What is a neural network? What are neural network robots?

The computer system today who have grown considerably, has been part of the technological evolution in society, living today without computers is like stop progress, as they are the most important tool in any activity of daily living, they help us to maintain our accounts, develop new technologies and connect us to the whole world.

A conventional computer is capable of performing millions of operations per second however, it’s not capable of making decisions by itself, in other words a computer can not perform tasks by itself, it requires to be programmed, but a single program is limited to do only a certain task, and for multiple tasks requires multiple programs. Manufacturers of computers, provide us with the hardware needed to operate the computer as well as software developers provides the programs that allow us to use hardware as a tool, but, what would happen if a computer would be able to making decisions without the use of large amounts of specialized software?

Developers of new technologies have sought alternatives to achieve this task. An astonishing discovery was made 50 years ago during the study of the human brain; it is possible to implement an artificial system based on the same architecture of the biological neural networks and their operation, so they develop artificial intelligence and neural networks.

Artificial intelligence is nothing more than a set of techniques based on the behavior of a human brain, primarily in learning and making decisions. In his majority, living beings are biological systems that learn and based on that learning are capable of making decisions, where these decisions are generally based on survival.

In the same way the artificial intelligence systems require learning and making decisions, but unlike biological systems, artificial system operates on the basis of mathematical algorithms and learning is induced for specific purposes. Within artificial intelligence there are many categories, but in the case of intelligent computers the most used are artificial neural network and genetic algorithms among others.

The neural networks are no more than an emulation of the neural system of the brain, where each element of the biological system is replaced by a mathematical equivalent. An artificial neural network can do similar tasks of the human brain, tasks that a regular computer is unable to perform as image recognition, speech recognition and making decisions, the disadvantage of this system compared to a programmed system, is that it needs to be trained, in other words, a neural network without training is like a newborn child coming into the world, so scientists developed various models of neural networks each with different abilities and different algorithms.

Today the use of artificial intelligence and neural networks are being implemented in software, emulating the parallel nature of a neural network to a linear system. The most common applications are voice recognition, character recognition (OCR), image reconstruction and more, but are also being implemented in hardware, where lineal structure of processors is changed to a neural structure, which takes advantage of the parallel nature of the neural networks. This new technology is called Neural Processors.

Currently neural processors are being used in specific applications, such as robotics, where at the moment, it’s easy to implement, but lately have been developing hybrid systems using microprocessors and neural processors, which are usually used in servers.

One might think that the technology of artificial intelligence will help society achieve a better standard of living, but also we must keep in mind, that inheriting our biological capabilities to a machine makes us more dependent on them, and that dependency, only inhibits our intellectual capacities. To avoid this, we must see these machines for what they are: tools to improve.