Short pieces on what is happening in the field — the news, the patterns, the failures — and what they mean for knowledge as a first-class layer. Each piece is written to be read in a couple of screens.
These are not research papers and they are not marketing copy. They are short pieces — usually two screens, two pages at most — on what I am seeing in the field and what I think it means for knowledge. They are written in my voice, on my view of the latest news, and published in chronological order so the oldest thinking is at the bottom and the newest is at the top.
A look at the most common way agentic projects go wrong: not the model, but the knowledge layer — stale procedure, unowned data, no deprecation, weak grounding.
Why packing a context window as one big text dump is the cheapest way to make an agent both more expensive and less dependable, and what to do instead.
The same discipline that makes an agent cheaper — sending less of the right knowledge — also makes it more dependable. Grounding and savings are not two goals; they are one discipline seen from two sides.
Enterprise agents fail less from not knowing enough than from acting on the wrong slice of what is known. A practical look at what "good enough knowledge" means at organizational scale.
More pieces are added as the field moves. The archive is chronological: newest first. If you want to write one, the form is simple — a short opinionated take on something in the news, in two screens or less.