Introduction to Artificial Intelligence

The aim of artificial intelligence (AI) is twofold. On the one hand, AI is concerned with solving problems related to human or animal activities of various kinds: perception, planning, data interpretation, diagnosis, decision-making, language comprehension, design. On the other hand, AI seeks to better understand and model intelligence. This brings it closer to the cognitive sciences, from which it also draws inspiration for the design of models (memory, reasoning, learning). The need to restrict activity to a limited field of application, and to draw on diverse types of knowledge, quickly became apparent in AI. This symbolic approach to AI gave rise to knowledge-based systems.

Another approach, known as connectionism, is based on the functioning of the cerebral cortex. A neural network is formed by the interconnection of a large number of artificial neurons. It has a number of interesting properties, including the ability to learn from large numbers of examples.

Recent successes have put AI in the media spotlight: chess, go, poker, Martian robots, the American question-and-answer game Jeopardy, image recognition, speech recognition, video games and more.

This article presents the different models of AI, as well as the associated learning methods. It also describes a wide range of applications (medicine, industry, military, banking, law, etc.) and discusses the ethical aspects associated with these applications.