People read content first; machines operate backstage

Cases, judgments, methods, principles and evidence matter, but they are the system's internal organisation—not most people's starting point. Requiring users to learn the taxonomy transfers the builder's work to the reader.

A usable knowledge system lets someone enter through the transaction, funding, integration, finance or AI problem already in front of them, then understand the deeper framework while solving it.

A more natural path is to understand one framework or begin with a real question, invoke relevant cases, judgments and methods, and then build personal system capability through action and outcomes.

This does not reject taxonomy. It changes its position: classification stays backstage for consistency, while problems and tasks move to the front so people can begin.

AI can find relationships across many objects, but without clear problem definitions, boundaries and accountability it merely produces plausible answers faster.

A good entrance helps people express the problem and helps AI identify the right knowledge objects. Human interface and machine structure do not conflict; they carry different responsibilities.

Open OS should not merely display Howard's existing views. It should help users move through understanding frameworks, solving problems and building capability.

The measure changes accordingly: not the number of pages or categories, but whether people enter real problems faster, whether judgment changes action and whether outcomes revise the system.