ARTIFICIAL INTELLIGENCE
Abstract
AI is a branch of science which deals with helping machines find solutions to computer computer problem problemss in a more human like fashion. we start by making a distin distincti ction on betwee between n mind mind and and cogn cogniti ition, on, and by positing cognition is an aspect aspect of mind .The term AI was was coined in 1956 by ohn !c"arthy at the !assachusetts Institute of Technology. Technology. The field was founded on the claim that a central property of humans, intelligence# the sapien sapience ce of $omo $omo sapie sapiens# ns#can can be so prec precise isely ly described that it can be simulated by a machine. This raises philosophical issues about the nature of the mind and limits of scientific hubris, issues which ha%e been addr addres esse sed d by myth myth,, fict fictio ion n and and phil philos osop ophy hy sinc sincee anti&uity. Artificial intelligence has been the sub'ect of optimism, but has also suffered setbacks and, today, has become become an essential essential part of the technology technology industry industry,, pro%iding the hea%y lifting for many of the most difficult problems problems in computer science, !echanical or (formal( reaso reasonin ning g has been been de%elo de%eloped ped by philos philosoph ophers ers and mathematicians mathematicians since anti&uity. AI generally in%ol%es borrowing characteristics from human intelligence, and applying them as algorithm in human friendly way. It is basically the ability of a machine to think for itself. It aims at getting getting computers computers to do tasks tasks which which re&uir re&uiree human human intell intellige igence nce.. In short short it can be descri described bed as simple things turn turn out to be the hardest hardest to automate like) like) *ecogni+ing a face, a%igating a busy street, street, -nderstanding what someone says. e tried to e/plain the brief ideas of AI and its application to %arious fields. It cleared the concept of computational and con%entional categories. It includes %ariou %ariouss ad%anc ad%anced ed system systemss such such as eural eural etwo etwork, rk, 0/pert ystems, 2u++y ystems and 0%olutionary computati computation. on. AI is used in typical typical problems problems such as 3attern recognition, recognition, atural language processing processing and more. This system is working throughout the world as an artificial brain. brain. This paper throws throws light on e%olution of ai in india along with pros and cons.
methods that are not obser%ed in people or that in%ol%e much more computing than people can do. e discussed conditions for considering a machine to be intelligent. e argued that if the machine could successfully pretend to be human human to a knowle knowledge dgeabl ablee obser% obser%er er then then you certainly should consider it intelligent.
1. Introduction It is the branch branch of Comput Computer er Scienc Sciencee conce concerne rned d with making computers behave like humans. It is the Science and and Engi Engine neer ering ing of maki making ng inte intelli llige gent nt mach machine ines, s, especially especially intelligent intelligent computer computer programs. programs. It is the hot topic on many boards and software houses. Although AI has a strong science fiction connotation, it forms a vital branch of computer science, dealing with intelligent behavior, learning and adaptation in machines. It combines with and contributes to several other disciplines, including: !sychology !hilosophy linguistics biology anthropology logic mathematics computer science " software engineering #esearch in AI is concerned with producing machines to automate tasks re$uiring intelligent behavior. E%amples include control, planning and scheduling, the ability to answer diagnostic and consumer $uestions, handwriting, speech, speech, and facial facial recognition. recognition. As such, it has become a scientific scientific discipline, discipline, focused focused on providing providing solutions to real life problems. AI syst system emss are are now now in rout routin inee use use in econ econom omic ics, s, medicine, engineering and the military, as well as being built into many common home computer software applications applications,, traditional traditional strategy strategy games like computer computer chess and other video games.
2. History
e can learn something about how to make machines sol%e problems by obser%ing other people or 'ust by obser%ing our own methods. 4n the other hand, most most work work in AI in%ol% in%ol%es es studyi studying ng the prob problem lemss the world world pres present entss to intellige intelligence nce rather rather than studying studying people or animals. AI researchers researchers are free to use
&
methods now classified as machine learning, characteri4ed by formalism and statistical analysis. 'his is also known as symbolic AI, logical AI, neat AI and (ood >ld -ashioned Artificial Intelligence 0(>-AI2. 0(>-AI2. !ethods include) include) E%pert systems: apply reasoning capabilities to reach a conclusion. An e%pert system can process large l arge amounts of known information and provide conclusions based on them. Case based reasoning =ayesian networks =ehavior based AI: a modular method of building AI systems by hand. 'he intellectual roots of AI, and the concept of intelligent machines, may be found in (reek mythology. Intelligent artifacts appear in literature since then, with real mechanical devices actually demonstrating behaviour with some degree of intelligence. After modern computers became available following )orld )orld )ar*II, )ar*II, it has become possible to create programs that perform difficult intellectual tasks. 195 196) 'he first working AI programs were written in &+& to run on the -erranti ark I machine of the /niversity of anchester 0/12: a draughts*playing program written by Christopher Strachey and a chess*playing chess*playing program written by 3ietrich !rin4. 1967198 ) 3uring the &+56s and &+76s arvin insky and Seymour !apert publish !erceptrons, demonstrating limits of simple neural nets and Alain Colmerauer developed the !rolog computer language. 'ed 'ed Shortliffe demonstrated the power of rule*based systems for knowledge representation and inference in medical diagnosis and therapy in what is sometimes called the first e%pert system. 8ans oravec developed the first computer*controlled vehicle to autonomously negotiate cluttered obstacle courses. 19:s onwards) In the &+96s, neural networks became widely used with the back propagation algorithm, first described by !aul ohn )erbos )erbos in &+7;. 'he &++6s marked ma
3.Categories of AI : AI divides roughly into two schools of thought: "on%entional AI) "omputational Intelligence ;"I<) "on%entional AI ) Conventional AI mostly involves
"omputational intelligence ;"I<) Computational Intelligence involves iterative development or learning 0e.g. parameter tuning in connectionist systems2. ?earning is based on empirical data and is associated with non*symbolic AI, scruffy AI and soft computing. !ethods include) include) eural networks: systems with very strong patt ern eural recognition capabilities. -u44y Systems: techni$ues for reasoning reasoning under uncertainity, has been widely used in modern industrial and consumer product control systems. Evolutionary computation: applies biologically inspired concepts such as population, mutation and survival of the fittest to generate increasingly better solution to the problem. 'hese methods most notably divide into evolutionary algorithms 0e.g. genetic algorithms2 and swarm intelligence 0e.g. ant algorithms2.
4. Mechanisms >ver the past five decades, AI research has mostly been focusing on solving specific problems. @umerous solutions have been devised and improved to do so efficiently and reliably. 'his e%plains why the field of Artificial Intelligence is split into many branches. Some of the branches have been e%plained below. 4.1 Planning !lanni !lanning ng progr programs ams start start with with gener general al facts facts about about the world 0especially facts about the effects of actions2, facts about the particular situation and a statement of a goal. -rom these, they generate a strategy for achieving the goal. In the most common cases, the strategy is
4.2 Pattern recognition: 'he main focus in AI today is getting a computer to recogni4e, make senses and recreate in what it sees and hears. 'he 'he two ma
3attern*ecognition= 3attern*ecognition=ision ision: ItBs goal is to get a computer to recogni4e pictures so that it can recogni4e ob
4.2 Ontology: >ntology is the study of what ob
4.4Artifcial 4.4Artifcial lie: From robot dreams dreams to reality: Artificial life is a field of scientific study that attempts to mode modell livi living ng biol biolog ogic ical al syste systems ms thro throug ugh h comp comple le% % algo algori rith thms ms.. Scie Scient ntis ists ts use use thes thesee mode models ls to test test and and e%periment e%periment with a multitude multitude of factors factors on the behaviour behaviour of the systems. It is a diverse field of research, but a common theme is testing out the fundamental principles of life by building detail detailed ed workin working g mod models els.. >ne of the mos mostt ambitio ambitious us goals goals of artific artificial* ial*lif lifee resear research ch is the constr construct uction ion of living systems out of non*living parts. Artificial life is a blanket term used to refer to human attempts at setting up systems systems with lifelike lifelike properties properties all biological biological organisms organisms possess, such as self*reproduction, homeostasis, adap adapta tabi bilit lity y, mutat mutatio iona nall vari variat atio ion, n, optim optimi4 i4at ation ion of e%ternal states, and so on. 4.! E"ist#$ology Epistemology is a study of knowledge that are re$uired for solving problems in the world.
!. Research challenges: Artificial Intelligence and robotics are likely to creep into our lives without us really noticing. 8owever, AI has spawned some useful applications like e%pert systems and game AI, but the truly pervasive use of AI is still to come as more research and improved technology surfaces in the future. 8ere are a few applied innovations that AI promises in the future and the technologies behind them.
#obotics is the study of how to design, build, use, and work with robots. #obots are mechanical devices that can move and react to sensory input giving them some degree of autonomous control. #obo #obots ts are are wide widely ly used used in the the indu indust stri rial al sect sector or performing high*precision
>A*3A >A*3A ?rand "hallenge) "hallenge) 'he 3A#!A (rand Challenge was a race for a D million pri4e where cars had to drive drive themselves themselves over a hundred miles of challenging desert terrain without any
communication with humans, using using GPS, GPS, computers and a sophisticated array of sensors. 'his was the first in a series of challenges aimed aimed at a congressional mandate stating stati ng that by 6& one*third of the operational ground ground combat vehicles of the /S Armed -orces should be unmanned. Telephone Translators)
>ne of the common cliches when one talks about the future is how the world is shrinking every day. 3istance used to be a barrier in travel and the invention of the airplane changed all that. 'ime used to be a factor in commun communica ication tion since the mail syste system m took months months to deliver a letter across the /nited States, but the telephone dissolved such a hurdle. 'he combinations of travel and commun communica ication tionss has brough broughtt whole whole nation nationss togeth together er
e%cept now the last barrier in international relationship is langua language. ge. 'his 'his is where where teleph telephone one transl translato ators rs will change all that. Essentially, a person from the /nited States says some thin thing gs in Engli nglissh into into his his tele teleph phon onee. Almo lmost inst instan anta tane neou ously sly,, a compu compute terr inte interc rcep epts ts the the voic voice, e, translates translates what was said, and synthetically synthetically generate the appropriate apanese words to the person on the other line. line. >f Course Course,, the transl translato atorr would would need need advanc advanced ed voice voice recog recogniti nition, on, natura naturall langua language ge proces processin sing g and inferencing to e%tract what was meant by the English* speaker, and then synthesi4e a human*sounding apanese personBs voice in conversational apanese *esearch *esearch Assistants) Assistants) 'he world is mov moving ing from the Indust Industria riall Age Age to the Information Age where the phrase knowledge is power is becoming a reality. )ith so much information out there, it has become harder and harder to find what is real really ly rele releva vant. nt. 'his 'his is wher wheree a rese resear arch ch assis assista tant nt powered by AI can help. @ot only can the assistant unders und erstan tand d what what one is lookin looking g for, for, which which re$uir re$uires es natural language processing, it is smart enough to know where to look and compare what it finds to what it is looking for to see how relevant the information is, so the person doesnBt have to do the Bdirty work.B #esearch assist assistant antss will be an import important ant tool in the future future by keeping the world of information from e%ploding into an infinite chaos of unorgani4ed facts and figures. A ?reater ?reater -se of 0/pert ystems) ystems) )ith ith such such succ succes esss as a diag diagno nost stic ic in medi medicc and and mech mechan anic icss pres presen ently tly,, e%pe e%pert rt syst system emss will will be more more prevalent in other applications that re$uire an e%pert with whom people can consult with. @eed to identify the perfect pet for a friendF A pet e%pert system could ask some $uestions related to the personBs personality so that it can conclude the types of animals that would be suited for them. )hat kinds of dishes can one make tonight with the food in the refrigeratorF Input the foods into a cook e%pert system and find out. 'he possibilities for e%pert systems are almost endless. If e%pert systems are design designed ed and built built correc correctly tly,, users users sho should uld be able able to easily program their own e%pert and should make better decisions in their lives.
3assing the Turing Turing Test) Test) 'he idea behind the test is that if a machine could make a person think heGshe was interacting with an intelligent person, why not consider the machine intelligent in its own rightF 'he controversy over the 'uring 'est will
probably continue into the future, but once a computer convincingly passes the test and becomes more and more integrated with society, this test would be at least the best appro%imation of intelligence possible.
%. A""lications Artificial Artificial Intelligence Intelligence is helping helping people in every every field to make make bette betterr use use of info inform rmat atio ion n to work work hard harder er not not sm smar arte terr. 'he 'he pote potent ntia iall appl applic icat atio ions ns of Artif rtific icia iall Inte Intelli llige genc ncee are are abun abunda dant. nt. 8owe 8oweve ver, r, some some of the the applications of AI have been listed below:
!edicine) ew ew blood test spots spots cancer In one of the biggest biggest advanc advances es in cancer cancer researc research h in years years,, scient scientists ists have develo developed ped a blood blood test test that that can detect detect cancer cancer with a greater than +6H accuracy accuracy.. 'his artificial intelligence **already tested for cancers of the breast, ovary, and lung**could one day be used to detect many types cancer. BAll thatBs needed is a single drop of bloodJ B'he computer does the rest.B...In tests on several hundred blood samples, some taken from women with ovarian cancer and others from healthy women, the test proved Ban astonishingB &66H accurate in detecting cancer, even at the earliest stages. Artificial nose) Scient Scientists ists have have endowe endowed d comput computers ers with eyes eyes to see, see, than thanks ks to digi digita tall came camera ras, s, and and ears ears to hear hear,, via via microp microphon hones es and sop sophis histica ticated ted recogn recognitio ition n softwa software re.. @ow theyBre taking computers further into the realm of the senses with the development of an artificial nose. 0nose to sniff sniff out hospital hospital superbugs) E*nose analyses gas samples by passing the gas over an array array of electr electrode odess coated coated with differ different ent conduc conducting ting polymers. Each electrode reacts to particular substance by changing its electrical resistance in a characteristic way. Combining the signals from all the electrodes gives a Bsmell*printB of the chemicals in the mi%ture that neural netwo network rk soft softwa ware re built built into into the the e*no e*nose se can can lear learn n to recogni4e. As a result, it can be detected from the smell alone that what the bacterial infections are.
!ilitary) A new model of army soldier rolls closer to the battlefield) 'he American military is working on a new generation of soldier, far different from the army it has. B'hey donBt feel hung hungry ry,B ,B said said (ord (ordon on ohn ohnso son n of the the oin ointt -orc -orces es
;
Command at the !entagon. B'hey are not afraid. 'hey donBt forget their orders. 'hey donBt care if the guy ne%t to them has
of tiny, subconscious rules*truths we that have learned from e%perience. Add them up and you get instinct: a doctorBs sense that a patientBs stomach*ache might really be appendicitis, for e%ample. !rogram those rules into a computer and you get an e%pert system* one of many that can screen screen lab tests, tests, diagno diagnose se blood blood infect infections ions,, and identify tumors on a mammogram.
&. Pros and Cons 3ros) Artif Artific icia iall inte intelli llige genc ncee find findss appli applica catio tions ns in spac spacee e%plor e%plorati ation. on. Intell Intellige igent nt robots robots can be used used to e%plor e%ploree space. 'hey are machines and hence have the ability to endure endure the hos hostile tile enviro environme nment nt of the interpl interplane anetar tary y space. 'hey can be made to adapt in such a way that planetary atmospheres do not affect their physical state and functioning.
Intelligent robots can be programmed to reach the EarthBs nadirs. 'hey can be used to dig for fuels. 'hey can be used for mining purposes. 'he intelligence of machines can be harnes harnessed sed for e%ploring e%ploring the depths depths of oceans oceans.. 'hese machines serve human so well especially where human intelligence has serious limitations. Intelligent machines can replace human beings in many areas of work. #obots can do certain l aborious tasks.
0/pert ystems) ystems) 'he primary goal of e%pert systems research is to make e%pertise available to decision makers and technicians who need need answer answerss $uickl $uickly y. 'here 'here is never never enough enough e%pe e%pert rtise ise to go arou around nd** **ce cert rtai ainly nly it is not not alwa always ys available at the right place and the right time. !ortable with comput computer erss loaded loaded with with in*dept in*depth h kno knowle wledge dge of specific sub
!ains taking activities, which have long been carried out by humans can be taken over by the robots. >wing to the intelli intelligen gence ce progr programm ammed ed in them, them, the machin machines es can shoulder responsibility to a certain e%tent. 'hey can be made to manage themselves and their time to complete the assigned tasks. Emotionss that Emotion that often often interce intercept pt ration rational al thinki thinking ng of a human being are not a hindrance for artificial thinkers. ?acking the emotional side, robots can think logically and take the right decisions. Sentiments are associated with moods that affect human efficiency. 'his is not the case with machines with artificial intelligence. 'hus 'hus artif artific icia iall inte intelli llige genc ncee can can be utili4 utili4ed ed in the the comple completion tion of repetit repetitive ive and time*co time*consu nsumin ming g tasks tasks efficiently. Intelligent machines can be employed to do cert certai ain n dang danger erou ouss task tasks. s. ach achin ines es e$ui e$uipp pped ed with with artificial intelligence can be made to thoughtfully plan towards the fulfillment of tasks tasks and accordingly ad
'he idea of machines replacing human beings sounds wonderful. It appears to save us from all the pain. =ut is it real really ly such such an e%ci e%citin ting g ideaF ideaF Conc Concep epts ts such such as whol wholeh ehea eart rted edne ness ss and and dedic dedicat atio ion n in work work bear bear no e%istence in the world of artificial intelligence. Imagine Imagine robots robots workin working g in hos hospita pitals. ls. 3o you pictur picturee them showing care and concern towards the patientsF Imagine Imagine intellig intelligent ent machin machines es employ employed ed in creativ creativee fields. 3o you think the robots will e%cel in such fieldsF 'hinking machines lack a creative mind. 8uman beings are emotional intellectuals. 'hey think and feel. 'heir feelings guide their thoughts. If robots begin replacing humans in every field, it may lead to unemployment. !eople will be left with nothing to do. Empty time may result result in its destruct destructive ive use. 'hin 'hinki king ng mach machin ines es will will gove govern rn all all the the fiel fields ds and and populate all positions pre*occupied by people.
Attributes
Natural Artificial Intelligence Intelligence (Human) (Machine)
The ability to use sensors (eyes, ears, touch, smell)
HIH
!"#
The ability to be creati$e an% imaginati$e
HIH
!"#
The ability to learn from e&'erience
HIH
!"#
The ability to be a%a'ti$e
HIH
!"#
The ability to affor% the cost of acuiring intelligence The ability to use a $ariety of information source
HIH
!"#
HIH
HIH
The ability to acuire large HIH amount of e&ternal information !"# The ability to mae com'le& calculations
HIH
The ability to transfer information
!"#
HIH
HIH
HIH The ability to mae a series !"# of calculations ra'i%ly an% accurately Apart Apart from from all these issues, issues, there there is a fear fear of robots superseding usL Ideally human beings should continue being the masters of machines. If things turn the other way round, the world will turn into chaos. Intelligent machines may prove to be smarter than usM they might enslav enslavee us and start ruling ruling the world. world. anBs anBs greedy greedy creativity may endanger mankindL
Eventually, it is up to you whether to stand by artificial intelligence or understand the likely disaster that it may lead to. In my view, there is no ideal replacement for human beings. Artificial intelligence can help alleviate the difficulties faced by man but intelligent machines can never be BhumanB
%.1 Li$itations If robots start replacing human resources in every field, we will ill have to deal with seriou ious issues like ike unemplo unemployme yment nt in turn turn leading leading to mental mental depres depressio sion, n,
5
poverty and crime in the society. 8uman beings deprived of their work life may not find any means to 8uman beings will be left with empty time. Secondly, replacing human beings with robots in every field may not be a right decision to make. 'here are many many ne of the ma
(.)ibliogra"*y P&Q!rograms with Common Sense :* ohn cCarthy, In !echani+ation of Thought 3rocesses, 3rocesses, 3roceedings 3roceedings of the ymposium ymposium of the ational ational 3hysics 3hysics @aboratory, &++. PQArtificial Intelligence, ?ogic and -ormali4ing Common Sense :* #ichmond 'homason, editor, 3hilosophical @ogic and Artificial Intelligence. 1lRver Academic, &+9+. PQConcepts of ?ogical AI :* 'om 'om itchell. !achine @earning . c(raw*8ill, &++7. P;Q?ogic and artificial intelligence :*#ichmond 'homason. PQIn Edward @. alta, editor, The tanford 0ncyclopedia of 3hilosophy. -all 66.
&. Goals 'he goal of artificial intelligence 0at least according to the field fieldss founde founders2 rs2 is to create create comput computers ers whose whose intelligence e$uals or surpasses humans. Achieving this goal is the famous NAI problem.O 'o some, AI is the manifest destiny of computer science. 'o others, its a failure: clearly, the AI problem is nowhere near being solved. )hyF -or the most part, the answer is simple: no one is really trying to solve it. 'his may come as a surprise to people outside the field. )hat have all those AI researchers been doing all these yearsF 'he reality is that they have largely given up on the grand ambitions of AI and are instead working on increasingly speciali4ed subpro sub proble blems: ms: not
'.Conclusion >ver >ver the past past few decade decades, s, despite despite many stumbl stumbling ing blocks, AI has grown from a do4en researchers, to thou thousa sand ndss of engi engine neer erss and and spec specia ialis listsM tsM and and from from programs capable of playing checkers, to systems designed designed to diagnose diagnose disease. disease. As we progress progress in the development of artificial intelligence,other theories are available,in addition to blocks, AI has grown from a do4e do4en n rese resear arch cher ers, s, to thous thousan ands ds of engi engine neer erss and and specialists specialistsMM and from programs programs capable capable of playing playing checkers, to systems designed to diagnose disease. As we prog progre ress ss in the deve develop lopme ment nt of artif artific icia iall intelligence intelligence,other ,other theories theories are available,in available,in addition addition to building on what we can do with AI.
7
