What makes a computer system "intelligent"? A framework for practitioners

A new paper from arXiv offers a foundational attempt to define what engineers and researchers mean when they refer to "intelligent systems"—a term that has gained prominence as artificial intelligence applications have matured across industries.

Defining the term#

The paper takes an integrative approach to characterize intelligent systems by identifying their core properties and recurring component structures. Rather than offering a narrow technical definition, the authors provide a pedagogical framework intended to give systems engineers a shared vocabulary and mental model for analyzing and designing these systems. This broader perspective aims to bridge the gap between abstract AI concepts and practical implementation challenges.

Scope and approach#

The research examines both established and contemporary examples of intelligent systems to demonstrate how the proposed framework applies across different contexts and generations of technology. By grounding the description in real-world cases, the authors illustrate where the conceptual model holds explanatory power and where it may need refinement.

The paper, initially submitted in August 2020, has undergone multiple revisions—most recently in August 2026—suggesting the authors have continued refining their framework in light of evolving AI capabilities and deployment patterns.

Why it matters#

As organizations increasingly deploy AI-powered systems in mission-critical contexts, having a clear, shared understanding of what constitutes an "intelligent system" becomes essential for communication among teams, risk assessment, and architectural decision-making. This work contributes a structured vocabulary that practitioners can apply across domains, from autonomous systems to knowledge management tools.