Data Modeling is Dead (Again), 2026 Edition. Part 1

2025-11-05 · Source: Practical Data Modeling · Field: Technology & Digital — Artificial Intelligence & Machine Learning, Data Science & Analytics · Depth: Advanced, short

Summary

The article addresses the recurring "data modeling is dead" narrative, which resurfaces with each new technological hype cycle, currently driven by AI, LLMs, and agents. It outlines several common arguments for this claim, including the idea that expanding context windows eliminate the need for curated schemas, enabling direct processing of messy JSON blobs. Another argument centers on the "One Big Table" (OBT) approach, favored for its simplicity and LLM-friendliness over complex SQL joins. The piece also discusses "Just-in-Time" modeling, where agentic AI infers and discards data models dynamically, and the "Synthetic Data" loop, suggesting AI-to-AI communication might bypass human-readable tables. Furthermore, it highlights the "Tabular Data is Dead" argument, asserting that unstructured data holds more value, and the "No Need to Learn Anything…Because AI" argument, which posits that AI's capabilities render traditional data skills like SQL and data modeling obsolete. The author plans to counter these arguments in subsequent articles.

Key takeaway

For data professionals evaluating the impact of AI on their roles, recognize that claims of "data modeling is dead" are part of a recurring hype cycle. Do not prematurely abandon foundational data modeling skills or dismiss structured data's value. Instead, prepare to understand how AI can augment, rather than replace, rigorous data design, as future articles will likely demonstrate its continued necessity.

Key insights

New AI capabilities are fueling a recurring narrative that traditional data modeling is obsolete.

Principles

• Hype cycles frequently declare core data practices dead.
• Abstraction layers often lead to devaluing foundational skills.

In practice

• Context windows enable "Schema-on-Read 2.0" via LLMs.
• One Big Table simplifies LLM querying by avoiding joins.
• Agentic AI can infer and discard data models dynamically.

Topics

• Data Modeling
• Large Language Models
• Data Architecture
• Synthetic Data
• NoSQL

Best for: Data Scientist, Data Engineer, Machine Learning Engineer

Related on AIssential

• Are the Levels of Data Modeling Outdated? (Redux) — Data modeling levels remain relevant, but require flexible application and human judgment, especially with AI acceleration.
• How Data Professionals Use AI — and What Really Matters — Widespread AI adoption among data professionals amplifies existing data quality challenges, making trust and verification paramount.
• No Free Lunch: The Debt, The Excuses, and The Reality of Data Modeling — Intentional data modeling is critical for agility, preventing chaos, and aligning data with business processes, regardless of company size or AI advancements.
• Modern Data Report 2026: What 500+ Data Leaders and Experts Say on AI-Readiness — Enterprise data foundations are insufficient for AI, hindering automation and decision-making due to issues with trust, context, and discoverability.
• Agents are ready but your architecture probably isn't — AI value stems from outcome-driven design, robust data infrastructure, and proactive governance, not just model selection.
• The 2026 AI Data Engineer Roadmap — AI shifts data engineering from syntax mastery to strategic system design and conceptual understanding.

Open in AIssential →

Editorial summary, takeaway, and curation by AIssential. Original article published by Practical Data Modeling.