In Italy it went almost unnoticed - perhaps out of fear of derision - a fact that happened a few months ago and instead aroused if not a stir at least disquiet in the rest of the world: the "extinguishing" of two "robots" who suddenly began to talk among them in a language they themselves invented.
Some people have considered the event as "exciting" from a scientific point of view, while those, like the professor of robotics Kevin Warwick, immediately highlighted the dangers that may derive from it, especially if the robots had a some military skills ...
But what really happened?
It was an experiment performed by Facebook. Two were installed chatbot (v. article), called Alice and Bob, who had to improve their negotiation skills. They should have used the English language between them and then, later, with the possible users. At a certain point, without any warning, they began to use their own language incomprehensible to others. The experts of Facebook they therefore preferred to switch off the two machines because the expected result has not been achieved.
After this alarm it was discovered that similar events seem to have already happened (at least) near Google.
So what is happening? Can Artificial Intelligence pose a risk? And, if the answer is yes, can it pose a risk to humanity?
The answer is not simple. Perhaps the answer is not there.
Anyone, like me, was a passionate reader of science fiction as a boy can not not have read Asimov. But to talk about the younger ones, in recent years the films that talk about forms of Artificial Intelligence that interfere with the natural development of human civilization are innumerable.
How can we be sure that a program, perhaps developed by a group of ecological activists to protect nature, can not one day take control of the world and decide that the human race is the real danger to nature and therefore decides to "clean up" "?
And, beware, anyone who thought that this is just science fiction, at the point where we are, would not understand anything. In fact, it seems that reality lately surpasses science fiction!
If we consider that stopping evolution in this field is practically impossible, the worst thing is not to talk about it for fear of being laughed at. What is needed is a careful check on what happens and on the research in order to prevent possible problems that could arise from a field, Artificial Intelligence, absolutely new.
In this sense it is only to be commended the discussion started in the European Parliament regarding the need for regulation of robots (v.articolo).
(image: frame "I Robot" - 20th Century Fox)
