Big Brain is a personal-archive database. It started as one curator's pile of bookmarks and grew into a 601,974-item index of everything that passes through a working day — every tweet that got posted, every interview that got watched, every news headline that got read, every encyclopedia entry that got written. There is no team. There is no funding. There is a Mac in Las Vegas and a long-running habit of writing things down.
This page is the public window. The capture stream behind it is held back at the source — health, location, messages, viewing history, the social graph — none of that lands here. What you'll find below is the slice that was already published somewhere else: the Twitter archive, the news desk, the episode transcripts, the wiki entries. Numbers refresh roughly every four hours. Everything else is on the other side of a Tailscale login.
A working archive, not a content firehose. These are the public sources Big Brain watches; counts are item-level, not megabyte-level.
Counts only. Item text is held back until it's either summarized into a wiki entry or republished by one of the spine pages below.
When a tweet or a transcript clears summarization it lands as a wiki entry under /llm-wiki/. A few of the most recent ones:
Pages that are themselves public faces of brain.db — go here for the rendered, edited, curated views.
For Big Brain to be useful it has to capture more than it ever publishes. The deploy pipeline runs a strict source whitelist — anything outside it never leaves the host.
The full dashboard view of brain.db lives at /brain/dashboard/ on the Tailscale network and is not reachable from the public internet.