The Semantics Gap, Part 1: Why AI Still Doesn’t Understand Your Content

Posted by Kerry Landis | Jul 10, 2026 | Content | 0 |

The Semantics Gap, Part 1: Why AI Still Doesn’t Understand Your Content

https://medium.com/@kerrylandis/the-semantics-gap-part-1-why-ai-still-doesnt-understand-your-content-f3f2c56029f4

Many in the AI industry are solving the wrong problem.

Everyone is obsessed with better models, embeddings, retrieval, and agents. Meanwhile, AI still hallucinates and the real problem sits untouched.

AI isn’t failing because it lacks intelligence. It’s failing because it lacks semantic structure.

The next AI battleground won’t be models. It will be meaning.

You can’t reason over what you don’t understand.

And AI doesn’t understand your content…yet. Is your content ready?

About The Author

Kerry Landis

Kerry is a Senior Technical Staff Member at IBM and Chief Architect of IBM’s Content Ecosystem, responsible for the documentation authoring and publishing tools as well as the AI search and retrieval systems that deliver answers to customer queries. These tools support the creation of 10’s of millions of pages of content annually delivered to the web and AI tools across the enterprise. He is focused on where IBM is headed over the next 3–5 years for content delivery and architecture to ensure the success of the very customers that purchase products. This means making critical decisions now that lay the foundation of an agentic framework to deliver highly accurate and detailed responses. Kerry works for IBM, but his opinions are his own.

The Semantics Gap, Part 1: Why AI Still Doesn’t Understand Your Content