Socratic Intelligence as Applied for Artificial and Human Agents

Abstract

This essay introduces the concept of Socratic Intelligence and analyzes the following premises: 1. Artificial agents and human agents learn through observations, interactions, and examination of the observations and interactions. 2. What is learned by artificial and human agents depends on metacognition and reinforcement. 3. Socratic Intelligence includes epistemic curiosity, critical self-awareness, assumption analysis, dialogical competence, and intellectual humility. 4. Generative Socratic Intelligence can tutor human agents and help them to learn by Socratic questioning integrated into Intelligent Tutoring Systems or Large Language Models. 5. Generative Socratic Intelligence can help itself self-regulate by Socratic method applied to computational processes.