The desire to survive can develop intelligence artificial intelligence

What distinguishes man from a robot? Of course, in addition to excellent mechanical internal components, a person has a unique ability — he is able to feel. Based on the difference between the living and the dead, the world cinema has created a huge number of films that support the theory about emotion as about the defining of a truly human quality. However, despite all the attempts of scientists to create at least partly a living organism, able to show genuine emotion, in reality, the robot senses no more than the brick. Despite this sad reality, neuroscientists, and Kingson man and Antonio Damasio believe that they managed to develop a way to give inanimate robots feelings. If so, what is the result of the first “humanized” robot? Well, let’s try to understand together.

Can a thinking being, created by man?

Is it possible to create a thinking robot?

As the portal sciencenews.org the only way to give robots the ability to feel, may be training them to fear for their own existence. In order to avoid death, the robots would have to develop feelings, guide their behaviour to ensure the safety of your being. Researcher Antonio Damasio proposes to fill a humanoid androids so-called “artificial equivalent feelings” in which the built-in program would allow the machine to realize his fortune.

Modern “smart” machines often are designed to perform specific tasks, among which can be diagnosis of diseases, driving a car or win in the game of chess. However, the implementation of the algorithm, leading to the successful completion of the set before the robot task, has nothing to do with the creative intelligence, the ability to fail and to learn from their mistakes. From the point of view of Damasio, in this case, that feelings are the key ingredient, which both need robots for his transformation into a truly thinking being.

The question of humanity creating artificial intelligence may be one of the major issues of humanity of the XXI century

Feelings arise from the need to survive. When people support the robot in a viable state, the robot does not need to worry about your own self-preservation. So he doesn’t need feelings — signals that something needs to be repaired. Feelings encourage sentient beings to seek the optimum condition for their survival in nature that can help the intelligent machines learn to feel our own vulnerability in nature in the same way. In order to perceive potential threats, the robot must be designed to be able to understand his own internal state.

Experts from the University of southern California argue that the prospect of creating machines with emotions has recently been expanded thanks to advances in the development of two key areas of research: the creation of the so-called soft robotics and the study of the processes of deep learning. Development of methods for deep learning could allow to perform robotic technology complex calculations required to translate feelings into behavior that supports their existence. For the process of deep learning, neural networks of the brain could replace particular sets of related computing elements, simulating the nerve cells in the human body.

For in-depth learning requires multiple layers of the neural network, and using which the intelligent robot is able to track its internal state. Presenting the state of the environment in the form of a computing machine with the function of deep learning has been able to combine various data into a coherent assessment of the surrounding situation. However, the mere ability to feel the internal state will not play much importance in the case that the viability of a robotic body would not be exposed to danger. Robots made of metal are not worried about mosquito bites, paper cuts or indigestion, but if the robot was made of proper soft materials, equipped with electronic sensors, it could detect such risk, as the cut on the skin, threatening his insides, utilizing in the future a program for injury recovery.

The robot is able to perceive the existential risks that could learn to develop new methods of protection, instead of relying on pre-programmed decisions. In turn, developing new opportunities of self-protection can also lead to improved thinking skills and the development of intelligence.