Physics is essential for the development of artificial intelligence
Artificial intelligence is gradually surpassing human capacities and some researchers even claim that it could one day replace physicists. However, physics represents a real field of study for a better understanding of AI and particularly in-depth learning.
ARTIFICIAL INTELLIGENCE IS EXPERIENCING A NEW GROWTH
The exact definition of Artificial Intelligence (AI) remains rather vague. In fact, when we approach this discipline, we have to rely not only on logic but also on statistics, mathematics, probability, computer science, and probably also on so-called “computational” neurobiology. But if you ask a person whose knowledge is limited to the real world, the answer often remains the same: artificial intelligence is a machine that imitates the cognitive capacity of a human being.
But the very notion of intelligence has its limits in certain fields… it has come to be described as a technology implemented in order to understand automatic learning. The very origin of AI stems from a frightening but well-founded theory: it comes from a network of artificial neurons inspired by the connected layers of mammalian neurons located in their visual cortex. They are called convolutional neural networks or CNN.
Convolutional neural networks have made it possible to learn about two-dimensional digital images, particularly in computer vision tasks such as deciphering words and objects in these images. New techniques for deep learning have resulted from this revolution. They have made it possible to explore the impossible in order to help humanity to be better off. Such is the case of lung tumors, which can now be easily detected with a CT scan. The same is true of computers, which are becoming the masters of their trade, capable of driving vehicles and leading to victory a game of chess or board games against a simple human being with great intelligence.