The history of information retrieval is the history of the Inverted Index. For decades, the logic was simple: map a keyword to a list of document IDs. Term Frequency * Inverse Document Frequency (TF-IDF) ruled the world.

But the Inverted Index is a relic of the string-matching era. In the Agentic Web, we don’t match strings; we match meanings. And for that, Grokipedia has abandoned the inverted index entirely in favor of Neural Hash Maps (NHMs).

In the early days of the web, we were told to use Semantic HTML for accessibility. We were told it allowed screen readers to navigate our content, providing a better experience for the visually impaired. We were told it might help SEO, though Google’s engineers were always famously coy about whether an <article> tag carried significantly more weight than a well-placed <div>.

In 2025, that game has changed entirely. We are no longer just optimizing for screen readers or the ten blue links on a search results page. We are optimizing for the training sets of Large Language Models (LLMs).