The Semantics of Semantics

2026-02-22 · Source: Intentional Arrangement · Field: Technology & Digital — Data Science & Analytics, Artificial Intelligence & Machine Learning · Depth: Advanced, long

Summary

The term "semantic layer" in the data industry, widely used by vendors like dbt, Looker, AtScale, and Power BI, refers to SQL abstraction and metric definition, not formal semantics as understood in computer science. Originating with Business Objects in the early 1990s as a SQL generation tool, the concept was retroactively named and patented to shield business users from writing SQL. Modern data stack products repurpose this term for YAML-based metric definitions and SQL compilation engines, which define metric calculation rules, join paths, and access policies, but do not perform reasoning, inference, or consistency checking. This contrasts sharply with formal semantics, which involves machine-interpretable meaning, formal logic, ontologies (like OWL), and automated reasoning, as envisioned by the Semantic Web. The market for these "semantic layers" is estimated at $1.73 billion in 2025, projected to grow to $4.93 billion by 2030, with 53% of organizations actively pursuing implementation.

Key takeaway

For data leaders evaluating "semantic layer" solutions for AI initiatives, you must distinguish between metric governance tools and true knowledge representation systems. Your organization risks confusing SQL compilers with business glossaries for the formal semantic infrastructure (ontologies, knowledge graphs) required for performant AI agents. Prioritize solutions that offer genuine logical reasoning and knowledge representation if your goal is to enable advanced AI capabilities, rather than just consistent metric reporting.

Key insights

The data industry's "semantic layer" primarily offers SQL abstraction and metric governance, not formal semantic reasoning.

Principles

• Metric definition is distinct from ontology engineering.
• SQL generation tools are not semantic reasoning systems.

Method

Modern "semantic layers" compile declarative metric and dimension definitions (YAML, DSL, JavaScript) into SQL queries at runtime, handling join resolution and aggregation semantics.

In practice

• Use dbt Semantic Layer for governed metrics via YAML.
• Implement LookML for proprietary declarative data modeling.
• Consider AtScale for virtual OLAP cubes and autonomous aggregate management.

Topics

• Semantic Layer
• Metric Definition
• SQL Compilation
• Formal Semantics
• Knowledge Graphs

Best for: CTO, VP of Engineering/Data, Director of AI/ML, Data Scientist, Data Engineer, AI Architect

Related on AIssential

• Why Semantic Data Layers matter to product teams — Semantic data layers standardize business concepts, enabling precise AI-driven self-service analytics and consistent data interpretation.
• Ontologies, Misunderstood — Ontologies reason over abstract forms and class relationships, distinct from data-centric relational logic.
• Why machines that can write poetry still can’t agree on what “good” means — Ontology-based semantic layers are crucial for grounding agentic AI in explicit, governed business meaning, overcoming inherent linguistic ambiguity.
• Formalizing the DATASUS RTS: An Ontological Model for a Resource Description Framework Knowledge Graph — Transforming relational health data into a W3C-compliant knowledge graph improves interoperability and semantic querying.
• The Semantic Infrastructure Opportunity — Semantic infrastructure, especially thesauri, transforms raw entity resolution matches into meaningful, contextualized knowledge.
• Ontologies Are the Intentional Core of a True Knowledge Graph — Formal ontologies provide the stable, intentional semantic backbone essential for reliable enterprise AI and knowledge graphs.

Open in AIssential →

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