=375-430}}
{{Harvnb|Poole|Mackworth|Goebel|1998|pp=281-309}},
{{Harvnb|Luger|Stubblefield|2004|pp=314-329}},
{{Harvnb|Nilsson|1998|loc=chpt. 10.1-2, 22}}
</ref> [[Partial order]] planning problems take into account the fact that sometimes it's not important which sub-goal the agent achieves first.<ref>
[[Partial order]] planning:
{{Harvnb|Russell|Norvig|2003|pp=387-395}},
{{Harvnb|Poole|Mackworth|Goebel|1998|pp=309-315}},
{{Harvnb|Nilsson|1998|loc=chpt. 22.2}}
</ref>
* If the environment is changing, or if the agent can't be sure of the results of its actions, it must periodically check if the world matches its predictions (conditional planning and execution monitoring) and it must change its plan as this becomes necessary (replanning and continuous planning).<ref>
Planning and acting in non-deterministic domains: conditional planning, execution monitoring, replanning and continuous planning:
{{Harvnb|Russell|Norvig|2003|pp=430-449}}
</ref>
* Some [[automated planning and scheduling|planning]] problems take into account the [[utility]] or "usefulness" of a given outcome. These problems can be analyzed using tools drawn from [[economic]]s, such as [[decision theory]] or [[decision analysis]]<ref name=DTA>
[[Decision theory]] and [[decision analysis]]:
{{Harvnb|Russell|Norvig|2003|pp=584-597}},
{{Harvnb|Poole|Mackworth|Goebel|1998|pp=381-394}}
</ref> and [[Applied information economics|information value theory]].<ref name=IVT>
[[Applied information economics|Information value theory]]:
{{Harvnb|Russell|Norvig|2003|pp=600-604}}
</ref>
* [[Multi-agent planning]] problems try to determine the best plan for a community of [[agent]]s, using [[cooperation]] and [[competition]] to achieve a given goal.<ref>
Multi-agent planning and emergent behavior
{{Harvnb|Russell|Norvig|2003|pp=449-455}}
</ref> These problems are related to emerging fields like [[evolutionary algorithm]]s and [[swarm intelligence]].

====Learning====
{{Main|machine learning}}
Important [[machine learning]]<ref>
[[machine learning|Learning]]:
{{Harvnb|ACM|1998|loc=I.2.6}},
{{Harvnb|Russell|Norvig|2003|pp=649-788}},
{{Harvnb|Poole|Mackworth|Goebel|1998|pp=397-438}},
{{Harvnb|Luger|Stubblefield|2004|pp=385-542}}
{{Harvnb|Nilsson|1998|loc=chpt. 3.3 , 10.3, 17.5, 20}}
</ref>
problems are:
* [[Unsupervised learning]]: find a model that matches a stream of input "experiences", and be able to predict what new "experiences" to expect.
* [[Supervised learning]], such as [[statistical classification|classification]] (be able to determine what category something belongs in, after seeing a number of examples of things from each category), or [[regression]] (given a set of numerical input/output examples, discover a continuous function that would generate the outputs from the inputs).
* [[Reinforcement learning]]:<ref>
[[Reinforcement learning]]:
{{Harvnb|Russell|Norvig|2003|pp=763-788}},
{{Harvnb|Luger|Stubblefield|2004|pp=442-449}}
</ref> the agent is rewarded for good responses and punished for bad ones. (These can be analyzed in terms [[decision theory]], using concepts like [[utility (economics)|utility]]).

====Natural language processing====
{{Main|natural language processing}}

[[Natural language processing]]<ref>
[[Natural language processing]]:
{{Harvnb|ACM|1998|loc=I.2.7}},
{{Harvnb|Russell|Norvig|2003|pp=790-831}},
{{Harvnb|Poole|Mackworth|Goebel|1998|pp=91-104}},
{{Harvnb|Luger|Stubblefield|2004|pp=591-632}}
</ref>
gives machines the ability to be read and understand the languages human beings speak. The problem of [[natural language processing]] involves such subproblems as: [[syntax]] and [[parsing]];<ref>
[[Syntax]] and [[parsing]]:
{{Harvnb|Russell|Norvig|2003|pp=795-810}},
{{Harvnb|Luger|Stubblefield|2004|pp=597-616}}
</ref>
[[semantics]] and [[word sense disambiguation|disambiguation]];<ref>
[[Semantics]] and [[word sense disambiguation|disambiguation]]:
{{Harvnb|Russell|Norvig|2003|pp=810-821}}
</ref> and discourse understanding.<ref>
Discourse understanding ([[coherence (linguistics)|coherence]] relations, [[speech act]]s, [[pragmatics]]):
{{Harvnb|Russell|Norvig|2003|pp=820-824}}
</ref>
Many researchers hope that a sufficiently powerful natural language processing system would be able to acquire knowledge on its own, by reading the existing text available over the internet.

Some straightforward applications of natural language processing include [[information retrieval]] (or [[text mining]]) and [[machine translation]].<ref>
Applications of natural language processing, including [[information retrieval]] (or [[text mining]]) and [[machine translation]]
{{Harvnb|Russell|Norvig|2003|pp=840-857}},
{{Harvnb|Luger|Stubblefield|2004|pp=623-630}}
</ref>

====Perception====
{{Main|machine perception|computer vision|speech recognition}}

[[Machine perception]]<ref>
[[Machine perception]]:
{{Harvnb|Russell|Norvig|2003|pp=537-581, 863-898}},
{{Harvnb|Nilsson|1998|loc=~chpt. 6}}
</ref>
is the ability to use input from sensors (such as cameras, microphones, sonar and others more exotic) to deduce aspects of the world. [[Computer vision]]<ref>
[[Computer vision]]:
{{Harvnb|ACM|1998|loc=I.2.10}},
{{Harvnb|Russell|Norvig|2003|pp=863-898}},
{{Harvnb|Nilsson|1998|loc=chpt. 6}}
</ref>
is the ability to analyze visual input. A few selected subproblems are [[speech recognition]],<ref>
[[Speech recognition]]:
{{Harvnb|ACM|1998|loc=~I.2.7}},
{{Harvnb|Russell|Norvig|2003|pp=568-578}}
</ref>
[[facial recognition]] and [[object recognition]].<ref>
[[Object recognition]]:
{{Harvnb|Russell|Norvig|2003|pp=885-892}}
</ref>

====Motion and manipulation====
{{Main|robotics}}

The field of [[robotics]]<ref>
[[Robotic]]s:
{{Harvnb|ACM|1998|loc=I.2.9}},
{{Harvnb|Russell|Norvig|2003|pp=901-942}},
{{Harvnb|Poole|Mackworth|Goebel|1998|pp=443-460}}
</ref>
is closely related to AI. Intelligence is required for robots to be able to handle such tasks as:
* [[navigation|navigate]], referred to as [[robotic mapping]] including the sub-problems of [[localization]] (knowing where you are), [[robotic mapping|mapping]] (learning what is around you) and [[robotic mapping#Path planning|path planning]] (figuring out how to get there).<ref>
[[Robotic mapping]] (localization, etc)
{{Harvnb|Russell|Norvig|pp=908-915}}
</ref>
* manipulate objects (usually described in terms of [[configuration space]]).<ref name=CS>
Moving and [[configuration space]]:
{{Harvnb|Russell|Norivg|pp=916-932}}
</ref>

====Social intelligence====
{{Main|affective computing}}
[[Image:Wikimania 2006 POLIMEREK 100-0093 IMG.JPG|thumb|213px|[[Kismet (robot)|Kismet]], a robot with rudimentary social skills.]]
Emotion and social skills play two roles for an intelligent agent:<ref>{{Harvnb|Minsky|2007}}, {{Harvnb|Picard|1997}}</ref>
*It must be able to predict the actions of others, by understanding their motives and emotional states. (This involves elements of [[game theory]], [[decision theory]], as well as the ability to model human emotions and the perceptual skills to detect emotions.)
*For good [[human-computer interaction]], an intelligent machine also needs to ''display'' emotions — at the very least it must appear polite and sensitive to the humans it interacts with. At best, it should appear to have normal emotions itself.

====General intelligence====
{{Main|strong AI|AI-complete}}
Most researchers hope that their work will eventually be incorporated into a machine with ''general'' intelligence (known as [[strong AI]]), combining all the skills above and exceeding human abilities at most or all of them.<ref name=GI/> A few believe that [[anthropomorphic]] features like [[artificial consciousness]] or an [[artificial brain]] may be required for such a project.

Many of the problems above are considered [[AI-complete]]: to solve one problem, you must solve them all. For example, even a straightforward, specific task like [[machine translation]] requires that the machine follow the author's argument ([[#Deduction, reasoning, problem solving|reason]]), know what it's talking about ([[#Knowledge representation|knowledge]]), and faithfully reproduce the author's intention ([[#Social intelligence|social intelligence]]). [[Machine translation]], therefore, is believed to be AI-complete: it may require [[strong AI]] to be done as well as humans can do it.<ref>{{Harvnb|Shapiro|1992|p=9}}</ref>

====General limitations====

There are three general limitations in AI, commonly stated as stupidity, ignorance, and laziness.{{Fact|date=January 2008}} Most real-world problems have one or more of these factors.

* Stupidity: One does not always know how to compute a perfect solution.
*: E.g. there is no known method to directly factor the multiple of two primes.
*:The solution to stupidity is generally to use an alternative method to approach the answer, or one that results in an answer that is "good enough". E.g. for prime factorization, there are various heuristics to determine whether a large number is prime.
* Ignorance: One does not always have the necessary information to compute a perfect solution.
*: E.g. in the game [[Stratego]], the opponent's pieces are of known position, but start as of unknown identity. In Texas hold 'em [[poker]], the order of the deck and thus the other players' cards as well as the flop cards are unknown.
*: The solution to ignorance is generally the strategic discovery of new information or acceptance of unknowns - e.g. in [[Stratego]] one can bait or attack pieces to uncover their identity, or guess that the opponent's flag is in a well-protected location rather than in an easily reachable one. In [[poker]], one can try to determine the other players' cards by their reactions during bidding, as well as knowing the simple probability of various flop cards and going with whatever is most likely to succeed overall.
* Laziness: One does not always have the time to compute a perfect solution.
*: E.g. in [[chess]], though the state is entirely known, as well as the rules of the game and the value of its outcomes, there is not enough computing power available to exhaustively go through all possible games. [[Checkers]], however, has been solved relatively recently by exactly this method.{{Fact|date=January 2008}}
*: The solution to laziness is generally a utility [[heuristic]] - e.g. in chess, one can take a guess at how likely a certain move is to result in a win or a loss even without having fully computed its outcomes, based on generalized ideas such as defensive positions, numeric piece values, etc.

=== Approaches to AI ===
Artificial intelligence is a young science and is still a fragmented collection of subfields. At present, there is no established unifying theory that links the subfields into a coherent whole.

====Cybernetics and brain simulation====
In the 40s and 50s, a number of researchers explored the connection between [[neurology]], [[information theory]], and [[cybernetics]]. Some of them built machines that used electronic networks to exhibit rudimentary intelligence, such as [[W. Grey Walter]]'s [[Turtle (robot)|turtles]] and the [[Johns Hopkins Beast]]. Many of these researchers gathered for meetings of the [[Teleological Society]] at Princeton and the [[Ratio Club]] in England.<ref>
Among the researchers who laid the foundations of [[cybernetic]]s, [[information theory]] and [[neural networks]] were [[Claude Shannon]], [[Norbert Weiner]], [[Warren McCullough]], [[Walter Pitts]], [[Donald Hebb]], [[Donald McKay]], [[Alan Turing]] and [[John Von Neumann]].
{{Harvnb|McCorduck|2004|pp=51-107}}
{{Harvnb|Crevier|1993|pp=27-32}},
{{Harvnb|Russell|Norvig|2003|pp=15,940}},
{{Harvnb|Moravec|1988|p=3}}.
</ref>

====Traditional symbolic AI====
When access to digital computers became possible in the middle 1950s, AI research began to explore the possibility that human intelligence could be reduced to symbol manipulation. The research was centered in three institutions: [[Carnegie Mellon University|CMU]], [[Stanford]] and [[MIT]], and each one developed its own style of research. [[John Haugeland]] named these approaches to AI "good old fashioned AI" or "[[GOFAI]]".<ref>
{{Harvnb|Haugeland|1985|pp=112-117}}
</ref>

;Cognitive simulation :[[Economist]] [[Herbert Simon]] and [[Alan Newell]] studied human problem solving skills and attempted to formalize them, and their work laid the foundations of the field of artificial intelligence, as well as [[cognitive science]], [[operations research]] and [[management science]]. Their research team performed [[psychology|psychological]] experiments to demonstrate the similarities between human problem solving and the programs (such as their "[[General Problem Solver]]") they were developing. This tradition, centered at [[Carnegie Mellon University]],<ref>Then called [[Carnegie Tech]]</ref> would eventually culminate in the development of the [[Soar (cognitive architecture)|Soar]] architecture in the middle 80s.<ref>
{{Harvnb|Crevier|1993|pp=52-54, 258-263}}, {{Harvnb|Nilsson|1998|p=275}}
</ref>

;Logical AI :Unlike [[Alan Newell|Newell]] and [[Herbert Simon|Simon]], [[John McCarthy (computer scientist)|John McCarthy]] felt that machines did not need to simulate human thought, but should instead try find the essence of abstract reasoning and problem solving, regardless of whether people used the same algorithms.<ref>
See [http://books.google.com/books?id=PEkqAAAAMAAJ&q=%22we+don't+care+if+it's+psychologically+real%22&dq=%22we+don't+care+if+it's+psychologically+real%22&output=html&pgis=1 Science at Google Books],
and [http://www.engagingexperience.com/2006/07/ai50_ai_past_pr.html McCarthy's presentation at AI@50]
</ref> His laboratory at [[Stanford University|Stanford]] ([[Stanford Artificial Intelligence Laboratory|SAIL]]) focussed on using formal [[logic]] to solve wide variety of problems, including [[knowledge representation]], [[automated planning and scheduling|planning]] and [[machine learning|learning]]. Work in logic led to the development of the programming language [[Prolog]] and the science of [[logic programming]].<ref>
{{Harvnb|Crevier|1993|pp=193-196}}
</ref>

;"Scruffy" symbolic AI :Researchers at [[MIT]] (such as [[Marvin Minsky]] and [[Seymour Papert]]) found that solving difficult problems in [[computer vision|vision]] and [[natural language processing]] required ad-hoc solutions -- they argued that there was no [[silver bullet]], no simple and general principle (like [[logic]]) that would capture all the aspects of intelligent behavior. An important realization was that AI required large amounts of [[commonsense knowledge]], and that this had to be engineered one complicated concept at a time. [[Roger Schank]] described their "anti-logic" approaches as "[[Neats vs. scruffies|scruffy]]" (as opposed to the "[[Neats vs. scruffies|neat]]" paradigms at [[CMU]] and [[Stanford]]),<ref>
{{Harvnb|Crevier|1993|pp=163-176}}. [[Neats vs. scruffies]]: {{Harvnb|Crevier|1993|pp=168}}.
</ref> and this still forms the basis of research into [[commonsense knowledge]], such as [[Doug Lenat]]'s [[Cyc]].

;Knowledge based AI :When computers with large memories became available around 1970, researchers from all three traditions began to build [[knowledge representation|knowledge]] into AI applications. This "knowledge revolution" led to the development and deployment of [[expert system]]s, the first truly successful form of AI software.<ref>{{Harvnb|Crevier|1993|pp=145-162}}</ref>

==== Sub-symbolic AI ====
During the 1960s, symbolic approaches had achieved great success at simulating high-level thinking in small demonstration programs. Approaches based on [[cybernetics]] or [[neural network]]s were abandoned or pushed into the background.<ref>
The most dramatic case of sub-symbolic AI being pushed into the background was the devastating critique of [[perceptron]]s by [[Marvin Minsky]] and [[Seymour Papert]] in 1969. See [[History of AI]], [[AI winter]], or [[Frank Rosenblatt]]. {{Harv|Crevier|1993|pp=102-105}}.
</ref>
By the 1980s, however, progress in symbolic AI seemed to stall and many believed that symbolic systems would never be able to imitate all the processes of human cognition, especially [[machine perception|perception]], [[robotics]], [[machine learning|learning]] and [[pattern recognition]]. A number of researchers began to look into "sub-symbolic" approaches to specific AI problems.<ref>{{Harvtxt|Nilsson|1998|p=7}} characterizes these newer approaches to AI as "sub-symbolic".</ref>

;Bottom-up, situated, behavior based or nouvelle AI :Researchers from the related field of [[robotics]], such as [[Rodney Brooks]], rejected symbolic AI and focussed on the basic engineering problems that would allow robots to move and survive.<ref>{{Harvnb|Brooks|1990}} and {{Harvnb|Moravec|1988}}</ref> Their work revived the non-symbolic viewpoint of the early [[cybernetic]]s researchers of the 50s and reintroduced the use of [[control theory]] in AI. These "bottom-up" approaches are known as [[behavior-based AI]], [[situated]] AI or [[Nouvelle AI]], and are closely tied to [[embodied cognitive science]].

;Computational Intelligence :Interest in [[neural networks]] and "[[connectionism]]" was revived by [[David Rumelhart]] and others in the middle 1980s.<ref>
{{Harvnb|Crevier|1993|pp=214-215}} and {{Harvnb|Russell|Norvig|2003|p=25}}</ref> These and other sub-symbolic approaches, such as [[fuzzy system]]s and [[evolutionary computation]], are now studied collectively by the emerging discipline of [[computational intelligence]].<ref>
See [http://www.ieee-cis.org/ IEEE Computational Intelligence Society]
</ref>

;The new neats :In the 1990s, AI researchers developed sophisticated mathematical tools to solve specific subproblems. These tools are truly [[scientific method|scientific]], in the sense that their results are both measurable and verifiable, and they have been responsible for many of AI's recent successes. The shared mathematical language has also permitted a high level of collaboration with more established fields (like [[mathematics]], [[economics]] or [[operations research]]). {{Harvtxt|Russell|Norvig|2003}} describe this movement as nothing less than a "revolution" and "the victory of the [[neats and scruffies|neats]]."<ref>
{{Harvnb|Russell|Norvig|2003|p=25-26}}
</ref>

====Intelligent agent paradigm====
The "[[intelligent agent]]" paradigm became widely accepted during the 1990s.<ref>
"The whole-agent view is now widely accepted in the field" {{Harvnb|Russell|Norvig|2003|p=55}}.
</ref><ref name=IA>
The [[intelligent agent]] paradigm is discussed in major AI textbooks, such as:
{{Harvnb|Russell|Norvig|2003|pp=27, 32-58, 968-972}},
{{Harvnb|Poole|Mackworth|Goebel|1998|pp=7-21}},
{{Harvnb|Luger|Stubblefield|2004|pp=235-240}}
</ref>
Although earlier researchers had proposed modular "divide and conquer" approaches to AI,<ref>
For example, both [[John Doyle]] {{Harv|Doyle|1983}} and [[Marvin Minsky]]'s popular classic ''[[The Society of Mind]]'' {{Harv|Minsky|1986}} used the word "agent" to describe modular AI systems.
</ref>
the [[intelligent agent]] did not reach its modern form until [[Judea Pearl]], [[Alan Newell]] and others brought concepts from [[decision theory]] and [[economics]] into the study of AI.<ref>
{{Harvnb|Russell|Norvig|2003|pp=27, 55}}
</ref>
When the [[economics|economist's]] definition of a [[agent (economics)|rational agent]] was married to [[computer science]]'s definition of an [[object oriented programming|object]] or [[module (programming)|module]], the [[intelligent agent]] paradigm was complete.

An [[intelligent agent]] is a system that perceives its [[agent environment|environment]] and takes actions which maximizes its chances of success. The simplest intelligent agents are programs that solve specific problems. The most complicated intelligent agents would be rational, thinking human beings.<ref name=IA/>

The paradigm gives researchers license to study isolated problems and find solutions that are both verifiable and useful, without agreeing on one single approach. An agent that solves a specific problem can use any approach that works — some agents are symbolic and logical, some are sub-symbolic [[neural network]]s and some can be based on new approaches (without forcing researchers to reject old approaches that have proven useful). The paradigm gives researchers a common language to describe problems and share their solutions with each other and with other fields—such as [[decision theory]]—that also use concepts of abstract agents.

==== Integrating the approaches ====

An [[agent architecture]] or [[cognitive architecture]] allows researchers to build more versatile and intelligent systems out of interacting [[intelligent agents]] in a [[multi-agent system]].<ref>
[[Agent architecture]]s, [[hybrid intelligent system]]s, and [[multi-agent system]]s:
{{Harvnb|ACM|1998|loc=I.2.11}},
{{Harvtxt|Russell|Norvig|1998|pp=27, 932, 970-972}} and
{{Harvtxt|Nilsson|1998|loc=chpt. 25}}
</ref>
A system with both symbolic and sub-symbolic components is a [[hybrid intelligent system]], and the study of such systems is [[artificial intelligence systems integration]].


===Tools of AI research===
In the course of 50 years of research, AI has developed a large number of tools to solve the most difficult problems in [[computer science]]. A few of the most general of these methods are discussed below.

====Search====
{{Main|search algorithm}}
Many problems in AI can be solved in theory by intelligently searching through many possible solutions:<ref>
[[Search algorithm]]s:
{{Harvnb|Russell|Norvig|2003|pp=59-189}},
{{Harvnb|Poole|Mackworth|Goebel|1998|pp=113-163}},
{{Harvnb|Luger|Stubblefield|2004|pp=79-164, 193-219}},
{{Harvnb|Nilsson|1998|loc=chpt. 7-12}}
</ref>
* [[Machine reasoning|Reasoning]] can be reduced to performing a search. For example, in game playing, the agent can search through a tree of possible moves and counter moves to find a strategy that improves its position. (Tools for two person games include [[minimax]] and [[alpha-beta pruning]].)<ref>
Adversarial search:
{{Harvnb|Russell|Norvig|2003|pp=161-185}},
{{Harvnb|Luger|Stubblefield|2004|pp=150-157}},
{{Harvnb|Nilsson|1998|loc=chpt. 12}}
</ref> Logical proof can be viewed as searching for a path that leads from [[premise]]s to [[conclusion]]s, where each step is the application of an [[inference rule]].<ref name=FC>
[[Forward chaining]], [[backward chaining]], [[Horn clause]]s, and logical deduction as search:
{{Harvnb|Russell|Norvig|2003|pp=217-225, 280-294}},
{{Harvnb|Poole|Mackworth|Goebel|1998|pp=~46-52}},
{{Harvnb|Luger|Stubblefield|2004|pp=62-73}},
{{Harvnb|Nilsson|1998|loc=chpt. 4.2, 7.2}}
</ref> Many other reasoning problems, such as [[constraint satisfaction]]<ref>
Constraint satisfaction:
{{Harvnb|Russell|Norvig|2003|pp=137-156}},
{{Harvnb|Poole|Mackworth|Goebel|1998|pp=pp. 147-163}}
</ref> and [[dynamic programming]]<ref>
[[Dynamic programming]]:
{{Harvnb|Russell|Norvig|2003|p=293}},
{{Harvnb|Poole|Mackworth|Goebel|1998|pp=145-147}},
{{Harvnb|Nilsson|1998|p=178}}
</ref> are solved using a form of search.
*[[automated planning and scheduling|Planning]] algorithms search through trees of goals and subgoals, attempting to find a path to a target goal.<ref>
[[State space search]] and [[automated planning and scheduling|planning]]:
{{Harvnb|Russell|Norvig|2003|pp=382-387}},
{{Harvnb|Poole|Mackworth|Goebel|1998|pp=298-305}},
{{Harvnb|Nilsson|1998|loc=chpt. 10.1-2}}
</ref> These sets of goals and subgoals can be represented with graphs (as in the [[graphplan]] algorithm),<ref>
[[Graphplan]]:
{{Harvnb|Russell|Norvig|2003|pp= 395-402}}
</ref> or in a [[hierarchical task network]].<ref>
[[Hierarchical task network]]:
{{Harvnb|Russell|Norvig|2003|pp=422-430}}
</ref>
*[[Robotics]] algorithms for moving limbs and grasping objects use [[local search]]es in [[configuration space]].<ref name=CS/>


There are several types of search algorithms:
* "Naive" search algorithms, such as [[breadth first search]], [[depth first search]] and general [[state space search]].<ref>
Naive searches:
{{Harvnb|Russell|Norvig|2003|pp=59-93}},
{{Harvnb|Poole|Mackworth|Goebel|1998|pp=113-132}},
{{Harvnb|Luger|Stubblefield|2004|pp=79-121}},
{{Harvnb|Nilsson|1998|loc=chpt. 8}}
</ref>
* [[Heuristic]] or "informed" search. The naive algorithms quickly run into problems: the [[search space]] quickly can expand to [[astronomical]] size in a "[[combinatorial explosion]]. "AI researchers realized from the beginning that they would need to trim the search space using [[heuristics]] or "rules of thumb".<ref>
[[John McCarthy (computer scientist)|John McCarthy]] writes that "the combinatorial explosion problem has been recognized in AI from the beginning" in [http://www-formal.stanford.edu/jmc/reviews/lighthill/lighthill.html Review of Lighthill report]
</ref> The use of [[heuristics]] led to the development of intelligent searches such as greedy [[best-first search|best first]] and [[A*]].<ref>
[[Heuristic]] or informed searches:
{{Harvnb|Russell|Norvig|2003|pp= 94-109}},
{{Harvnb|Poole|Mackworth|Goebel|1998|pp=pp. 132-147}},
{{Harvnb|Luger|Stubblefield|2004|pp= 133-150}},
{{Harvnb|Nilsson|1998|loc=chpt. 9}}
</ref>
* Local searches, such as [[hill climbing]], [[simulated annealing]] and [[beam search]], use techniques borrowed from [[optimization (mathematics)|optimization theory]].<ref>
[[optimization (mathematics)|Optimization]] searches:
{{Harvnb|Russell|Norvig|2003|pp=110-116,120-129}},
{{Harvnb|Poole|Mackworth|Goebel|1998|pp=56-163}},
{{Harvnb|Luger|Stubblefield|2004|pp= 127-133}}
</ref>
* [[Genetic algorithm]]s are a form of [[optimization (mathematics)|optimization]] search that imitates the process of [[natural selection]], searching for an artificial [[phenotype]] (i.e. any sort of pattern) which passes a [[fitness (biology)|fitness]] measure by producing many copies of the most successful versions (imitating [[inheritance]]) and modifying them slightly (imitating [[mutation]]).<ref>
[[Genetic algorithms]]:
{{Harvnb|Russell|Norvig|2003|pp=116-119}},
{{Harvnb|Poole|Mackworth|Goebel|1998|pp=162}},
{{Harvnb|Luger|Stubblefield|2004|pp=509-530}},
{{Harvnb|Nilsson|1998|loc=chpt. 4.2}}
</ref>

==== Logic ====
{{Main|logic programming}}

[[Logic]]<ref>
[[Logic]]:
{{Harvnb|ACM|1998|loc=~I.2.3}},
{{Harvnb|Russell|Norvig|2003|pp=194-310}},
{{Harvnb|Luger|Stubblefield|2004|pp=35-77}},
{{Harvnb|Nilsson|1998|loc=chpt. 13-16}}
</ref>
was introduced into AI research by [[John McCarthy (computer scientist)|John McCarthy]] in his 1958 [[Advice Taker]] proposal. The most important technical development was [[J. Alan Robinson]]'s discovery of the [[resolution (logic)|resolution]] and [[unification]] algorithm for logical deduction in 1963. This procedure is simple, complete and entirely algorithmic, and can easily be performed by digital computers.<ref>
[[Resolution (logic)|Resolution]] and [[unification]]:
{{Harvnb|Russell|Norvig|2003|pp=213-217, 275-280, 295-306}},
{{Harvnb|Poole|Mackworth|Goebel|1998|pp=56-58}},
{{Harvnb|Luger|Stubblefield|2004|pp=554-575}},
{{Harvnb|Nilsson|1998|loc=chpt. 14 & 16}}
</ref> However, a naive implementation of the algorithm quickly leads to a [[combinatorial explosion]] or an [[infinite loop]]. In 1974, [[Robert Kowalski]] suggested representing logical expressions as [[Horn clauses]] (statements in the form of rules: "if ''p'' then ''q''"), which reduced logical deduction to [[backward chaining]] or [[forward chaining]]. This greatly alleviated (but did not eliminate) the problem.<ref name=FC/><ref name=HLP>
History of logic programming:
{{Harvnb|Crevier|1993|pp=190-196}}.
Advice Taker:
{{Harvnb|McCorduck|2004|p=51}},
{{Harvnb|Russell|Norvig|2003|pp=19}}
</ref>

Logic is used for knowledge representation and problem solving, but it can be applied to other problems as well. For example, the [[satplan]] algorithm uses logic for [[automated planning and scheduling|planning]],<ref>
[[Satplan]]:
{{Harvnb|Russell|Norvig|2003|pp=402-407}},
{{Harvnb|Poole|Mackworth|Goebel|1998|pp=300-301}},
{{Harvnb|Nilsson|1998|loc=chpt. 21}}
</ref>
and [[inductive logic programming]] is a method for [[machine learning|learning]].<ref>
[[Explanation based learning]], [[relevance based learning]], [[inductive logic programming]], [[case based reasoning]]:
{{Harvnb|Russell|Norvig|2003|pp=678-710}},
{{Harvnb|Poole|Mackworth|Goebel|1998|pp=414-416}},
{{Harvnb|Luger|Stubblefield|2004|pp=~422-442}},
{{Harvnb|Nilsson|1998|loc=chpt. 10.3, 17.5}}
</ref>

There are several different forms of logic used in AI research.
* [[Propositional logic]]<ref>
[[Propositional logic]]:
{{Harvnb|Russell|Norvig|2003|pp=204-233}},
{{Harvnb|Luger|Stubblefield|2004|pp=45-50}}
{{Harvnb|Nilsson|1998|loc=chpt. 13}}
</ref> or [[sentential logic]] is the logic of statements which can be true or false.
* [[First order logic]]<ref>
[[First order logic]] and features such as [[equality]]:
{{Harvnb|ACM|1998|loc=~I.2.4}},
{{Harvnb|Russell|Norvig|2003|pp=240-310}},
{{Harvnb|Poole|Mackworth|Goebel|1998|pp=268-275}},
{{Harvnb|Luger|Stubblefield|2004|pp=50-62}},
{{Harvnb|Nilsson|1998|loc=chpt. 15}}
</ref> also allows the use of [[quantifier]]s and [[predicate]]s, and can express facts about objects, their properties, and their relations with each other.
* [[Fuzzy logic]], a version of first order logic which allows the truth of statement to represented as a value between 0 and 1, rather than simply True (1) or False (0). [[Fuzzy system]]s can be used for uncertain reasoning and have been widely used in modern industrial and consumer product control systems.<ref name=FL>
[[Fuzzy logic]]:
{{Harvnb|Russell|Norvig|2003|pp=526-527}}
</ref>
* [[Default logic]]s, [[non-monotonic logic]]s and [[circumscription]] are forms of logic designed to help with default reasoning and the [[qualification problem]].<ref name=NML/>
* Several extensions of logic have been designed to handle specific domains of [[knowledge representation|knowledge]], such as: [[description logic]]s;<ref name=DL/> [[situation calculus]], [[event calculus]] and [[fluent calculus]] (for representing events and time);<ref name=SC/> [[Causality#causal calculus|causal calculus]];<ref name=CC/> [[belief calculus]]; and [[modal logic]]s.<ref name=BC/>

==== Stochastic methods ====
{{Main|Bayesian network|hidden Markov model|Kalman filter}}

Starting in the late 80s and early 90s, [[Judea Pearl]] and others championed the use of [[stochastic]] or [[probability|probabilistic]] methods in artificial intelligence.<ref>
{{Harvnb|Russell|Norvig|2003|pp=25-26}} (on [[Judea Pearl]]'s contribution).
Stochastic methods are described in all the major AI textbooks:
{{Harvnb|ACM|1998|loc=~I.2.3}},
{{Harvnb|Russell|Norvig|2003|pp=462-644}},
{{Harvnb|Poole|Mackworth|Goebel|1998|pp=345-395}},
{{Harvnb|Luger|Stubblefield|2004|pp=165-191, 333-381}},
{{Harvnb|Nilsson|1998|loc=chpt. 19}}
</ref>
Researchers have used principles from [[probability]] theory<ref>
[[Probability]]:
{{Harvnb|Russell|Norvig|2003|pp=462-489}},
{{Harvnb|Poole|Mackworth|Goebel|1998|pp=346-366}},
{{Harvnb|Luger|Stubblefield|2004|pp=~165-182}},
{{Harvnb|Nilsson|1998|loc=chpt. 19.1}}
</ref>
to devise a number of powerful tools.

[[Bayesian network]]s<ref>
[[Bayesian network]]s:
{{Harvnb|Russell|Norvig|2003|pp=492-523}},
{{Harvnb|Poole|Mackworth|Goebel|1998|pp=361-381}},
{{Harvnb|Luger|Stubblefield|2004|pp=~182-190, ~363-379}},
{{Harvnb|Nilsson|1998|loc=chpt. 19.3-4}}
</ref>
have been applied to a large number of problems, such as: uncertain reasoning (using the [[Bayesian inference]] algorithm),<ref>
[[Bayesian inference]] algorithm:
{{Harvnb|Russell|Norvig|2003|pp=504-519}},
{{Harvnb|Poole|Mackworth|Goebel|1998|pp=361-381}},
{{Harvnb|Luger|Stubblefield|2004|pp=~363-379}},
{{Harvnb|Nilsson|1998|loc=chpt. 19.4 & 7}}
</ref>
[[Machine learning|learning]] (using the [[expectation-maximization algorithm]]),<ref name=BL>
[[Bayesian]] learning and the [[expectation-maximization algorithm]]
{{Harvnb|Russell|Norvig|2003|pp=712-724}},
{{Harvnb|Poole|Mackworth|Goebel|1998|pp=424-433}},
{{Harvnb|Nilsson|1998|loc=chpt. 20}}
</ref>
and [[Automated planning and scheduling|planning]] (using [[decision network]]s).<ref name=BDN>
[[Bayesian]] [[decision network]]s:
{{Harvnb|Russell|Norvig|2003|pp=597-600}}
</ref>

Probabilistic methods have been particularly successful at dealing with processes that occur over time. Several successful algorithms have been developed for filtering, prediction, smoothing and finding explanations for streams of data,<ref>
{{Harvnb|Russell|Norvig|2003|pp=537-581}}
</ref>
such as [[hidden Markov model]]s,<ref>
[[Hidden Markov model]]:
{{Harvnb|Russell|Norvig|2003|pp=549-551}}
</ref>
[[Kalman filter]]s<ref>
[[Kalman filter]]:
{{Harvnb|Russell|Norvig|2003|pp=551-557}}
</ref>
and [[dynamic Bayesian network]]s.<ref>
[[Dynamic Bayesian network]]:
{{Harvnb|Russell|Norvig|2003|pp=551-557}}
</ref>
These tools are used for the problems of [[machine perception|perception]] (such as [[pattern matching]]) and [[machine learning|learning]].

==== Economic models ====
{{Main|utility theory|decision theory|game theory}}

AI has been able to use tools drawn from [[economic]]s, such as [[decision theory]] and [[decision analysis]],<ref name=DTA/> Bayesian [[decision network]]s,<ref name=BDN/> [[applied information economics|information value theory]],<ref name=IVT/>
[[Markov decision process]]es,<ref name=MDPDDN>
[[Markov decision process]]es and dynamic [[decision network]]s:
{{Harvnb|Russell|Norvig|2003|pp=613-631}}
</ref>
dynamic [[decision network]]s,<ref name=MDPDDN/>
[[game theory]] and [[mechanism design]]<ref name=GTMD>
[[Game theory]] and [[mechanism design]]:
{{Harvnb|Russell|Norvig|2003|pp=631-643}}
</ref> These tools have been especially important for [[automated planning and scheduling|planning]] problems.

==== Classifiers and statistical learning methods ====
{{Main|classifier (mathematics)|statistical classification|machine learning}}

The simplest AI applications can be divided into two types: classifiers ("if shiny then diamond") and controllers ("if shiny then pick up"). Controllers do however also classify conditions before inferring actions, and therefore classification forms a central part of many AI systems.

[[Classifier (mathematics)|Classifiers]]<ref>
Statistical learning methods and [[Classifier (mathematics)|classifiers]]:
{{Harvnb|Russell|Norvig|2003|pp=712-754}},
{{Harvnb|Luger|Stubblefield|2004|pp=453-541}}
</ref> are functions that use [[pattern matching]] to determine a closest match. They can be tuned according to examples, making them very attractive for use in AI. These examples are known as observations or patterns. In supervised learning, each pattern belongs to a certain predefined class. A class can be seen as a decision that has to be made. All the observations combined with their class labels are known as a data set.

When a new observation is received, that observation is classified based on previous experience. A classifier can be trained in various ways; there are mainly statistical and [[machine learning]] approaches.

A wide range of classifiers are available, each with its strengths and weaknesses. Classifier performance depends greatly on the characteristics of the data to be classified. There is no single classifier that works best on all given problems; this is also referred to as the "no free lunch" theorem. Various empirical tests have been performed to compare classifier performance and to find the characteristics of data that determine classifier performance. Determining a suitable classifier for a given problem is however still more an art than science.

The most widely used classifiers are the [[Artificial neural network|neural network]],<ref name=NN/>
[[kernel methods]] such as the [[support vector machine]],<ref>
[[Kernel methods]]:
{{Harvnb|Russell|Norvig|2003|pp=749-752}}
</ref>
[[k-nearest neighbor algorithm]],<ref>
[[K-nearest neighbor algorithm]]:
{{Harvnb|Russell|Norvig|2003|pp=733-736}}
</ref> [[Gaussian mixture model]],<ref>
[[Gaussian mixture model]]:
{{Harvnb|Russell|Norvig|2003|pp=725-727}}
</ref>
[[naive Bayes classifier]],<ref>
[[Naive Bayes classifier]]:
{{Harvnb|Russell|Norvig|2003|pp=718}}
</ref>
and [[decision tree]].<ref name=DT>
[[Alternating decision tree|Decision tree]]:
{{Harvnb|Russell|Norvig|2003|pp=653-664}},
{{Harvnb|Poole|Mackworth|Goebel|1998|pp=403-408}},
{{Harvnb|Luger|Stubblefield|2004|pp=408-417}}
</ref>
The performance of these classifiers have been compared over a wide range of classification tasks<ref>
{{cite-web| last=van der Walt | first=Christiaan | url=http://www.patternrecognition.co.za/publications/cvdwalt_data_characteristics_classifiers.pdf|title= Data characteristics that determine classifier performance}}</ref>
in order to find data characteristics that determine classifier performance.

==== Neural networks ====
{{main|neural networks|connectionism}}

[[Image:Artificial neural network.svg|thumb|180px|A neural network is an interconnected group of nodes, akin to the vast network of [[neuron]]s in the [[human brain]].]]
The study of [[neural network]]s<ref name=NN>
Neural networks and connectionism:
{{Harvnb|Russell|Norvig|2003|pp=736-748}},
{{Harvnb|Poole|Mackworth|Goebel|1998|pp=408-414}},
{{Harvnb|Luger|Stubblefield|2004|pp=453-505}},
{{Harvnb|Nilsson|1998|loc=chpt. 3}}
</ref>
began with [[cybernetic]]s researchers, working in the decade before the field AI research was founded. In the 1960s [[Frank Rosenblatt]] developed an important early version, the [[perceptron]].<ref>
[[Perceptrons]]:
{{Harvnb|Russell|Norvig|2003|pp=740-743}},
{{Harvnb|Luger|Stubblefield|2004|pp=458-467}}
</ref>
[[Paul Werbos]] discovered the [[backpropagation]] algorithm in 1974,<ref>
[[Backpropagation]]:
{{Harvnb|Russell|Norvig|2003|pp=744-748}},
{{Harvnb|Luger|Stubblefield|2004|pp=467-474}},
{{Harvnb|Nilsson|1998|loc=chpt. 3.3}}
</ref>
which led to a renaissance in neural network research and [[connectionism]] in general in the middle 1980s. The [[Hopfield net]], a form of attractor network, was first described by [[John Hopfield]] in 1982.

Neural networks are applied to the problem of [[machine learning|learning]], using such techniques as [[Hebbian learning]]<ref name=LS474>
[[Competitive learning]], [[Hebbian theory|Hebbian]] coincidence learning, [[Hopfield network]]s and attractor networks:
{{Harvnb|Luger|Stubblefield|2004|pp=474-505}}.
</ref>
and the relatively new field of [[Hierarchical Temporal Memory]] which simulates the architecture of the [[neocortex]].<ref>
{{Harvnb|Hawkins|Blakeslee|2004}}
</ref>

==== Social and emergent models ====
{{Main|evolutionary computation}}

Several algorithms for [[machine learning|learning]] use tools from [[evolutionary computation]], such as [[genetic algorithms]]<ref>
[[Genetic algorithm]]s for learning:
{{Harvnb|Luger|Stubblefield|2004|pp=509-530}},
{{Harvnb|Nilsson|1998|loc=chpt. 4.2}}
</ref> and [[swarm intelligence]].<ref>
[[Artificial life]] and society based learning:
{{Harvnb|Luger|Stubblefield|2004|pp=530-541}}
</ref>

==== Control theory ====
{{Main|intelligent control}}

[[Control theory]], the grandchild of [[cybernetics]], has many important applications, especially in [[robotics]].<ref>
[[Control theory]]:
{{Harvnb|ACM|1998|loc=~I.2.8}},
{{Harvnb|Russell|Norvig|2003|pp=926-932}}
</ref>

==== Specialized languages ====
{{Main|IPL|Lisp (programming language)|Prolog|STRIPS|Planner}}

AI researchers have developed several specialized languages for AI research:
* [[IPL]], one of the first programming languages, developed by [[Alan Newell]], [[Herbert Simon]] and [[J. C. Shaw]].<ref>
{{Harvnb|Crevier|1993|p=46-48}}
</ref>
* [[Lisp programming language|Lisp]]<ref>
[[Lisp (programming language)|Lisp]]:
{{Harvnb|Luger|Stubblefield|2004|pp=723-821}}
</ref> was developed by [[John McCarthy (computer scientist)|John McCarthy]] at [[MIT]] in 1958.<ref>
{{Harvnb|Crevier|1993|pp=59-62}},
{{Harvnb|Russell|Norvig|2003|p=18}}
</ref> There are many dialects of Lisp in use today.
* [[Prolog]],<ref>
[[Prolog]]:
{{Harvnb|Poole|Mackworth|Goebel|1998|pp=477-491}},
{{Harvnb|Luger|Stubblefield|2004|pp=641-676, 575-581}}
</ref> a language based on [[logic programming]], was invented by [[France|French]] researchers [[Alain Colmerauer]] and [[Phillipe Roussel]], in collaboration with [[Robert Kowalski]] of the [[University of Edinburgh]].<ref name=HLP/>
* [[STRIPS]], a planning language developed at [[Stanford]] in the 1960s.
* [[Planner]] developed at [[MIT]] around the same time.
AI applications are also often written in standard languages like [[C++]] and languages designed for mathematics, such as [[Matlab]] and [[Lush]].

===Competitions and prizes===
[[Image:Robocup.legged.leauge.2004.nk.jpg|thumb|right|200px|A legged league game from RoboCup 2004 in Lisbon, Portugal.]]
The [[Loebner prize]] is an annual competition to determine the best [[Turing test]] competitors. The winner is the computer system that, in the judges' opinions, demonstrates the "most human" conversational behaviour (with learning AI [[Ultra Hal]] winning in [[2007]], [[Jabberwacky]] in [[2005]] and [[2006]], and [[Artificial Linguistic Internet Computer Entity|A.L.I.C.E.]] before that), they have an additional prize for a system that in their opinion passes a Turing test. This second prize has not yet been awarded.

The [[DARPA Grand Challenge]] is an annual race for a $2 million prize where [[driverless car]]s must travel over a hundred miles without any communication with humans, using [[GPS]], computers and a sophisticated array of sensors. The challenge is aimed at a congressional mandate stating that by 2015 one-third of the operational ground combat vehicles of the US Armed Forces should be unmanned.<ref>[http://www.darpa.mil/grandchallenge04/sponsor_toolkit/congress_lang.pdf Congressional Mandate] DARPA</ref> In November 2007, DARPA introduced the [[DARPA Urban Challenge]], a sixty-mile urban area race.

A popular challenge amongst AI research groups is the [[RoboCup]] and [[Federation of International Robot-soccer Association|FIRA]] annual international robot soccer competitions. Hiroaki Kitano has formulated the International RoboCup Federation challenge: "In 2050 a team of fully autonomous humanoid robot soccer players shall win the soccer game, comply with the official rule of the FIFA, against the winner of the most recent World Cup."<ref>[http://www.robocup.org/Press/pr/RoboCup2003_020603eng.pdf The RoboCup2003 Presents: Humanoid Robots playing Soccer] PRESS RELEASE: 2 June 2003</ref>

A lesser known challenge to promote AI research is the annual [[Arimaa]] challenge match. The challenge offers a $10,000 prize until the year 2020 to develop a program that plays the board game [[Arimaa]] and defeats a group of selected human opponents.

==Applications of artificial intelligence==
===Business===
Banks use artificial intelligence systems to organize operations, invest in stocks, and manage properties. In August 2001, robots beat humans in a simulated [[stock trader|financial trading]] competition ([[BBC News]], 2001).<ref name="xppr">{{cite web
|url=http://news.bbc.co.uk/2/hi/business/1481339.stm
|title=Robots beat humans in trading battle
|year=[[August 8]] [[2001]]
|accessdate=2006-11-02
|work=BBC News, Business
|publisher=The British Broadcasting Corporation
}}</ref> A medical clinic can use artificial intelligence systems to organize bed schedules, make a staff rotation, and provide medical information. Many practical applications are dependent on [[artificial neural networks]], networks that pattern their organization in mimicry of a brain's neurons, which have been found to excel in pattern recognition. [[Financial institution]]s have long used such systems to detect charges or claims outside of the norm, flagging these for human investigation. Neural networks are also being widely deployed in [[homeland security]], speech and text recognition, [[medical diagnosis]] (such as in [[Concept Processing]] technology in [[EMR]] software), [[data mining]], and [[e-mail spam]] filtering.

[[Robot]]s have become common in many industries. They are often given jobs that are considered dangerous to humans. Robots have proven effective in jobs that are very repetitive which may lead to mistakes or accidents due to a lapse in concentration and other jobs which humans may find degrading. General Motors uses around 16,000 robots for tasks such as painting, welding, and assembly. [[Japan]] is the leader in using and producing robots in the world. In 1995, 700,000 robots were in use worldwide; over 500,000 of which were from Japan.<ref>[http://encarta.msn.com/encyclopedia_761564255/Robot.html "Robot," Microsoft® Encarta® Online Encyclopedia 2006 ]</ref>

===Toys and games===
The 1990s saw some of the first attempts to mass-produce domestically aimed types of basic Artificial Intelligence for education, or leisure. This prospered greatly with the [[Digital Revolution]], and helped introduce people, especially children, to a life of dealing with various types of AI, specifically in the form of [[Tamagotchi]]s and [[Giga Pet]]s, the [[Internet]] (example: basic search engine interfaces are one simple form), and the first widely released robot, [[Furby]]. A mere year later an improved type of [[domestic robot]] was released in the form of [[Aibo]], a robotic dog with intelligent features and [[autonomy]].

===List of applications===
;Typical problems to which AI methods are applied:
{{MultiCol}}
*[[Pattern recognition]]
**[[Optical character recognition]]
**[[Handwriting recognition]]
**[[Speech recognition]]
**[[Facial recognition system|Face recognition]]
*[[Artificial Creativity]]
{{ColBreak}}
*[[Computer vision]], [[Virtual reality]] and [[Image processing]]
*[[Diagnosis (artificial intelligence)]]
*[[Game theory]] and [[Strategic planning]]
*[[Game artificial intelligence]] and [[Computer game bot]]
*[[Natural language processing]], [[Translation]] and [[Chatterbot]]s
*[[Non-linear control]] and [[Robot]]ics
{{EndMultiCol}}

;Other fields in which AI methods are implemented:
{{MultiCol}}
*[[Artificial life]]
*[[Automated reasoning]]
*[[Automation]]
*[[Biologically-inspired computing]]
*[[Colloquis]]
*[[Concept mining]]
*[[Data mining]]
*[[Knowledge representation]]
*[[Semantic Web]]
*[[E-mail spam]] filtering
{{ColBreak}}
* [[Robot]]ics
**[[Behavior-based robotics]]
**[[Cognitive]]
**[[Cybernetics]]
**[[Developmental robotics]]
**[[Epigenetic robotics]]
**[[Evolutionary robotics]]
*[[Hybrid intelligent system]]
*[[Intelligent agent]]
*[[Intelligent control]]
*[[Litigation]]
{{EndMultiCol}}

;Lists of researchers, projects & publications
*[[:Category:Artificial intelligence researchers|List of AI researchers]]
*[[List of notable artificial intelligence projects|List of AI projects]]
*[[List of important publications in computer science#Artificial intelligence|List of important AI publications]]


==See also==
: ''Main list: [[List of basic artificial intelligence topics]]''

==Notes==
{{reflist|3}}

== References ==
=== Major AI textbooks===
* {{Citation|first=George|last= Luger |first2=William|last2= Stubblefield|year=2004|title=Artificial Intelligence: Structures and Strategies for Complex Problem Solving|edition=5th|publisher=The Benjamin/Cummings Publishing Company, Inc.|pages= 720|isbn=0-8053-4780-1|url=http://www.cs.unm.edu/~luger/ai-final/tocfull.html}}
* {{Citation|last=Nilsson|first=Nils|author-link=Nils Nilsson|year=1998|title=Artificial Intelligence: A New Synthesis|publisher=Morgan Kaufmann Publishers|isbn=978-1-55860-467-4}}
* {{Russell Norvig 2003}}
* {{Citation | first = David | last = Poole | first2 = Alan | last2 = Mackworth | first3 = Randy | last3 = Goebel | publisher = Oxford University Press | publisher-place = New York | year = 1998 | title = Computational Intelligence: A Logical Approach | url = http://www.cs.ubc.ca/spider/poole/ci.html | author-link=David Poole }}
=== Other sources ===
* {{Citation| last=ACM | first=(Association of Computing Machinery) |year=1998|title=ACM Computing Classification System: Artificial intelligence|url=http://www.acm.org/class/1998/I.2.html |author-link=Association of Computing Machinery}}
* {{Citation | last=Brooks | first=Rodney | author-link=Rodney Brooks | year =1990 | title = Elephants Don't Play Chess | journal = Robotics and Autonomous Systems | volume=6 | pages=3-15 | url=http://people.csail.mit.edu/brooks/papers/elephants.pdf | accessdate=30 August 2007}}
* {{Citation | last=Buchanan | first = Bruce G. | year= 2005 | title = A (Very) Brief History of Artificial Intelligence | magazine = AI Magazine | pages=53-60 | url=http://www.aaai.org/AITopics/assets/PDF/AIMag26-04-016.pdf | accessdate=30 August 2007 }}
* {{Crevier 1993}}
* {{Citation | last=Haugeland | first=John | author-link = John Haugeland | year = 1985 | title = Artificial Intelligence: The Very Idea | publisher=MIT Press| location= Cambridge, Mass. | isbn=0-262-08153-9 }}.
* {{Citation | last=Hawkins | first=Jeff | author-link=Jeff Hawkins | last2=Blakeslee | first2=Sandra | year=2004 | title=On Intelligence | publisher=Owl Books | location=New York, NY | isbn=0-8050-7853-3 }}.
* {{Citation | last=Kahneman | first=Daniel | author-link=Daniel Kahneman | last2=Slovic | first2= D. | last3=Tversky | first3=Amos | author3-link=Amos Tversky | year=1982 | title=Judgment under uncertainty: Heuristics and biases | location=New York |publisher=Cambridge University Press}}.
* {{Citation | last=Kurzweil | first=Ray | author-link=Ray Kurzweil | year=1999 | title=The Age of Spiritual Machines | publisher=Penguin Books | isbn=0-670-88217-8 }}
* {{Citation | last=Kurzweil | first=Ray | author-link=Ray Kurzweil | year=2005 | title=The Singularity is Near | publisher=Penguin Books | isbn=0-670-03384-7 }}
* {{Citation | last=Lakoff | first=George | author-link=George Lakoff | last2=Núñez | first2=Rafael E. | author2-link=Rafael Núñez | year=2000 | title=[[Where Mathematics Comes From|Where Mathematics Comes From: How the Embodied Mind Brings Mathematics into Being]] | publisher=Basic Books | isbn= 0-465-03771-2}}.
* {{Citation | last=Lenat | first = Douglas | year = 1989 | title = Building Large Knowledge-Based Systems | publisher = Addison-Wesley| author-link=Douglas Lenat }}
* {{Citation | last=Lighthill | first = Professor Sir James | year = 1973 | contribution= Artificial Intelligence: A General Survey | title = Artificial Intelligence: a paper symposium| publisher = Science Research Council|author-link=James Lighthill }}
* {{Citation | last=McCarthy | first=John | last2 = Minsky | first2 = Marvin | last3 = Rochester | first3 = Nathan | last4 = Shannon | first4 = Claude | url = http://www-formal.stanford.edu/jmc/history/dartmouth/dartmouth.html | title = A Proposal for the Dartmouth Summer Research Project on Artificial Intelligence | year = 1955 | author-link = John McCarthy (computer scientist) | author2-link = Marvin Minsky | author3-link = Nathan Rochester | author4-link = Claude Shannon}}.
* {{Citation | last=McCarthy | first=John | last2 = Hayes | first2=P. J. | year = 1969 | url=http://www-formal.stanford.edu/jmc/mcchay69.html | title= Some philosophical problems from the standpoint of artificial intelligence | journal =Machine Intelligence | volume= 4 | pages = 463-502 | author-link = John McCarthy (computer scientist) }}
* {{Citation | last=McCorduck | first=Pamela | year = 2004 | title = Machines Who Think | publisher=A. K. Peters, Ltd. | location=Natick, MA | edition=2nd | isbn=1-56881-205-1}}.
* {{Citation | last=Minsky | first=Marvin | author-link=Marvin Minsky | year = 1967 | title = Computation: Finite and Infinite Machines | publisher = Prentice-Hall | location=Englewood Cliffs, N.J. }}
* {{Citation | last=Minsky | first=Marvin | author-link=Marvin Minsky | year = 2006 | title = The Emotion Machine | publisher = Simon & Schusterl | publication-place=New York, NY | isbn=0-7432-7663-9 }}
* {{Citation | last=Moravec | first=Hans | year = 1976 | url= http://www.frc.ri.cmu.edu/users/hpm/project.archive/general.articles/1975/Raw.Power.html | title = The Role of Raw Power in Intelligence | author-link=Hans Moravec }}
* {{Citation | last=Moravec | first=Hans | year = 1988 | title = Mind Children | publisher = Harvard University Press | author-link =Hans Moravec }}
* {{Citation | last=NRC |chapter=Developments in Artificial Intelligence|chapter-url=http://www.nap.edu/readingroom/books/far/ch9.html|title=Funding a Revolution: Government Support for Computing Research|publisher=National Academy Press|year=1999| author-link=United States National Research Council | access-date=30 August 2007}}
* {{Citation | last=Newell | first = Allen | last2 = Simon | first2=H. A. | year = 1963 | contribution=GPS: A Program that Simulates Human Thought| title=Computers and Thought | editor-last= Feigenbaum | editor-first= E.A. |editor2-last= Feldman |editor2-first= J. |publisher= McGraw-Hill|publisher-place= New York | author-link=Allen Newell|author2-link=Herbert Simon}}
* {{Citation | last=Searle | first = John | author-link=John Searle | year = 1980 | url = http://www.bbsonline.org/documents/a/00/00/04/84/bbs00000484-00/bbs.searle2.html | title = Minds, Brains and Programs | journal = Behavioral and Brain Sciences | volume = 3| issue = 3| pages= 417-457}}
* {{Citation | last=Shapiro| first= Stuart C. | year=1992 | url=http://www.cse.buffalo.edu/~shapiro/Papers/ai.ps | contribution =Artificial Intelligence | editor-first=Stuart C. | editor-last=Shapiro | title=Encyclopedia of Artificial Intelligence |edition=2nd |pages=54-57| location=New York |publisher= John Wiley}}.
* {{Citation | last=Simon | first = H. A. | year = 1965 | title=The Shape of Automation for Men and Management | publisher =Harper & Row | publication-place = New York | author-link=Herbert Simon}}
* {{Citation | last=Turing | first = Alan | year=1950 | title = [[Computing machinery and intelligence]] | journal=Mind | issn=0026-4423 | volume = LIX | issue = 236 |date=October 1950 | pages= 433-460 | url =http://loebner.net/Prizef/TuringArticle.html | authorlink = Alan Turing | doi=10.1093/mind/LIX.236.433}}
* {{Citation | last=Wason | first=P. C. | author-link=Peter Cathcart Wason | coauthors=Shapiro, D. | editor=Foss, B. M. | title=New horizons in psychology | year=1966 | location=Harmondsworth | publisher=Penguin | chapter=Reasoning }}
* {{Citation | last=Weizenbaum | first = Joseph | title = Computer Power and Human Reason | publisher = W.H. Freeman & Company | location = San Francisco | year = 1976 | authorlink=Joseph Weizenbaum | isbn = 0716704641}}

== Further reading ==
* R. Sun & L. Bookman, (eds.), ''Computational Architectures: Integrating Neural and Symbolic Processes''. Kluwer Academic Publishers, Needham, MA. 1994.

==External links==
{{linkfarm}}
{{sisterlinks|Artificial Intelligence}}
*{{dmoz|Computers/Artificial_Intelligence/|AI}}
*[http://www.learnartificialneuralnetworks.com AI with Neural Networks]
*[http://www.ai-tools.org AI-Tools, the Open Source AI community homepage]
*[http://www.ai-directory.com Artificial Intelligence Directory, a directory of Web resources related to artificial intelligence]
*[http://www.aaai.org/AITopics/html/welcome.html The Association for the Advancement of Artificial Intelligence]
*[http://www.vega.org.uk/video/programme/16 Freeview Video 'Machines with Minds' by the Vega Science Trust and the BBC/OU]
*[http://www.geocities.com/francorbusetti Heuristics and artificial intelligence in finance and investment]
*[http://www-formal.stanford.edu/jmc/whatisai/whatisai.html John McCarthy's frequently asked questions about AI]
*[http://www.aiai.ed.ac.uk/events/jonathanedwards2007/bbc-r4-jonathan-edwards-2007-03-28.mp3 Jonathan Edwards looks at AI (BBC audio)]
*[http://csdir.org/Artificial_Intelligence/ Artificial Intelligence in the Computer science directory]
*[http://www.generation5.org/ Generation5 - Large artificial intelligence portal with articles and news].
*[http://www.mindmakers.org Mindmakers.org, an online organization for people building large scale A.I. systems]
*[http://www.kurzweilai.net/ Ray Kurzweil's website dedicated to AI including prediction of future development in AI]
*[http://www.acceleratingfuture.com/michael/blog/?cat=15 AI articles on the Accelerating Future blog]
*[http://aigp.csres.utexas.edu/~aigp/ AI Genealogy Project]
*[http://www.ailibrary.net/ Artificial intelligence library and other useful links]
*[http://www.waset.org/ijci/ International Journal of Computational Intelligence]
*[http://www.waset.org/ijis/ International Journal of Intelligent Technology]
*[http://dictionary.laborlawtalk.com/Artificial_intelligence AI definitions at Labor Law Talk]
*[http://www.virtualhumansforum.com Virtual Humans Forum and Directory]

{{clear}}
{{Technology}}

[[Category:Artificial intelligence| ]]
[[Category:Intelligence by type]]
[[Category:Cybernetics]]

[[ar:ذكاء اصطناعي]]
[[bn:কৃত্রিম বুদ্ধিমত্তা]]
[[zh-min-nan:Jîn-kang tì-hūi]]
[[bs:Vještačka inteligencija]]
[[bg:Изкуствен интелект]]
[[ca:Intel·ligència artificial]]
[[cs:Umělá inteligence]]
[[da:Kunstig intelligens]]
[[de:Künstliche Intelligenz]]
[[et:Tehisintellekt]]
[[el:Τεχνητή νοημοσύνη]]
[[es:Inteligencia artificial]]
[[eo:Artefarita inteligenteco]]
[[eu:Adimen artifizial]]
[[fa:هوش مصنوعی]]
[[fr:Intelligence artificielle]]
[[gl:Intelixencia artificial]]
[[ko:인공지능]]
[[hi:आर्टिफिशियल इंटेलिजेंस]]
[[hr:Umjetna inteligencija]]
[[io:Artifical inteligenteso]]
[[id:Kecerdasan buatan]]
[[ia:Intelligentia artificial]]
[[is:Gervigreind]]
[[it:Intelligenza artificiale]]
[[he:בינה מלאכותית]]
[[lv:Mākslīgais intelekts]]
[[lt:Dirbtinis intelektas]]
[[jbo:rutni menli]]
[[hu:Mesterséges intelligencia]]
[[mr:कृत्रिम बुद्धिमत्ता]]
[[nl:Kunstmatige intelligentie]]
[[ja:人工知能]]
[[no:Kunstig intelligens]]
[[nn:Kunstig intelligens]]
[[pl:Sztuczna inteligencja]]
[[pt:Inteligência artificial]]
[[ksh:Artificial Intelligence]]
[[ro:Inteligenţă artificială]]
[[ru:Искусственный интеллект]]
[[simple:Artificial intelligence]]
[[sk:Umelá inteligencia]]
[[sl:Umetna inteligenca]]
[[sr:Вјештачка интелигенција]]
[[sh:Umjetna inteligencija]]
[[fi:Tekoäly]]
[[sv:Artificiell intelligens]]
[[th:ปัญญาประดิษฐ์]]
[[vi:Trí tuệ nhân tạo]]
[[tr:Yapay zekâ]]
[[tk:Ýasama akyl]]
[[uk:Штучний інтелект]]
[[zh:人工智能]]