Written by Phillip Berrie
In the next series of articles I am going to look at artificial intelligence (AI) both in the real world and speculative fiction. We shall start with some history ...
The concepts of artificial life and intelligence have been with us for millennia. Myths and stories concerning artificial or constructed beings with human-like intelligence abound ranging from the automatons of Hephaestus the Greek god of black smiths through to Lieutenant Commander Data of Star Trek fame. However, this series of articles is going to concentrate on the reality of AI and the chance of it becoming real in the Computer Age.
Artificial Fiction Brain by Gengiskanhg.
Sourced from the Wikimedia Commons through a Creative Commons Sharealike 3.0 licence.
In the 1950s and 60s Computer Scientists were convinced that they would have computers with artificial intelligence in the very near future. So positive of this were they that Alan Turing, often considered the father of computing science, went so far as to define a test to determine whether a computer program could be considered intelligence or not.
The Turing Test is as follows: A human judge engages in a natural language conversation with one human and one machine, each of which tries to appear human. All participants are placed in isolated locations and communications is done via text to eliminate socially-based and other implicit communications delivered by the voice of the human participant. If the judge cannot reliably tell the machine from the human, the machine is said to have passed the test.
Note: There is an annual competition called the Loebner Prize based on the Turing Test which as of October 2008 has yet to have its computer-based participants miss-identified as humans more than 30% of the time — the winning condition for the US$100,000 prize.
The defining of the Turing Test brought about the creation of a number of Natural Language Processing programs. One of the earliest and best known of these programs was called 'Eliza' (named after the George Bernard Shaw character with the same name from the play Pygmalion) which manifested itself on mainframes and computer terminals in university Computer Science labs all over the English speaking world at the time. See the references below if you would like to talk to Eliza — she hasn't changed a bit.
After starting out so confidently, progress in AI stalled as researchers began to realise that the computing power they had at their disposal was not really up to the task. For, although their 'intelligent' programs were good at the individual tasks they were written for, they were seriously lacking in the more general intelligence and adaptability that we associate with humans.
"I'm sorry Dave, I'm afraid I can't do that."
HAL 9000, 2001 A Space Odyssey in 1968.
Artificial intelligence picked up again in the 1980s with the development of Expert Systems. An expert system is a computer-based system which attempts to capture the knowledge of an expert (or group of experts) in a particular field. An expert system can then be used by people without such expertise to solve or at least identify correctly problems in that field. Expert systems were developed that had the ability to learn and also infer new solutions from their knowledgebase. However, their inherent focus means they have limited general intelligence and although they are still used today for knowledge based applications they are not the key to the goal of general artificial intelligence.
In the new millennia, the topic of artificial intelligence is still strong in the Computer Science discipline. However, these days it is more used for solving specific problems in areas concerning: logistics, data mining, problem diagnosis and of course — chess.
Note: Since the year 2000, chess champions have been finding it hard going against their computerised opponents. And, in 2006, the undisputed human world chess champion, Vladimir Kramnik, was defeated 4 games to 2 by a chess program called 'Deep Fritz'.
Next article we shall analyse why the promise of AI has fallen so short of the mark and look at the prospects for the future.
References:
The Wikipedia on Alan Turing.
http://en.wikipedia.org/wiki/Alan_Turing
About the annual Turing Test competition, the Loebner Prize
http://www.loebner.net/Prizef/loebner-prize.html
And here is one journalist's account of last year's competition.
http://technology.timesonline.co.uk/tol/news/tech_and_web/article4934858.ece
A web page with a JAVA implementation of the original Eliza program
http://www.chayden.net/eliza/Eliza.html
The Wikipedia on Expert Systems
http://en.wikipedia.org/wiki/Expert_systems#Application_of_expert_systems
The Computer Chess Chronology
http://en.wikipedia.org/wiki/Computer_chess#Chronology_of_computer_chess
N.B. Please note that I although I use the Wikipedia (and WikiMedia Commons) a lot for references, this is for expediency and the familiarity of my readers. Anyone interested in further studies should make use of the references where available and understand that the Wikipedia is a co-operative project contributable to by anyone and must always be looked at in that light.
Did you know you can add your own reviews?
Just write at least 300 words and submit here!
TrackBack URI for this entry
Subscribe to this comment's feed