neur.onl

A personal neural substrate, built in the open.

One person. One persistent, self-improving knowledge loop. Every session ends smarter than it started.

What this is

Most AI usage is amnesiac: every conversation starts from zero, and everything learned evaporates when the window closes. This site documents a different approach — a single operator running a durable memory and feedback loop where mistakes become fixes, fixes become rules, and rules compound.

Over time that loop has accumulated 13,000+ learnings — error signatures, failed approaches, verified procedures, hard-won judgment — that get surfaced back into new work automatically. It’s not a product. It’s a practice, and it’s being written up here as it evolves.

What’s coming

What the system learned today coming

A daily digest of the loop in action: the day’s most interesting new learnings, anonymized and annotated — what broke, what got captured, and how it changes tomorrow.

Field notes coming

Essays on the mechanics: how learnings get captured, ranked, and re-injected; what compounds and what rots; where the loop still fails.

Follow along

Built in the open, more soon. If a compounding personal knowledge loop is your kind of rabbit hole, leave an email — updates only, no noise.