In 1950, British decoder and war-hero, Alan Turing issued a challenge to all the computer programmers of the world. He challenged them to create software that could converse with humans in Natural Language in such a manner that it would be indistinguishable from a human. The premise of the test was that a panel of human judges would converse anonymously, in text form, with a computer program and also with a human operating a computer and try to determine which of the participants is the human and which is the program. The Turing Test, as it came to be called, inspired computer scientists all over the world to make it their goal to build a Conversational Artificial Intelligence (Conversational AI) system that could beat the test. ELIZA, ALICE, Jabberwacky and all the other chatbots and conversational agents that we’ve come to know and experience were born because of the Turing Test. In 1991, the American philanthropist Hugh Loebner started the annual Loebner Prize competition, promising a $100,000 payout to the first computer to pass the Turing test. However, no AI program has even come close to passing an undiluted Turing test so they began awarding $2,000 each year to the best effort.
Alan Turing, born in 1912 in London, studied Mathematics at the University of Cambridge. He later moved to Princeton University in 1936 to study for a Ph.D. in mathematical logic which he completed in 1938.
In the summer of 1938, Alan Turing returned to the United Kingdom and joined the Government Code and Cypher School. When World War II broke out, he moved to the organization’s wartime headquarters at Bletchley Park, Buckinghamshire.
Over the autumn of 1939 and into the spring of 1940, Alan Turing in collaboration with some others designed a code-breaking machine known as the Bombe. This machine had the capability to break even the German Enigma code. For the rest of the war, the Allies received large quantities of military intelligence thanks to the Bombes.
Alan Turing was considered to be a war-hero and after the war ended he was made an Officer of the Most Excellent Order of the British Empire (OBE) for his code-breaking work.
Turing was considered to be one of the founding fathers of Artificial Intelligence and modern cognitive science. He was a leading early exponent of the theory that the human brain is in large part a digital computing machine. He theorized that the cortex is an “unorganized machine” at birth and that through “training” it gets organized “into a universal machine or something like it.”
In 1950, he put forth the Turing Test. This was essentially a challenge to computer scientists to create a computer program that could fool a panel of human judges into thinking that it was human.
Over the decades, there have been quite a large amount of claims by computer scientists about having passed the Turing Test, while only a few programs have even come vaguely close to passing this test. Here are some of the most noteworthy claims.
Eugene Goostman was a program developed by a trio of programmers. Created by Russian-born Vladimir Veselov, Sergey Ulasen, and Ukranian born Eugene Demchenko.
Eugene Goostman played the role of a 13-year-old Ukrainian boy. A brilliant idea, intended to induce forgiveness in those with whom it interacted for its grammatical errors and dearth of general knowledge.
The bot has participated in a large number of Turing test contests since its creation. It even managed to rank second in the 2005 and 2008 Loebner Prize contests. At a contest marking the 60th anniversary of Alan Turing's death, on 7 June 2014, 33% of the event's judges thought that Eugene Goostman was human. The event’s organizer had declared that Eugene Goostman had managed to pass the Turing Test.
There’s a lot of cynicism about this success. Critics argue that the bot never actually passed the Turing Test in the first place. They said that the questioning lasted only 5 minutes (which was not a sufficient amount of time) and that Eugene used wit to take the judge’s attention away from its non-human traits. The fact that Eugene played a 13-year-old boy whose first language wasn’t English also gave it a huge advantage because the judges could simply write off its mistakes to these facts.
Cleverbot, the successor to Jabberwacky was created by the British Artificial Intelligence scientist, Rollo Carpenter. Cleverbot's responses are not programmed in advance. Instead, it learns from human input: Humans type into the chat window and the system finds keywords or an exact phrase matching the input. After scanning through its pre-saved conversations, it responds to the input by finding how a human responded to that input when Cleverbot asked that question itself.
In 2011, Cleverbot participated in a formal Turing test at the Techniche festival at the Indian Institute of Technology, Guwahati. It managed to fool about 60% of the voters into thinking that it was human. However, on interacting it with Cleverbot, it is clear that it is not as intelligent as it was portrayed to be.
PC Therapist was able to convince 5 out of 10 judges that it was human in 1991. The author behind the program, Joseph Weintraub was a psychology graduate who became a Mainframe Computer Programmer. He built PC Therapist after getting frustrated using ELIZA. PC Therapist even sported a humor reminiscent of Woody Allen’s.
ELIZA was a chatbot created by Joseph Weizenbaum in the MIT Artificial Intelligence library in 1966. Eliza simulated conversation making use of a pattern matching and substitution methodology. She passed the words that users entered through a program and paired these words with scripted responses. This gave the users an illusion that the program had some level of understanding. But ELIZA had no built-in framework for contextualizing events.
Even though ELIZA was capable of engaging in discourse, she could not manage to converse with true understanding.
ELIZA acted as a Rogerian psychologist. In this capacity, all she really needed to do was to repeat the users’ statements back to them. In 1972, ELIZA had a conversation with another Artificial Intelligence system. ELIZA played a psychologist, while PARRY took on the role of a paranoid schizophrenic.
During some experiments that took place in the 1960s, some people had actually been fooled by ELIZA into thinking that they were conversing with an actual, human therapist that was talking to them via another computer.
Secretaries and non-technical staff at the MIT Artificial Intelligence Lab were convinced that they were having conversations with a real psychologist that could help and benefit them in a constructive way.
ELIZA was considered to be the first computer program capable of taking the Turing Test. However, she did not manage to pass the Turing Test. Inspite of this, she was able to convince a fair number of people that she was actually human. A large number of these people started talking to ELIZA about their problems and believed that ELIZA could help them deal with these problems.
But, most modern Artificial Intelligence researchers have not had needed a lot of convincing to realize that ELIZA was more of a gimmick than an actual, serious AI project.
Develop your Bot-Building skills with 2 months of FREE Bot-Building on the Engati Platform. No credit card required; start now!