Cognitive Approaches to Intelligence: Asymptotic Rationality and Continuum of Distinction Tasks

Having artificially intelligent machines that think, learn, reason, experience, and can function autonomously, without supervision, is one of the most intriguing goals in all of Computer Science. The goal of this paper is to outline some basic principles and concepts that are the backbone of real, autonomous Artificial Intelligence. The design of hybrid artificial intelligence methods and algorithms is presented using the limiting generalization paradigm. The results of the study can be used to create cognitive agents and robots, as well as the cognitive Internet.