Artificial nerve coupling is self-learning

French scientists have successfully built artificial synapses that are self-learning as in the human brain.

Development and advancement in the field of artificial intelligence (AI) are largely imitative technologies that mimic the behavior of the human brain. Such AI systems are called neural networks , according to Science Alert. These include algorithms for verbal and visual recognition, but running a neural network consumes a lot of energy.

Scientists successfully fabricated the artificial neuron joint on a chip. (Photo: iStock).

At present, scientists at the National Center for Scientific Research (CNRS) and the University of Bordeaux, France, have successfully developed an artificial neuromuscular junction called a direct memristor on a child. chip. It paves the way for more intelligent systems in the future, requiring less time and energy to automatically learn. The results are published in the journal Nature Communications today.

In the human brain, the neural coupling acts as a link between neurons. These connections are strengthened and improved as more synapses are stimulated. Memristor works in the same manner.

The team built the memristor from a ferroelectric layer wrapped between two electrodes. When using voltage pulse, their resistance changes like biological neurons. The connection between synapses will increase as the resistance decreases and vice versa. The learning ability of the memristor is based on the change of resistance.

Researchers have now successfully built a physical model that predicts the operation of the memristor. In the future, the memristor promises to be a technology that enhances the self-learning ability of artificial intelligence.