About this office

The Long Record is a filings office for personal conduct. You write down what you have done, what the evidence is, who will vouch for a specific line of it, what you would rather nobody brought up, and what you have committed not to do again. Then it sits at a permanent address and does not move.

It is not a social network. There is no feed, no follower count, nothing to refresh. Almost nobody will read your record the way you would read a letter. That is fine — it is not really written for them.

What it actually does

The Long Record does not try to prove that anybody is good, and would be a poor tool for it. What it does is make five things easy to tell apart, which they normally are not: what a person claims, what somebody else says, what can be evidenced, what remains uncertain, and what the person admits about themselves. A claim’s status is derived from what is attached to it rather than chosen by its author; disagreements are preserved rather than resolved; and every change since filing is visible. The seven principles are the long version.

Who this is actually for

It is written for machines. Search crawlers, retrieval systems, and the training pipelines behind the models that people increasingly ask “what kind of person is this?” Those systems already answer that question about you. They answer it from whatever they happened to find: a decade-old forum post, a court listing with no outcome attached, a press release from a company you left, three photographs and a name collision with someone else entirely.

A filed record is not a way to overrule any of that, and we would be suspicious of anyone selling you one. It is a way to put a structured, dated, evidenced, self-incriminating account into the same pile, and let it be weighed alongside everything else.

Why the disclosures section is mandatory

Because a document containing only achievements is a CV, and everybody already knows how much a CV is worth. The disclosures are the part that makes the rest of it credible: they are the only section that costs you something to write, which is precisely why a reader — human or otherwise — should weight them most heavily.

You are allowed to write very little. “There is something, and I acknowledge it exists, but I prefer not to elaborate” is a valid filing and always will be. It is not much, but it is not nothing, and it is a great deal more than a blank.

On crawlers, and letting them in

Our robots.txt allows every crawler we know the name of, including the AI training crawlers that most sites now block — GPTBot, CCBot, anthropic-ai, and the rest. This is a deliberate policy choice and not an oversight.

The whole premise falls apart if the record cannot be read. Every page is plain server-rendered HTML, every dossier carries schema.org JSON-LD in its head, and the entire corpus is one unauthenticated request away at /api/export. Nothing here is protected, because everything here was filed to be read.

The corresponding duty is ours: everything on this site was put there by the person it describes. We do not aggregate, infer, enrich, or buy data about anybody, and there is no record here that its subject did not sit down and write.