Ontological Primitives
Entities
The irreducible objects. Every AI-business configuration is a composition of these entities and the interfaces that connect them.
Business
An organization with boundaries, processes, inputs, outputs, and external interfaces. The container.
Human
A node that reasons, decides, acts, and bears accountability. Can interface with any other entity and the external world.
Model
The foundational capability. General-purpose reasoning, broad knowledge, and adaptability. Invoked on demand — the engine that powers everything above it.
Agent
A model situated within the business — given domain knowledge, tools, persistent state, and a role. Its degree of autonomy is not inherent; it emerges from where the agent is positioned.
Trait Definitions
What the Tags Mean
Each entity carries a set of defining traits. Click any tag above to jump here.
Business
bounded
—
Has defined edges — legal, contractual, operational. An inside and an outside.
has interfaces
—
Connects to the external world through specific touchpoints: customers, vendors, regulators.
has inputs / outputs
—
Consumes resources and produces value — revenue, products, services, data.
Human
adaptive
—
Can learn, reframe, improvise. Changes approach based on novel context.
accountable
—
Bears responsibility for decisions and outcomes. Can be held liable.
goal-directed
—
Acts toward objectives with intention and judgment, not just execution.
Model
general-purpose
—
Broad reasoning capability, not specialized to any single domain or task.
on-demand
—
Invoked when needed, returns a result, has no persistent presence or state.
foundational
—
The base capability layer — everything above (agents, workflows) is built on it.
Agent
situated
—
Placed within the business — given a role, a position in the network, a context.
stateful
—
Maintains persistent context across interactions — memory, history, accumulated knowledge.
domain-aware
—
Equipped with specific domain knowledge — processes, terminology, constraints of its field.
role-bound
—
Operates within a defined scope — responsibilities, permissions, and boundaries.
Interface Taxonomy
Types of Interfaces
Interfaces are first-class ontological objects. The distribution of interface types determines the class.
H ↔ H
Human–Human
Delegation, collaboration, handoffs, reporting. The default wiring.
H ↔ A
Human–Agent
Direction, delegation, review, approval, escalation.
A ↔ A
Agent–Agent
Orchestration and handoff without human mediation.
H ↔ Ext
Human–External
Sales, support, negotiation. Human at the boundary of the firm.
A ↔ Ext
Agent–External
Automated transactions, conversations, API exchange. Agent at the boundary.
Gov → ✱
Governance
Oversight, constraint, audit. Principals bounding operators.
The Six Classes
From Traditional to Autonomous AI
Each class is a distinct structural relationship. The diagrams show the essential pattern — the entities, the interfaces, and their position relative to the business boundary.
Class 0
Traditional
"No AI in the loop"
An all-human network. People hold every role, every interface, every boundary touchpoint. This is the structure AI enters.
Remove AI → Nothing changes.
Class 1
AI-Equipped
"AI on the side"
The human network is unchanged. Individuals invoke AI products — ChatGPT, Claude, Copilot — on demand. Powerful, but external to the network. No interfaces change.
Remove AI → People slow down, nothing breaks.
Class 2
AI-Augmented
"AI in the team"
The model is situated — given a role, domain knowledge, tools, and persistent state. It becomes an agent: a participant in the network. The human directs it. The H↔A interface emerges. Autonomy is not yet required.
Remove AI → Capacity drops, humans can cover.
Class 3
AI-Extended
"AI at some doors"
The agent takes over some boundary interfaces. Because it now faces the outside world without a human mediating, autonomy emerges — the position demands it. The human retains other boundaries and remains the principal.
Remove AI → Some channels go dark, core survives.
Class 4
AI-Native
"AI runs the shop"
The agent holds all boundary interfaces and runs operations. Autonomy now extends inward — the agent directs workflows. The human is still inside the machine: managing, configuring, handling exceptions. They know how the work is done.
Remove AI → Operations halt. Humans can rebuild.
Class 5
Autonomous AI
"AI is the business"
The agent network and the business are coextensive. The agent network is self-coordinating. The human is outside the machine: setting strategy, defining constraints, auditing outcomes — but not managing or configuring.
Remove AI → The entity ceases to exist.
Native vs Autonomous — The Critical Distinction
AI-Native: Humans are inside the machine. They manage agents, handle exceptions, reconfigure workflows. Remove the AI and operations halt — but the people know how to do the work.
Autonomous AI: Humans are outside the machine. They set strategy and audit outcomes. The agent network is self-coordinating. Remove the AI and there's nothing left — no one knows how to do the work manually.
The test: "Could the humans in this company run it without the agents?"
Native → Yes, painfully. Autonomous → No.
In Conversation
"We're Traditional — no AI in the loop. Everything runs on people, process, and spreadsheets."
"Most enterprises are AI-Equipped — same org, same process, people just use copilots on the side."
"We became AI-Augmented when we gave agents their own tasks on the sprint board."
"We're AI-Extended — agents handle support and procurement, but sales is still all human."
"That startup is AI-Native — agents run everything, two engineers keep them tuned."
"A true Autonomous AI company doesn't need an AI strategy. It needs a governance strategy."
Key Distinctions
Ideas That Underpin the Framework
Model → Agent: Situation, Not Promotion
A Model is the foundational reasoning engine — general-purpose, powerful, invoked on demand. An Agent is what happens when you situate that model within a business: give it a role, persistent context, domain knowledge, and tools.
The boundary between Class 1 and Class 2 is this act of situation — the model goes from being an external utility to an embedded participant with a place in the network.
Autonomy is Emergent, Not Inherent
An agent doesn't start autonomous. In Class 2, it's directed — the human tells it what to do. Autonomy emerges from position: when an agent is placed at a boundary interface (Class 3), it must act independently because there's no human mediating that interface. The architecture forces the autonomy.
Class 2 — Agent is directed. Human says what to do.
Class 3 — Agent becomes autonomous at the boundary. It must decide on its own because it faces outward.
Class 4 — Agent is autonomous internally. Directs workflows, humans handle exceptions.
Class 5 — Agent network is self-coordinating. Humans govern, not manage.
A Note on Scheduled Autonomy
There is a category of agents that exhibit autonomy through scheduling — given a heartbeat, waking at intervals to poll for conditions, execute routines, or perform maintenance without human prompting. This is a real form of autonomous behavior.
However, this framework deliberately excludes polling or schedule-based autonomy. The classes here describe relational autonomy — autonomy that emerges from an agent's position in the organizational network — not temporal autonomy driven by a cron job or heartbeat. A scheduled agent can run inside any class from 2 upward. It doesn't change the class.
The progression tracks three shifts: what the AI is — from foundational engine, to situated participant, to operational core; where it sits — from outside, to inside, to coextensive with the boundary; and what emerges — autonomy is not designed in, it's demanded by position. Place an agent at a boundary and it must act independently. Make it the core and it must self-coordinate. The architecture creates the autonomy.
Prepared by Alex Collet & Akhil Aryanv9 · February 2026