Definition
Plain language
Giving newer pages a boost so search results aren't full of stale material.
As stated in the literature
A recency prior applied at ranking time, boosting recently published or updated documents; interacts badly with feedback loops where new content is itself derived from prior model output.
Also called: freshness signal, recency weighting
Why it matters: It keeps results current, but when the newest pages are themselves rewrites of earlier AI answers, the freshness boost accelerates the spread of whatever the model said first.
For example, a search for a phone's battery life pushes this year's review above a well-written one from six years ago.
Heard on the show
“Real search stacks have deduplication, freshness weighting, domain authority, maybe explicit AI-content handling, and the authors say straight out they don't know how those systems work.”Episode 248 — One Self-Written Page Is Enough to Collapse an AI Search Answer