Russian computer program for the first time in history passed the Turing test
Developed back in 2001, the Russian program was the first in storieswhich was able to perform the test of the deceased in the 1954 year of Turing. For the first time in history, a computer managed to deceive a person by posing as a 13-year-old child from Odessa. Thus, for the first time in history, engineers were able to succeed in developing some kind of artificial intelligence. Although, in fairness, it is worth noting that around the test itself and its results for many years, disputes do not abate. Many people disagree that a program that passes the Turing test can be considered a full-fledged artificial intelligence. And they are certainly right. We are still very far from full-fledged artificial intelligence, which could replace a person.
The famous Turing test is an empirical test, the main idea of which belongs to Alan Turing and was put forward by him in the article “Computers and Mind”. This article was first published in the philosophical journal Mind in 1950. In his article, the British mathematician set himself the goal of determining whether the machine is capable of thinking. The standard interpretation of this test is: “A person interacts with one computer and one person. Based on the answers to his questions, he must decide for himself whether he communicates with him: with a computer program or with a person. The main task of the computer program is to mislead the person, forcing him to make the wrong choice.
All participants in this test do not see each other. If the judge is unable to say with certainty which of his interlocutors is a person, then the computer is considered to pass the test. In order to test directly the intelligence of the computer during the test, and not its ability to recognize speech, the conversation between the participants is conducted only in the text-based communication mode. In this case, the correspondence should take place at controlled and predetermined intervals so that the judge could not make his conclusion based on the speed of responses. In the days of Alan Turing, computers reacted more slowly than people, and now this rule is necessary because they react much faster than people.
According to Turing's test, a computer could be called thinking if he could convince 30% of the “negotiators” who communicate with him in the text messaging mode, that they are talking to a living person. In the 2014 year, for the first time in history, the test failed to pass the Russian program, which was written back in 2001 in St. Petersburg. The program was created by Russian engineer Vladimir Veselov (currently residing in the USA) together with Ukrainian Yevgeny Demchenko, who now lives in Russia. The computer program they compiled during the competition, which was conducted by the University of Reading (located in the UK), was able to convince 33% of the jury members that 13-year-old teenager Yevgeny Gustman hails from the Ukrainian city of Odessa.
The author of the computer program (bot) Vladimir Veselov, after winning the contest, told reporters that the main idea of the developers was that the 13-year-old boy could say that he knows everything, but his age at the same time indicated that he cannot know everything. According to him, a lot of time and effort was spent on creating and developing a character that would have a believable personality. It is noteworthy that for the first time in history the computer was able to pass the Turing test namely 8 June 2014 of the year, on the 60 anniversary of the death of the famous mathematician and logic, which more than half a century ago had a significant impact on the development of information technology worldwide.
Vladimir Veselov, noting the success of his program, stressed that passing the Turing test is a remarkable achievement for the creators of the bot program. In addition, he hopes that this achievement will raise a wave of interest in the development of artificial intelligence and advanced chat bots. At the same time, not everyone expresses this achievement as extremely positive. The Englishman Kevin Warwick from the University of Coventry remarked on this point that the success of Veselov and Demchenko’s computer program once again confirms the fears of many experts about the future development of AI. According to him, computers that will be able to deceive a person will be able to become assistants to cybercriminals.
It is worth noting that the Turing test, as already stated above, has quite a few opponents. Some of them criticize the test itself, some say that passing this test is a kind of “sport” for programmers, since the test does not carry any special practical burden. Yuri Chekhovich, a senior research fellow at the Computing Center of the Russian Academy of Sciences, Candidate of Physical and Mathematical Sciences, notes that the question of whether Alan Turing's test is the equivalent of a thinking person is most likely a question of philosophy. It is no coincidence that scientists throughout the world have been arguing about it for 50 for years. In Chekhovich’s view, computer programs that try to pass the Turing test are no different from a large number of other computer programs that exist today.
For example, nowadays computers are already quite easily beating up a person in chess, but it does not occur to anyone to say that computers are able to think. In this case, the computer solves a very specific problem. In addition, there are a large number of computer programs that can write poetry. At the same time, modern computers can compose small verses by millions per second, but it is unlikely that the computer ability to put words into rhymes can be called a sign of intelligence.
Information sources:
http://russian.rt.com/article/35721
http://www.rg.ru/2014/06/09/intellekt-site.html
http://www.interfax.ru/world/380143
http://www.gazeta.ru/science/news/2014/06/09/n_6216801.shtml
Information