Home | Site Map |

Artificial intelligence


This article is about modelling human thought with computers. For other uses of the term AI, see Ai .

Artificial intelligence, also known as machine intelligence, is defined as intelligence exhibited by anything manufactured (i.e. artificial ) by humans or othersentient beings or systems (should such things ever exist on Earth or elsewhere ). It is usually hypothetically applied togeneral-purpose computers . The term is also used to refer to the field ofscientific investigation into the plausibility of and approaches to creating such systems.



The question of what artificial intelligence is can be reduced to two parts: "what is the nature of artifice" and "what isintelligence"? The first question is fairly easy to answer, though it does point to the question of what it is possible tomanufacture (within the constraints of certain types of system, e.g. classical computational systems, of available processes ofmanufacturing and of possible limits on human intellect, for instance).

The second is much harder, raising questions of consciousness and self , mind (including the unconscious mind ) and the question of what components are involved in theonly type of intelligence it is universally agreed wehave available to study: that of human beings. Intelligent behavior in humans is complex and difficult to study or understand.Study of animals and artificial systems that are not just models of what exists already are also considered widely pertinent.

Several distinct types of artificial intelligence have been elucidated below. Also, the subject divisions, history, proponentsand opponents and applications of research in the subject are described. Finally, references to fictional and non-fictionaldescriptions of AI are provided.

Strong AI and weak AI

One popular and early definition of artificial intelligence research, put forth by John McCarthy at the DartmouthConference in 1956 , is "making a machine behave in ways that would be calledintelligent if a human were so behaving." However this definition seems to ignore the possibility of strong AI (see below).Another definition of artificial intelligence is intelligence arising from an artificial device. Most definitions could be categorized asconcerning either systems that think like humans, systems that act like humans, systems that thinkrationally or systems that act rationally.

Strong artificial intelligence

Strong artificial intelligence research deals with the creation of some form of computer-based artificialintelligence that can truly reason and solve problems ; a strong form of AI is said to be sentient ,or self-aware. In theory, there are two types of strong AI:

  • Human-like AI, in which the computer program thinks and reasons much like a human mind .
  • Non-human-like AI, in which the computer program develops a totally non-human sentience, and a non-human way of thinking andreasoning.

Weak artificial intelligence

Weak artificial intelligence research deals with the creation of some form of computer-based artificialintelligence that cannot truly reason and solve problems; such a machine would, in some ways, act as if it wereintelligent, but it would not possess true intelligence or sentience.

There are several fields of weak AI, one of which is naturallanguage . Many weak AI fields have specialised software or programming languages created for them. For example, the'most-human' natural language chatterbot A.L.I.C.E. uses a programming language AIML that is specific to itsprogram.

To date, much of the work in this field has been done with computer simulationsof intelligence based on predefined sets of rules. Very little progress has been made in strong AI. Depending on how one definesone's goals, a moderate amount of progress has been made in weak AI.

Philosophical criticism and support of strong AI

Several philosophers, notably John Searle and Hubert Dreyfus , have argued on philosophical grounds against the feasibility ofbuilding human-like consciousness or intelligence in a disembodied machine. Searle is most known for his Chinese room argument, which claims to demonstrate that even amachine that passed the Turing test would not necessarily be conscious in thehuman sense. Dreyfus, in his book What Computers Still Can't Do: A Critique of Artificial Reason, has argued thatconsciousness cannot be captured by rule- or logic -based systems or by systems that arenot attached to a physical body, but leaves open the possibility that a robotic system using neural networks or similar mechanisms might achieve artificial intelligence.

Other philosophers hold opposing views. Many see no problem with Weak AI but there is much support for Strong AI too. Daniel C. Dennett argues in Consciousness Explained that if there is no magic spark or soul, then Man is just amachine, and he asks why the Man-machine should have a privileged position over all other possible machines when it comes tointelligence.

Some philosophers hold that if Weak AI is accepted as possible then so must Strong AI. The Weak AI position, that intelligencemight be apparent but would not be real, is debunked in many ways, but one accessible example can be found in Simon Blackburn 's introduction to philosophy, Think. Blackburnpoints out that you might appear intelligent but there is no way of telling if that intelligence is real(whatever that means in this context): We have to take it on trust or faith.

Supporters of Strong AI claim that the anti-AI argument boils down in the end to some combination of

  1. arrogance as a privileged position is claimed, a magic spark is introduced (by God , for instance)
  2. defining intelligence as that of whichmachines are incapable.

An argument supporting Strong AI which those who deny its possibility mustnecessarily attack:

Given that
  1. the mind is a finite state machine (so the Church-Turing thesis applies to the brain)
  2. the mind is software (a finite state machine )
  3. the brain is purely hardware (i.e. only follows the rules of a classical computer)
  4. it exists exclusively within the brain
The possibility of Strong AI must be accepted.

Some (including Roger Penrose ) attack the applicability of theChurch-Turing thesis. Others say the mind is not completely physical. Roger Penrose's argument rests on the conception of hypercomputation being possible in our universe. Quantum mechanics and newtonian mechanics do not allow hypercomputation but it is thought that some strange space times would. However there seems to be agreement that our universe is notsufficiently convoluted to allow such hypercomputation.


Development of AI theory

Much of the (original) focus of artificial intelligence research draws from an experimental approach to psychology , and emphasizes what may be called linguistic intelligence (best exemplifiedin the Turing test ).

Approaches to artificial intelligence that do not focus on linguistic intelligence include robotics and collective intelligence approaches, which focus on active manipulation of an environment, or consensus decision making , and draw from biology and political science when seeking models of how "intelligent"behavior is organized.

Artificial intelligence theory also draws from animal studies, in particular withinsects, which are easier to emulate as robots (see artificial life ),as well as animals with more complex cognition, including apes , who resemble humans in manyways but have less developed capacities for planning and cognition. AI researchers argue that animals, which are simpler thanhumans, ought to be considerably easier to mimic. But satisfactory computational models for animal intelligence are notavailable.

Seminal papers advancing the concept of machine intelligence include A Logical Calculus of the Ideas Immanent in NervousActivity ( 1943 ), by WarrenMcCulloch and Walter Pitts , and On Computing Machinery and Intelligence ( 1950 ), by Alan Turing , and Man-ComputerSymbiosis by J.C.R. Licklider. See cybernetics and Turing test for further discussion.

There were also early papers which denied the possibility of machine intelligence on logical or philosophical grounds such as Minds, Machines and Gödel ( 1961 ) by John Lucas [1] .

With the development of practical techniques based on AI research, advocates of AI have argued that opponents of AI haverepeatedly changed their position on tasks such as computer chess or speech recognition that were previously regarded as"intelligent" in order to deny the accomplishments of AI. They point out that this moving of the goalposts effectively defines"intelligence" as "whatever humans can do that machines cannot".

John von Neumann (quoted by E.T. Jaynes ) anticipated this in 1948 by saying, in response to a commentat a lecture that it was impossible for a machine to think: "You insist that there is something a machine cannot do. If you willtell me precisely what it is that a machine cannot do, then I can always make a machine which will do just that!". VonNeumann was presumably alluding to the Church-Turing thesis which states that any effective procedure can be simulated by a (generalized) computer.

In 1969 McCarthy and Hayes started the discussion about the frame problem with their essay, "Some Philosophical Problems from the Standpointof Artificial Intelligence".

Experimental AI research

Artificial intelligence began as an experimental field in the 1950s with such pioneers as Allen Newell and Herbert Simon , who founded the firstartificial intelligence laboratory at Carnegie-Mellon University , and McCarthy and Marvin Minsky , who founded the MIT AI Lab in 1959. They allattended the aforementioned Dartmouth College summer AIconference in 1956 , which was organized by McCarthy, Minsky, Nathan Rochester of IBM and Claude Shannon .

Historically, there are two broad styles of AI research - the "neats" and "scruffies". "Neat", classical or symbolic AI research, in general, involves symbolic manipulation ofabstract concepts, and is the methodology used in most expert systems. Parallel to this are the "scruffy", or "connectionist",approaches, of which neural networks are the best-known example, whichtry to "evolve" intelligence through building systems and then improving them through some automatic process rather thansystematically designing something to complete the task. Both approaches appeared very early in AI history. Throughout the 1960sand 1970s scruffy approaches were pushed to the background, but interest was regained in the 1980s when the limitations of the"neat" approaches of the time became clearer. However, it has become clear that contemporary methods using both broadapproaches have severe limitations.

Artificial intelligence research was very heavily funded in the 1980s by the Defense Advanced ResearchProjects Agency in the United States and by the Fifth Generation Computer project in Japan . The failure of the work funded at the time to produce immediate results, despite thegrandiose promises of some AI practitioners, led to correspondingly large cutbacks in funding by government agencies in the late1980s, leading to a general downturn in activity in the field known as AI winter . Over the following decade, many AI researchers moved into related areas with moremodest goals such as machine learning , robotics , and computer vision , though researchin pure AI continued at reduced levels.

Practical applications of AI techniques

Whilst progress towards the ultimate goal of human-like intelligence has been slow, many spinoffs have come in the process.Notable examples include the languages LISP and Prolog , which were invented for AI research but are now used for non-AI tasks. Hacker culture first sprang from AI laboratories, in particular the MIT AI Lab , home at various times to such luminaries as McCarthy, Minsky, Seymour Papert (who developed Logo there), Terry Winograd (whoabandoned AI after developing SHRDLU ).

Many other useful systems have been built using technologies that at least once were active areas of AI research. Someexamples include:

The vision of artificial intelligence replacing human professional judgment has arisen many times in the history of the field,in science fiction and today in some specialized areas where" expert systems " are used to augment or to replace professional judgmentin some areas of engineering and of medicine.

Hypothetical consequences of AI

Some observers foresee the development of systems that are far more intelligent and complex than anything currently known. Onename for these hypothetical systems is artilects. With the introduction of artificially intelligentnon-deterministic systems, many ethical issues will arise. Many of these issues havenever been encountered by humanity .

Over time, debates have tended to focus less and less on "possibility" and more on "desirability", as emphasized in the" Cosmist " (versus " Terran ") debatesinitiated by Hugo de Garis and Kevin Warwick . A Cosmist, according to de Garis, is actually seeking to build more intelligent successors tothe human species. The emergence of this debate suggests that desirability questions may also have influenced some of the earlythinkers "against".

Some issues that bring up interesting ethical questions are:

  • Determining the sentience of a system we create.
  • Can AI be defined in a graded sense?
  • Freedoms and rights for these systems
  • Can AIs be "smarter" than humans in the same way that we are "smarter" than other animals?
  • Designing systems that are far more intelligent than any one human
  • Deciding how many safe-guards to design into these systems
  • Seeing how much learning capability a system needs to replicate human thought, or how well it could do tasks without it (e.g. expert systems )
  • The Singularity
  • Effect on careers and jobs. The problems may resemble problems seen under freetrade .

Famous figures

Machines displaying some degree of "intelligence"

There are many examples of programs displaying some degree of intelligence. Some of these are:

  • The StartProject - a web-based system which answersquestions on English.
  • Cyc , a knowledge base with vast collection of facts about the real world and logicalreasoning ability.
  • ALICE , a chatterbot
  • Alan , another chatterbot
  • ELIZA , a program which pretends to be a psychotherapist, developed circa 1970.
  • PAM (Plan Applier Mechanism) - a story understanding system developed by John Wilensky in 1978.
  • SAM (Script applier mechanism) - a story understanding system, developed in 1975.
  • SHRDLU - an early natural language understanding computer program developed in1968-1970.
  • Creatures , a computer game with breeding, evolving creatures coded from thegenetic level upwards using a sophisticated biochemistry and neural network brains.
  • BBC news story on the creator of Creatures latest creation. Steve Grand 's Lucy.
  • Eurisko - a language for solving problems which consists of heuristics, includingheuristics describing how to use and change its heuristics. Developed in 1978 by Douglas Lenat.
  • X-Ray Vision for Surgeons - a group in MIT which researches medicalvision.
  • Neural networks-based programs forbackgammon and go .

AI researchers

There are many thousands of AI researchers around the world at hundreds of research institutions and companies. Among the manywho have made significant contributions are:

To some computer scientists, the phrase artificial intelligence has acquired somewhat of a bad name due to the largediscrepancy between what has been achieved so far in the field and some more usual notions of intelligence. This problem has beenaggravated by various popular science writers and media personalities such as Kevin Warwick whose work has raised the expectations of AI research far beyond its current capabilities. Forthis reason, some researchers working on topics related to artificial intelligence say they work in cognitive science , informatics , statistical inference or information engineering . However, progress has infact been made, and AI is today routinely employed in thousands of industrial systems around the world. See Raj Reddy 's AAAI paper for a huge review of real-world AI systems in deploymenttoday.


Further reading


  • ArtificialIntelligence: A Modern Approach by Stuart J. Russell and Peter Norvig
  • Gödel, Escher, Bach  : An Eternal GoldenBraid by Douglas R. Hofstadter
  • Shadows of theMind and The Emperor's New Mind by Roger Penrose
  • Consciousness Explained by Dennett.
  • The Age of SpiritualMachines by Ray Kurzweil
  • Understanding Understanding: Essays on Cybernetics and Cognition by Heinz von Foerster
  • In the Image of the Brain: Breaking the Barrier Between Human Mind and Intelligent Machines by Jim Jubak
  • Today's Computers, Intelligent Machines and Our Future byHans Moravec, Stanford University


The following is a list of influential works See also longer lists at:-

AI related organizations


  • John McCarthy: Proposal for the Dartmouth Summer Research Project On Artificial Intelligence. [2]

See also

Important publications in artificial intelligence .

Sub-fields of AI research

Logic programming was sometimes considered a field ofartificial intelligence, but this is no longer the case.






External links

rtificial intelligence, systems, artificial intelligece, machine, artiifcial intelligence, computer, artificial intellignece, many, artificial intelligenc, weak, artificial intelligence, mind, artificial ntelligence, one, artfiicial intelligence, system, artificial intelligecne, non, artificial nitelligence, developed, artificial inteligence, consciousness, artificial intelliegnce, humans, , programming, artifiical intelligence, philosophical, artificial intellignce, neural, artificial intellience, much, artificial itelligence, problems, aritficial intelligence, logic, atrificial intelligence, applications, artifciial intelligence, minsky, artifcial intelligence, world, artificial intellgience, vision, artificial intelilgence, cannot, ratificial intelligence, history, artificial intellgence, known, artificial intleligence, progress, artifical intelligence, real, artificail intelligence, question, artifiial intelligence, position, artificial inetlligence, recognition, artificial intlligence, thought, artificiali ntelligence, man, artificila intelligence, animals, artficial intelligence, areas, artificialintelligence, famous, artificial inelligence, related, artificial itnelligence, theory, arificial intelligence, early, artificil intelligence, test, artificial intelligene, fields, artificia lintelligence, today, artificial intelligenec, marvin, artificia intelligence, philosophers, atificial intelligence, models, artiicial intelligence, study...

This article is completely or partly from Wikipedia - The Free Online Encyclopedia. Original Article. The text on this site is made available under the terms of the GNU Free Documentation Licence. We take no responsibility for the content, accuracy and use of this article.

Anoca.org Encyclopedia