Institutional Memory in the Age of Models

The most valuable internal corpus is not a pile of documents. It is a living record of decisions, counterfactuals, and the assumptions that made a choice reasonable at the time.

The important question is not whether the system is changing, but which assumptions remain useful while it does.

The research archive is intentionally text-first. External references are linked rather than stored, preserving fast reads and modest database costs while keeping every source legible to the reader.

That constraint is not an aesthetic compromise. It is a discipline: preserve the argument, the date, the relevant model, and the source path.