Yoyao HsuehFounder, Floyi · Founder at Floyi

Information Gain: Why AI Content Says Nothing New

31 Jul 2026 · 1 min

Watch on YouTube

0:57

In this concise monologue, the speaker argues that AI can rapidly produce articles, but the output often mirrors existing top‑ranking pages, offering no new value. He introduces the concept of "information gain"—a term from Google patents that measures the net new value a page contributes beyond what already exists. The speaker outlines a two‑step process to achieve information gain: first, cover all essential, shared claims (the "table stakes") identified by reviewing the top ten ranking pages for a query; second, add unique perspectives or angles that none of those pages cover. By ensuring both completeness and originality, a page can provide genuine information gain and stand out to Google.

Chapters

AI can write a 1,000-word article in 10 seconds. The problem is, so can everyone else. And it's the same article, same point, same structure, same facts already sitting on page 1. Now there's a concept from Google's own patents for exactly what's missing. Information gain. The net new value your page adds that no one else has. Now information gain is 2 moves. Cover everything the topic requires. Then say the things only you can say. Here's how to do that. Put your query into Google and open the top 10 ranking pages for your topic. List every claim they make and compare it to the others. The claims 2 or more of them share, those are your table stakes. Miss them and your page looks incomplete and Google notices. Now find what none of them say. The angles nobody covers. And this is where information gain lives.