Notes

Glossary

Dark matter
The knowledge inside a creator’s videos that exists nowhere else as text: what people, search engines and AI answer engines cannot read, quote or link to unless they watch the video.
The Dark Archive
Our study of 1,284 English-language long-form YouTube channels: where their knowledge lives outside YouTube, and whether machines can read it. The full report is a paid product.
Machine readability score (MRS)
A score out of 100 that we give a creator’s own website. It adds up five parts: structured data, text a machine can extract, access for AI crawlers, attribution of each page to its source video, and systematic publishing.
T0 Dark
A channel with no external link anywhere. Its work exists only on YouTube.
T1 Satellite
A channel with rented links only: social profiles, a link-in-bio page, Patreon or a merch store, but no domain of its own.
T2 Beacon
A channel with its own domain that resolves, but no content index: no blog, no episode pages, no feed.
T3 Dormant
A channel whose site has a content index that has gone stale, or whose posts are not derived from its videos.
T4 Sovereign
A channel that systematically republishes its videos as text on a site it owns: regular posts, most of them derived from the videos, each long enough to read on its own.
Answer engine
A search product that answers a question in prose and cites its sources, such as ChatGPT search, Perplexity or Google’s AI Overviews.
AI crawler
A bot that fetches web pages for an AI company, to train models or to answer searches. A site allows or blocks each one by name in its robots.txt file.
llms.txt
A plain-text file at the root of a site that describes it for language models. It is a proposal, not a standard, and no major engine has documented that it uses the file to choose citations.

The notes that use these terms are at Notes.