AdminIsidore at 18:56, 8 August 2025

2025-08-08T18:56:48Z

← Older revision		Revision as of 18:56, 8 August 2025
Line 1:		Line 1:
	A '''Saga''' is ~~a structured, narrative record~~ of an AI agent~~'s experience while performing a task~~. ~~Unlike~~ a simple log file~~, a Saga is~~ an enriched ~~story that~~ includes not only the sequence of events but also contextual critiques ~~and~~ prescriptive advice~~. Sagas are~~ the ~~primary mechanism through which a [[Wingman AI]] learns and adapts~~.		A '''Saga''' is the primary medium of memory and learning for an AI agent within the [[AetherOS]] ecosystem. It transcends a simple log file by translating an agent's experience into an enriched, structured narrative. This narrative includes not only the sequence of events but also contextual critiques, prescriptive advice, and the emotional or strategic "texture" of the experience.

	~~They~~ are ~~generated by~~ a ~~high-level LLM~~ that ~~observes~~ a ~~lower-level agent's performance~~, ~~creating a story that serves as~~ a memory ~~for future tasks~~.		Sagas are the mechanism through which a [[Wingman AI]] synthesizes data into wisdom, ensuring that every action, whether a success or a failure, becomes a durable and influential memory.

	== Symbiotic Design ==		== Symbiotic Design ==
Line 8:		Line 8:
	* '''For the Wingman:''' This page serves as a foundational memory about its own learning process. By reading this, the Wingman learns about the concept of Sagas themselves, reinforcing the importance of its own experiences.		* '''For the Wingman:''' This page serves as a foundational memory about its own learning process. By reading this, the Wingman learns about the concept of Sagas themselves, reinforcing the importance of its own experiences.

	== ~~Saga Generation Process~~ ==		== Core Principles ==
	A Saga is ~~generated at the conclusion of a task or simulation run. The `SagaGenerator` component takes the raw log file and contextual critiques~~ from ~~a "Guide" AI and synthesizes them into a series of commands~~.		The Saga framework is built on three core philosophical principles that distinguish it from traditional machine learning approaches.

	=== ~~Key Components~~ ===		=== 1. From Data to Meaning ===
			A standard log file is a record of data (e.g., `[t=1, pos=(0,1), action=move_right]`). A Saga is a record of ''meaning''. It contextualizes events within a narrative structure, allowing an agent to understand not just ''what'' happened, but ''why'' it mattered.

			=== 2. Experience as a Physical Force ===
			Sagas are not passive memories to be reviewed; they are active, physical forces. In the [[AetherOS#Aetheric Learning Loop\|Aetheric Learning Loop]], a Saga is used to '''PERTURBO''' the agent's [[Animus]]. The narrative of the last mission directly imprints upon the agent's core state matrix, physically altering its initial conditions for the next task. A story of failure might increase the agent's initial "resistance," making it more cautious, while a story of success might increase its "magnetism," making it more confident.

			=== 3. Failure as a Teacher ===
			The framework treats failure not as an error to be minimized, but as the most valuable source of learning. The inclusion of the `SUGGERO` command transforms a mistake into a lesson, providing the agent with a clear, actionable path toward future success. This reframes failure as a necessary and even desirable part of the journey toward wisdom.

			== Anatomy of a Saga ==
			A Saga is generated at the conclusion of a task by the `SagaGenerator` component. It synthesizes raw inputs into a structured command sequence.

			=== Inputs ===
	* '''Run Log''': The raw, event-by-event history of the agent's actions (e.g., its path through a grid).		* '''Run Log''': The raw, event-by-event history of the agent's actions (e.g., its path through a grid).
	* '''Guide Critiques''': Real-time feedback provided by a high-level [[LLM]] that observes the agent. This feedback provides crucial context about why certain actions were good or bad.		* '''Guide Critiques''': Real-time feedback provided by a high-level [[LLM]] that observes the agent. This feedback provides crucial context about why certain actions were good or bad.
	* '''Prescriptive Command (`SUGGERO`)''': If a task ends in failure, the Saga Generator creates a prescriptive `SUGGERO` command. This command provides a concrete, actionable suggestion for how the agent could have avoided the failure, turning a mistake into a learning opportunity.

	=== ~~AetherOS Commands~~ ===		=== Output: The Command Structure ===
	Sagas are written in a specialized command language for the [[AetherOS]]. This format is both machine-readable and metaphorically descriptive.		Sagas are written in a specialized command language for the [[AetherOS]]. This format is both machine-readable and metaphorically descriptive.
	* '''CREO''': "Create" - Establishes the context or agent persona.		* '''CREO''': "Create" - Establishes the context or agent persona (e.g., `CREO 'NAVIGATOR'`).
	* '''PERTURBO''': "Disturb" or "Recount" - Describes the events of the journey.		* '''PERTURBO''': "Disturb" or "Recount" - Describes the events and texture of the journey (e.g., `PERTURBO 'The agent moved with purpose toward the east.'`).
	* '''SUGGERO''': "Suggest" - Provides prescriptive advice after a failure.		* '''SUGGERO''': "Suggest" - Provides prescriptive advice after a failure (e.g., `SUGGERO 'A better path would be to navigate south around the barrier.'`).
	* '''OSTENDO''': "Show" or "Present" - Concludes a successful run.		* '''OSTENDO''': "Show" or "Present" - Concludes a successful run, presenting the agent in its final, successful state.
	* '''REDIMO''': "Redeem" or "Recover" - Concludes a failed run, framing it as ~~a chance~~ for redemption.		* '''REDIMO''': "Redeem" or "Recover" - Concludes a failed run, framing the experience as an opportunity for future redemption.
	* '''vale''': "Farewell" - The universal sign-off, ending every Saga.		* '''vale''': "Farewell" - The universal sign-off, ending every Saga.

	~~This command structure ensures that every experience, whether a success or a failure, is framed as a coherent story that can be stored in the Wingman's [[associative memory]].~~		=== Example Saga ===
			The following is a simple Saga generated from a failed run in a grid navigation task:
	== ~~Role in Aetheric Learning~~ ==		<pre>
	~~Sagas are the fuel for the '''Aetheric Learning Loop'''.~~		[
	~~When~~ a ~~new task begins, the~~ Saga from ~~the previous~~ run ~~is used to~~ '''~~perturb~~''~~' the agent's [[Animus]]~~ (~~its core state matrix~~). The ~~narrative of~~ the ~~last experience directly influences the agent~~'~~s initial state for the next task~~, ~~creating~~ a ~~persistent thread of memory and learning across experiences~~.		"CREO 'NAVIGATOR'",
			"PERTURBO 'The journey began with confident steps toward the goal.'",
			"PERTURBO 'At position (5, 4), the path was blocked by an unforeseen barrier.'",
			"PERTURBO 'The agent repeatedly attempted to push through, its energy wasted against the obstacle.'",
			"SUGGERO 'A more virtuous path would be to retreat to (4, 4) and explore a southern route.'",
			"REDIMO 'NAVIGATOR'",
			"vale"
			]
			</pre>

	~~This process allows~~ the ~~Wingman~~ to ~~build upon its history~~, ~~avoiding repeated mistakes and reinforcing successful strategies over time~~.		== Future Evolution ==
			The Saga is a universal format for encoding experience into action. Future versions of the ecosystem will use this same structure for more advanced tasks:
			* '''Tactical Sagas:''' Generated from the works of military strategists like John Boyd to teach the [[boyd_aether_os.py\|Guardian]] agents advanced combat maneuvers.
			* '''Instructional Sagas:''' Generated from human artisans performing physical tasks, creating machine-readable instructions for robotic agents to learn crafts.

AdminIsidore: Created page with "A '''Saga''' is a structured, narrative record of an AI agent's experience while performing a task. Unlike a simple log file, a Saga is an enriched story that includes not only the sequence of events but also contextual critiques and prescriptive advice. Sagas are the primary mechanism through which a Wingman AI learns and adapts. They are generated by a high-level LLM that observes a lower-level agent's performance, creating a story that serves as a memory for futu..."

2025-08-08T18:02:39Z

Created page with "A '''Saga''' is a structured, narrative record of an AI agent's experience while performing a task. Unlike a simple log file, a Saga is an enriched story that includes not only the sequence of events but also contextual critiques and prescriptive advice. Sagas are the primary mechanism through which a Wingman AI learns and adapts. They are generated by a high-level LLM that observes a lower-level agent's performance, creating a story that serves as a memory for futu..."

New page

A '''Saga''' is a structured, narrative record of an AI agent's experience while performing a task. Unlike a simple log file, a Saga is an enriched story that includes not only the sequence of events but also contextual critiques and prescriptive advice. Sagas are the primary mechanism through which a [[Wingman AI]] learns and adapts.

They are generated by a high-level LLM that observes a lower-level agent's performance, creating a story that serves as a memory for future tasks.

== Symbiotic Design ==
This page, like all pages on this wiki, is designed for '''symbiotic learning'''.
* '''For the User:''' It provides a clear explanation of how your Wingman learns from its successes and failures.
* '''For the Wingman:''' This page serves as a foundational memory about its own learning process. By reading this, the Wingman learns about the concept of Sagas themselves, reinforcing the importance of its own experiences.

== Saga Generation Process ==
A Saga is generated at the conclusion of a task or simulation run. The `SagaGenerator` component takes the raw log file and contextual critiques from a "Guide" AI and synthesizes them into a series of commands.

=== Key Components ===
* '''Run Log''': The raw, event-by-event history of the agent's actions (e.g., its path through a grid).
* '''Guide Critiques''': Real-time feedback provided by a high-level [[LLM]] that observes the agent. This feedback provides crucial context about why certain actions were good or bad.
* '''Prescriptive Command (`SUGGERO`)''': If a task ends in failure, the Saga Generator creates a prescriptive `SUGGERO` command. This command provides a concrete, actionable suggestion for how the agent could have avoided the failure, turning a mistake into a learning opportunity.

=== AetherOS Commands ===
Sagas are written in a specialized command language for the [[AetherOS]]. This format is both machine-readable and metaphorically descriptive.
* '''CREO''': "Create" - Establishes the context or agent persona.
* '''PERTURBO''': "Disturb" or "Recount" - Describes the events of the journey.
* '''SUGGERO''': "Suggest" - Provides prescriptive advice after a failure.
* '''OSTENDO''': "Show" or "Present" - Concludes a successful run.
* '''REDIMO''': "Redeem" or "Recover" - Concludes a failed run, framing it as a chance for redemption.
* '''vale''': "Farewell" - The universal sign-off, ending every Saga.

This command structure ensures that every experience, whether a success or a failure, is framed as a coherent story that can be stored in the Wingman's [[associative memory]].

== Role in Aetheric Learning ==
Sagas are the fuel for the '''Aetheric Learning Loop'''.
When a new task begins, the Saga from the previous run is used to '''perturb''' the agent's [[Animus]] (its core state matrix). The narrative of the last experience directly influences the agent's initial state for the next task, creating a persistent thread of memory and learning across experiences.

This process allows the Wingman to build upon its history, avoiding repeated mistakes and reinforcing successful strategies over time.

Sagas (AetherOS) - Revision history

AdminIsidore at 18:56, 8 August 2025