Not so long ago, American entrepreneur Elon Musk and physicist Stephen Hawking in an open letter urged the developers of artificial intelligence (AI) not to create something that cannot be controlled. They believe that the threat posed by AI can be much more dangerous than nuclear weapons.
Rival
In March 2016, the AlphaGo computer, developed by DeepMind, in five games of the ancient Chinese board game Go defeated the Korean Li Sedol, one of the strongest in this discipline. This is a significant step forward compared to the victory of the Deep Blue computer over Garry Kasparov in 1997, because if chess has 35 variants of moves, then there are about 250 of them in Go.
The computer lost only in the 4th game, when Sedol made a non-standard move, the probability of which AlphaGo estimated as 1 in 10,000, but in the 5th game this mistake was taken into account. The president of the American Go Association, Andrew Okan, was a little annoyed, as he believed that in this game, the machine would reach the advantage over a person in a few decades.
Frightening speed
Scientists are already confronted with the fact that AI is capable of self-learning almost to the same extent as humans. The work of an artificial brain, like a human one, is based on a variety of neural connections that are accumulated as a result of the experience accumulated by a machine. Considering that the performance of the AI is much higher than that of the human brain, the learning process of the robot will go much faster.
Promotional video:
So, in 2014, a supercomputer was able to solve one of Erdрдs' mathematical problems, the solution of which is so complicated that a person is not able to check it - one file with an equation takes 13 gigabytes. It is obvious that in terms of intellectual work, AI has already outstripped humans, and after 2030, according to experts, artificial intelligence will surpass us in terms of the general level of consciousness.
How to understand it?
American blogger Tim Urban, who deals with the problem of artificial intelligence, noted that as AI becomes smarter than humans, we will have less and less chance of understanding it. Explaining what has been said, he gives an example. The chimpanzee can comprehend what a person is and what a skyscraper is, but a primate can never understand that a skyscraper was built by humans.
Likewise, we will never be able to understand the capabilities of AI, even if the machine tries to explain it to us - let alone try to do it ourselves. A computer could spend years trying to teach us the simplest notions of what it knows, and that would be hopeless,”sums up Urban.
You're fired
In the foreseeable future, we will face massive unemployment, futurologists are sure. And even those who consider themselves indispensable. This applies not only to physical or intellectual work, but also to creative activity. Machines are already making music and writing articles. Professor Moshe Vardi of Rice University in the US believes that most jobs will be occupied by robots within 30 years.
Some scientists believe that there is nothing wrong with this, since technological unemployment will open the door to a future where people will work not out of necessity, but for pleasure. Actually, the governments of some Western countries are already taking steps in this direction, in particular, experiments are being carried out to introduce a universal basic income.
Self-sustaining
Given the speed of AI development, smart machines will soon be able to do without a human at all. For example, a team of robotics from the French University of Pierre and Marie Curie has created a high-tech spider machine that can independently change the way of movement in the event of a breakage or loss of one of the limbs. It takes about 20 minutes to recognize the breakdown and test new modes of movement.
In the very near future, robots will learn how to independently replace broken or worn parts. The prospects for such machines, according to scientists, are truly endless. They can be used in places of natural or man-made disasters, as well as for exploring the ocean depths and distant space.
Almost like a human
In early 2017, Boston Dynamics demonstrated a new humanoid robot PETMAN, which is used to test army workwear. PETMAN mimics human physiology, moving and maintaining balance in the same way as a human would. Moreover, the machine is able to regulate the temperature and humidity of the "body", bringing the test conditions closer to natural.
The Atlas robot, another Boston Dynamics development, is capable of solving more complex tasks: opening doors, navigating difficult terrain, and even overcoming vertical obstacles. The machine is much stronger than a person and can carry loads over 100 kilograms. No wonder the developers say that the new robots strike terror into them.
Two ways
According to scientists, with the intensive development of AI, humanity has two options for the future: by the end of the XXI century, we can either disappear from the face of the earth, or become immortal. Pessimists are concerned that the slightest mistake along the way will lead to a swift end of the human race. For example, an AI programmed to tackle the challenges of climate change will determine that humans are the main obstacle to its work and will use its entire arsenal of robotic tools to eliminate this threat.
It will be virtually impossible to defend or hide from AI. So, one of the most powerful supercomputers at the moment, Nautilus, was able to foresee where bin Laden would be hiding, as well as predict the beginning of the "Arab spring".
Optimists are confident that superintelligence will be able to solve all our pressing everyday problems. AI will provide not only the comfort and safety of our life, but will also be much more intensive than a human to develop the fields of fundamental and applied science. And the first thing that is expected of him is the development of drugs that will make the human body immortal.
Taras Repin