Wednesday, October 12, 2016

Alan Turing's Test


Turing Test, as we mentioned, is designed to determine the potential of artificial intelligence, close to human intelligence. The classic interpretation of this test can be expressed as: a person interacts with a computer or a person. Based on the answers to certain questions, one must determine who is his companion: the computer or human. And in the computer program functions include human deception and wrap it to the wrong conclusion. During the test, none of the participants can not see each other.

Alan Turing's Test

Alan Turing said that the machine can recognize the thinking, if it can be misleading 30% of the people-in the correspondence. And for decades to implement this test no one could, but the 2014 edition of «The Independent» reported that the computer program from this reliable academic company was able to convince people that they exchange messages with a boy of thirteen. It is, in fact, means that in fact the Turing test has been passed.

Interesting Tests

It is interesting that a computer program called "Eugene Gustman" was created Russian programmers, and the test was organized by scientists from the University of Reading in London at the Royal Society. In the experiment, 33% of participants believe that talking with a real person. By the way, "Eugene" claimed he was a boy of thirteen, who lives in Odessa.

This fact, of course, is a huge step in research on artificial intelligence, but scientists say that this achievement can be used by cybercriminals.

"Eugene Gustman" - this is the first program, passed the test, though there are also other programs that are close to this result. These include such as «JFRED», «Elbot the Robot», «Ultra Hal» and «Cleverbot».

"Eugene Gustman" - the New Way

According to one of the founders of "Eugene" Vladimir Veselov, the basic idea of ​​the developers was to "Eugene" knew everything and, at the same time, did not know anything, and to develop a program with the "real" person was spent huge amount of time. Professor at the University of Reading's Kevin Warwick says that in the field of artificial intelligence has so far been no more than controversial and landmark event.

We, in this case, could not agree more with the specialists, because the very fact that the car was able to convince people that it is a man, can transform the entire global computer system. Imagine what can generally be consequences if cybernetic sides possessing artificial intelligence, will conduct a live conversation with people, but spam bots will not be recognized?