A place to read our work.
We opened Architecturalism, where we keep our writing: research on how AI systems work, case studies of the systems we build, and the occasional note like this one.
We put up a new section of the site, Architecturalism. It is where our writing lives: research on how AI systems actually work, case studies of what we build, and short notes like this one.
Two pieces are live to start. The first is a short case for why the architecture, not the model, is where the intelligence lives. The second is a case study on how a custom architecture holds a clinical-grade standard for Longevity Science Daily: sourced, reproducible, and fail-closed.
More research is coming. Short pieces, evidence-based, written to be read by anyone, not only engineers. The same care good science writing takes, applied to AI.
We are not asking you to subscribe. There is no list. If something here is useful, or you have a system that keeps breaking, that is the reason to get in touch.