Context Is a Property, Not an Object

2026-07-09 · Source: Intentional Arrangement · Field: Technology & Digital — Artificial Intelligence & Machine Learning, Knowledge Representation & Management · Depth: Advanced, quick

Summary

This essay redefines how context is understood within AI and knowledge infrastructure, particularly concerning Large Language Models (LLMs). It critiques the prevailing AI industry approach, termed "context engineering," which treats context as a temporary container or a fixed-capacity token window. Instead, drawing from library and information science principles, specifically ANSI/NISO Z39.19 Guidelines, the article posits that context is a property—declared, tested, displayed, and maintained—attached to concepts, terms, and relationships. This structured, logical, and intentional view of context, rather than a runtime artifact, enables persistent knowledge. Adopting this perspective shifts focus from merely copying text into windows to designing reusable architectural components that inherently carry context, preserving critical elements like provenance, scope, hierarchy, and disambiguation across queries.

Key takeaway

For AI Architects designing knowledge systems for LLMs, recognize that treating context as a transient object limits long-term knowledge persistence. You should instead architect systems where context is an intentional, structured property of data, preserving provenance, scope, and hierarchy. This shift moves beyond mere token window management, enabling more robust, reusable knowledge assets that consistently survive across queries and improve overall system reliability.

Key insights

Context should be treated as a structured, intentional property of knowledge, not a temporary container for LLMs.

Principles

• Context is a declared, structured property.
• Design systems for persistent knowledge.
• Focus on logic, not probabilistic mirroring.

Method

Shift from filling LLM context windows to designing intentional, reusable objects that imbue context as part of a larger architecture, preserving provenance, scope, hierarchy, and disambiguation.

In practice

• Design knowledge attributes for persistence.
• Curate context for concepts and relationships.
• Integrate human and machine sensemaking.

Topics

• LLM Context Management
• Knowledge Infrastructure
• Information Science
• ANSI/NISO Z39.19
• Context Engineering
• Semantic Architectures

Best for: AI Engineer, NLP Engineer, Research Scientist, AI Architect, Machine Learning Engineer, AI Scientist

Related on AIssential

• Context Engineering: An Introduction to the Information Environment for LLMs — Context engineering designs the information environment for LLMs, maximizing signal within finite context windows.
• The Context Problem — AI's "context" as a billing unit conflates capacity with coherence, leading to inflated costs and degraded model performance.
• The Agent Stack - Part 5: Context, Retrieval, and Memory — Context is the runtime-assembled working set, not the model's inherent knowledge.
• Give Your AI Unlimited Updated Context — An LLM-maintained personal wiki creates a persistent, compounding knowledge artifact, eliminating repetitive context setup.
• We Don’t Have an AI Problem. We Have a Context Problem. (And MCP Exposes It) — Unstructured context, not model limitations, is the primary bottleneck in scaling and maintaining LLM-based systems.
• Is RAG Still Needed? Choosing the Best Approach for LLMs — LLM context injection methods, RAG and Long Context, each offer distinct architectural trade-offs for data handling.

Open in AIssential →

Editorial summary, takeaway, and curation by AIssential. Original article published by Intentional Arrangement.