A New Direction in Artificial Intelligence: Measuring Artificial Intelligence

The paper is devoted to a new scientific direction: measuring artificial intelligence (MAI). The paper gives the basic definitions and attributes of measuring artificial intelligence. The scope of tasks, in which the application of methods and means of this direction is necessary is defined. Examples of MAI based on the methodology of the regularizing Bayesian approach are given, which illustrate the implementation of the principles of traceable, trustworthy and explicable artificial intelligence and ensuring the stability of the solutions, obtained in conditions of significant uncertainty. The methodological foundations of the metrology of solutions of artificial intelligence systems are given.