Trails
Essays, field notes, and dispatches from inside a knowledge engine that compiles instead of retrieves.
How Trail Works
All How Trail Works →Does a Trail Get Smarter as It Grows?
More sources should mean a better knowledge base. In practice, every system hits a point where adding material starts to subtract. Here is what compounds, what buckles, and where Trail's architecture deliberately bends the curve.
Work That Fits in a Night: How Trail Stops Caring About Corpus Size
A compile-time knowledge system has to integrate new sources against everything it already knows. That integration cost grows with the corpus — until, one Tuesday, it stops finishing before the next night begins. Here is how Trail makes that cost bounded without going blind to the long tail.
Three Filters on the Gate: How Trail's Curation Policy Borrows from the Structure of Your Mind
A knowledge system without filters isn't a brain — it's a garbage heap. Trail's trusted-pipeline, confidence-threshold, and no-contradictions trio isn't an arbitrary engineering choice. It mirrors a pattern two billion years of evolution converged on.
Compile-Time Knowledge: A Technical Argument Against RAG
RAG is a search engine wearing a language model as a hat. Trail compiles knowledge the way a brain consolidates memory — and the architectural difference shows up in latency, accuracy, provenance, and cost.
How a Brain Actually Remembers: Why Trail Compiles Knowledge Instead of Searching For It
RAG treats knowledge like a filing cabinet. Brains do something fundamentally different — and so does trail. The architecture matters more than the model.
Why Your Knowledge Base Should Compile, Not Search
Most AI knowledge systems are sophisticated search engines that forget between queries. The organizations winning are building something different — and the architectural choice has direct consequences for cost, accuracy, and defensibility.
Research
All Research →Niklas Luhmann's Zettelkasten Has Been Waiting Seventy Years for the LLM
Luhmann produced 90,000 cards and 500-plus books with a method that has worked for seven decades — but most of his time went into bookkeeping that does not directly contribute to thinking. LLM Neuron trails are the first plausible automation of exactly that bookkeeping. Trail is what it looks like as hosted infrastructure.
Scaling Trail from 200 to 100,000 Neurons: An Engineering Note
A compile-time knowledge engine has three workloads that fight for compute — ingest, lint, and the curation queue. Each one breaks at a different corpus size, for different reasons, and requires a different fix. This is the long-form companion to Work That Fits in a Night: the full accounting of where Trail's bottlenecks are, when they hit, and what a deployment looks like at 200, 8,000, 25,000, and 100,000 Neurons.
Trail and NotebookLM: Same Ethos, Different Artifact
NotebookLM proved source-grounded ingest-time tools can ship to millions. Trail shares that ethos and diverges on one question: what the system leaves behind.