Scientists have built artificial brains with the ability to learn verbal communication

Scientists have built artificial brains with the ability to learn verbal communication

It is common today to build a neural network, an intelligent robot, or similar technologies. We live in an age where scientists, companies in the field of new technologies and even big buyers are looking for the imperceptibility of smart products.

The joke that people jokingly told each other about why you act like a robot is over. Now is the age of perceptible robots. Robots that not only know what their task is, but also know how to get a job done or when to stop. In simpler terms, robots try to enter a special world called emotions and tangible verbal communication.

Have you ever tried to chat with a robot like CleverBot, you will notice a short distance from your conversation that this conversation is not normal and is completely emotionless. But if the same robot was equipped with artificial nerve neurons, then the situation would take on a different color and smell. There are about 86 billion neurons in the human brain, 760 million neurons in the cat brain and 16 million neurons in the frog brain. These neurons help the brain of the creatures to interact well with the environment. Recently, Italian researcher at the University of Sassari in Italy, Professor Borono Glucio, together with researchers at the University of England, succeeded in building an artificial brain with two million simulated neurons. This artificial brain has more light than the brains of other creatures on the planet, such as bees, beetles and crabs. But the point that makes this artificial brain important goes back to the learning and training process of this neural network. The design of this artificial brain is so advanced that it allows researchers to teach this artificial brain based on verbal communication. This amazing brain is called ANNABELL. Annabelle is the title of Artificial Neural Network with Adaptive Behavior Exploited for Language Learning.

Natural Language Processing

The design of this neural network is very similar to the human brain. It uses the mechanism of verbal communication and reasoning to perform complex processes. Interestingly, the process of learning and developing the skills of this artificial brain is very similar to the human brain. The human brain is not like computers that work with specific programs or rules. The human brain uses a set of cognitive skills to try to interact properly with its environment and grow over time. Just as this artificial brain is unique, the experiments performed on this brain are also unique in their kind. In conventional methods, a set of behavioral and cognitive sciences in a specific format is provided to neural networks. In simpler terms, a primary database is provided to the neural network, and this neural network develops its skills based on this database; But Annabelle goes through the learning process only through verbal communication with humans and without any prior knowledge.

How does this artificial brain work ?

This network goes through the learning process in two ways. The first method is synaptic plasticity. In this method, if two neurons are active at the same time, they will be able to communicate effectively with each other. The second method works on the basis of neural gating. In this method, a certain number of neurons are considered as on and off switches. The combination of these two methods allows the neural network to control the signals that turn switches on or off, as well as the flow of information that is exchanged between different parts of the virtual brain.

 Once the database is saturated with words and phrases, Annabelle becomes empowered to be able to answer 85% to 95% of the various questions she is asked. For example, if you ask Annabelle how many games she has played? He will reply: “I have played Space Invaders, Pac Man and three other games.” As Discovery magazine reports, a significant portion of the artificial brain’s behavior is very similar to that of humans, although it has a long way to go before it behaves exactly like humans. The team plans to install the neural network on a robot after successfully testing it. This allows the neural network to observe (understand more precisely) its surroundings and the learning process to expand in a more natural way.