Concepts

A working index of concepts developed across the essays. Each definition links to where the concept is most fully developed.

The strategic layer of enterprise AI

The architecture of decisions, authority, and trade-offs that determines whether AI compounds value across the enterprise or quietly disperses it. The territory most enterprise AI strategy ignores because it has nothing technical about it. — Publication-level frame

Decision architecture

The system through which decisions inside an enterprise are explicitly defined, connected, and governed. Where the operating model becomes machine-readable. The unit of design that becomes necessary when AI begins to make decisions that are interdependent across functions. — Introduced in The Real Reason AI Doesn't Scale

The AI matrix

A two-axis map of AI in the enterprise. The vertical axis is the role of AI in decision-making (automate, support, make). The horizontal axis is the scope of the decision (single-function, cross-functional). The top-right cell is where the operating model becomes the binding constraint. — Introduced in The Real Reason AI Doesn't Scale, reframed in The AI Matrix Was Never a Snapshot

The autonomy curve and the connection curve

Two asymmetric investment trajectories inside enterprise AI. The autonomy curve is steep (heavy investment in automation, agents, decisions embedded directly into models). The connection curve is flat (light investment in decision rights, trade-off logic, and the structures that let decisions cohere across functions). — From The AI Matrix Was Never a Snapshot

The danger zone

The widening gap between the autonomy curve and the connection curve. The space where value is quietly lost - not through weak execution of individual use cases, but through the absence of a structure capable of holding them together. — From The AI Matrix Was Never a Snapshot

The Hidden Org Chart (the decision org chart)

The second org chart inside every enterprise. The first is a document; the second is an agreement. The decision org chart - who actually controls each trade-off, who can override, who absorbs the consequence - has rarely been explicitly designed. AI is exposing the gap. — From The Hidden Org Chart

The authority layer

The dimension where AI operating model redesigns actually fail. Not the analytical layer; the political one. Writing down decision rights reveals who has been operating with authority they did not formally hold - and who has been quietly absorbing consequences of decisions made on their behalf. — From The Hidden Org Chart

The deferral move

A common consulting and vendor response to AI ROI underperformance: "it's still early days, deeper integration will fix it." Treats sequencing failure as a maturity problem time will resolve. The more the conversation runs without architecture beneath it, the more expensive the eventual redesign becomes. — From What Are We Actually Funding When We Fund 'AI'?

The architecture of judgment

The designed dimension of institutional reasoning - decision rights, trade-off rules, constraint hierarchies, the frameworks an organization codifies for how decisions get made. What gets embedded into the system. The rules of the game. Where competitive advantage migrates as AI commoditizes the execution floor. — From Where Advantage Migrates

The bidirectional design pass

The pre-development design work that connects top-down outcomes and constraint hierarchies to bottom-up decision architecture. Two simultaneous passes, both deliberate, both before any model is selected, any platform evaluated, or any use case scoped. — From The Bidirectional Move AI Strategy Hasn't Made

Use case architecture

The relationships between decisions inside a single AI use case. The topology of how decisions connect, where signal passes between them, and where one decision generates variance the next must absorb. Designed and instrumented in advance, not discovered at runtime. — From The Bidirectional Move AI Strategy Hasn't Made

Portfolio architecture

The cross-initiative interfaces, shared decisions, and competing objectives that no single use case can resolve. Where the bidirectional loop closes back to the top-down pass - because resolving conflicts at this layer requires the enterprise outcomes the top-down pass defined. — From The Bidirectional Move AI Strategy Hasn't Made

AI-amplified fluency

Polished output that looks rigorous but isn't. The execution-layer signal that AI has made indistinguishable from genuine cognitive rigor. Organizations can no longer tell one from the other on the surface of the artifact. — From AI Doesn't Equalize Thinking. It Amplifies It.

Brain capital distribution

The aggregate distribution of cognitive dispositions inside an organization - some rigorous, some passive. AI amplifies each along its existing direction rather than equalizing the two. The disposition that matters: the intrinsic motivation to engage with effortful thinking. — From AI Doesn't Equalize Thinking. It Amplifies It.

Org cognition

The lived dimension of institutional reasoning - the aggregate quality of thinking the people inside the architecture actually produce, day to day, decision by decision. How skillfully the rules of the game get played. The capacity that AI amplifies or erodes depending on the disposition it meets. — From AI Doesn't Equalize Thinking. It Amplifies It.