The current computer system, which has evolved considerably, is part of the technological evolution of the company. Living without a computer is like a brake on progress, because they are the most important tool in any activity of daily life, they help us maintain our accounts, develop new technologies and connect to the world.
A conventional computer is capable of performing millions of operations per second, but it can not make decisions on its own. In other words, a computer can not perform tasks on its own. It must be programmed, but only one program is limited to one task. A specific task, and for several tasks requires several programs. Computer manufacturers provide us with the hardware necessary to run the computer, and software developers provide programs that allow us to use hardware as a tool, but what would happen if a computer could make decisions without using large amounts of specialized software?
The new technology developers have looked for alternatives to carry out this task. An amazing discovery was made 50 years ago when studying the human brain; It is possible to implement an artificial system based on the same architecture as the biological neural networks and their operation, so that 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, mainly to learn and make decisions. For the most part, living beings are biological systems that learn and are capable of making decisions, decisions that are generally based on survival.
Similarly, artificial intelligence systems require learning and decisions, but unlike biological systems, artificial systems work on the basis of mathematical algorithms and learning is induced for specific purposes. In artificial intelligence, there are many categories, but in the case of smart computers, the most used are artificial neural networks and genetic algorithms, among others.
Neural networks are nothing more than an emulation of the neuronal system of the brain, and each element of the biological system is replaced by a mathematical equivalent. A network of artificial neurons can perform similar tasks in the human brain, tasks that a common computer can not perform, such as image recognition, voice recognition and decision making. However, this system has the disadvantage of having a disadvantage compared to a programmed system. In other words, a network of untrained neurons is like a newborn baby coming into the world. Therefore, scientists have developed different models of neural networks, each with different capabilities and algorithms.
Nowadays, the use of artificial intelligence and neural networks is implemented in software, imitating the parallel nature of a neural network to a linear system. The most common applications are speech recognition, character recognition (OCR), image reconstruction, etc., but they are also implemented in the hardware, where the linear structure of the processors is replaced by a neural structure, which is extracts Part of the parallel nature of neural networks. . This new technology is called neuronal processors.
Neural processors are currently used in specific applications, such as robotics, for which it is easy to implement at the moment, but recently they have developed hybrid systems that use microprocessors and neural processors, generally used in servers.
I might think that artificial intelligence technology will help society improve their standard of living, but we must also bear in mind that inheriting our biological capabilities from a machine makes us more dependent on them. and that this addiction only slows down our intellectual abilities. capabilities To avoid this, you have to see these machines for what they are: tools to improve.
No comments:
Post a Comment