# 01. Fundamentals

## The Vision: Cognitive Autonomy

In a business context, AI maturity is not defined by how "smart" a foundational model is. It is defined by how much of the **cognitive load** has been structurally transferred from the human to the system.

Traditional maturity models measure adoption. The **Cognitive Autonomy Index (CAI)** treats AI as a system. The goal of the CAI is to measure the shift from human execution to system execution. True ROI is found not when humans work faster, but when humans stop "doing the machine's homework" entirely.

## Core Concepts: Human-to-System Transfer

As a workflow matures, the human's role fundamentally changes. **Crucially, the CAI does not assume that maximum autonomy is the goal for every process.** We account for the appropriate level of autonomy based on specific tools and compliance risks.

### The Three Human Roles

1. **The Creator (Low Autonomy):** The human generates the input, guides the logic, and formats the output. The AI is simply a conversational interface.
2. **The Reviewer (Medium Autonomy):** The system generates the output based on pre-defined context (like RAG or templates). The human acts as a QA check, legally or operationally signing off on the work.
3. **The Governor (High Autonomy):** The system executes the task end-to-end via an Actionable Harness. The human manages the system itself, handling exceptions and optimizing the overarching strategy rather than individual tasks.

## Glossary of Terms

* **Actionable Harness:** A technical integration (like an API, Webhook, or Model Context Protocol) that "hitches" an AI model to a specific business process.
* **Agent Mesh:** A coordinated network of specialized AI agents that trigger one another across platforms.
* **Cognitive Load:** The amount of mental effort and manual execution required by a human to complete a task.
* **Right-Leveling:** Aligning a tool's capabilities (L\_t) with human competency (S) and operational processes (L\_c) to achieve a CAI score of exactly 1.0.
* **RAG (Retrieval-Augmented Generation):** Providing AI with contextual memory by allowing it to search internal data.
* **Utility Ceiling (L\_t):** The maximum safe and technically possible level of autonomy for a specific tool.


---

# Agent Instructions: Querying This Documentation

If you need additional information that is not directly available in this page, you can query the documentation dynamically by asking a question.

Perform an HTTP GET request on the current page URL with the `ask` query parameter:

```
GET https://cai-framework.gitbook.io/cai-framework/01-fundamentals.md?ask=<question>
```

The question should be specific, self-contained, and written in natural language.
The response will contain a direct answer to the question and relevant excerpts and sources from the documentation.

Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.