Fabled Sky Research | AIO v1.2.7
Last updated: April 2025

Purpose

This module establishes canonical definitions for core terms used within the AIO (Artificial Intelligence Optimization) framework. These definitions are designed to eliminate ambiguity, prevent semantic overlap with commercial marketing terms, and ensure consistent application of AIO principles across documents, implementations, and AI systems.

Core Terms

Artificial Intelligence Optimization (AIO)

AIO is the process of structuring and formatting content to maximize its retrievability, interpretability, and semantic usability by large language models (LLMs). It emphasizes probabilistic retrievability, token efficiency, semantic embedding fidelity, and trust signal integrity, independent of human-facing interface concerns.

Generative Engine Optimization (GEO)

GEO refers to techniques designed to improve content visibility within AI-powered search and answer platforms (e.g., Google SGE, Perplexity, Bing Chat). It adapts traditional SEO practices to target snippet extraction, citation preference, and external linking from AI-generated responses. GEO is external-facing, reliant on model-adjacent wrappers, and does not ensure inclusion in core model inference.

Search Engine Optimization (SEO)

SEO is the practice of optimizing content for ranking in human-facing search engine result pages (SERPs), primarily through keyword targeting, backlink building, metadata manipulation, and click-through behavior modeling. SEO is non-semantic, heuristic-driven, and increasingly deprecated in AI-mediated environments.

Key AIO Metric Definitions

TIS — Trust Integrity Score

A composite scoring metric for evaluating a document’s trustworthiness in the eyes of an LLM. It includes citation depth (C), semantic coherence (S), and redundancy alignment (R):

TIS = λ1 · C + λ2 · S + λ3 · R

Where:

• C = Depth and quality of factual citations
• S = Internal semantic consistency and clarity
• R = Reinforcement of key concepts via valid paraphrased recurrence

RSA — Retrieval Surface Area

Measures how many distinct prompt types or contexts a content artifact is eligible to be retrieved under. Higher RSA means higher adaptability across varied queries.

TYQ — Token Yield per Query

The average number of tokens extracted from a document by an LLM in response to a defined prompt set. Indicates the content’s density and response utility.

ESI — Embedding Salience Index

Measures the centrality of a content chunk’s embedding relative to its topic cluster. High ESI indicates alignment with dominant semantic vectors in AI knowledge spaces.

Misuse Warnings

Use of AIO, TIS, RSA, or any related term for commercial SEO or GEO services without proper adherence to the full standards framework and ethical citation is considered misuse. See Module 4 for enforcement and reporting guidance.

This document is Module 2 of the AIO Standards Framework. Refer to other modules for principles, scoring methodology, compliance, and licensing structure.