ae-whitepapers

This repository hosts two foundational texts for the AEGNTIC project:

• Computational Amplification Through Aegntic AI (computational-amplification-whitepaper.pdf) — a practical framework for orchestrating parallel, agentic development with measurable ROI.
  File:computational-amplification-whitepaper.pdf

• Aegntic Whitepaper Notice (Aegntic Whitepaper Notice.pdf) — the language/ontology rationale behind “AEGNT: Algorithms Evolving Generative Neural Thresholds.”
  File: aegntic-whitepaper-notice.pdf

---

TL;DR

• Core principle: “Computational Power = Engineering Success.” Intelligent orchestration converts raw compute into non-linear productivity gains.
• Amplification Stack (L1→L5): single agents → parallel agents → isolated parallel branches → automated best-of-breed selection → infinite learning loops.
• Three-Folder System: IDocs (persistent knowledge), Specs (planning/architecture), .cloud (reusable assets) to compound team learning.
• MCP servers: standardized action surfaces for Docker, GitHub/FS, databases/APIs, and browser automation—unlocking richer, verifiable agent behaviors.
• Safety & QA: hard limits for loops/parallelism, multi-criteria evaluation (correctness, performance, security, maintainability, innovation).
• Economic case: documented productivity/ROI deltas from agentic patterns and parallel worktrees.
• Ontology: “AEGNT” reframes so-called “AI” as evolving systems that generate new thresholds of understanding (not mere automation).

---

Papers

• Computational Amplification Through Aegntic AI
  ./computational-amplification-whitepaper.pdf computational-amplification-whitepaper.pdf

  Highlights: amplification equation; parallel/isolated branches via Git worktrees; infinite agent loops; MCP integration; safety rails; empirical ROI.

• Aegntic Whitepaper Notice
  ./aegntic-whitepaper-notice.pdf aegntic-whitepaper-notice.pdf

  Highlights: “AEGNT: Algorithms Evolving Generative Neural Thresholds,” language precision, and the motivation to replace “artificial intelligence” with a more accurate ontology.

---

Key Concepts

• “Computational Power = Engineering Success.” The central lever is orchestration—turning compute into exponential, not linear, output.
• Amplification Stack:
  1. Single agent → 2) Parallel agents → 3) Isolated parallel agents (Git worktrees) → 4) Intelligent selection → 5) Infinite learning loops.
• Three-Folder System: IDocs, Specs, .cloud to institutionalize learnings across projects and time.
• Model Context Protocol (MCP): consistent interfaces for tools (Docker, GitHub/FS), data, APIs, and the browser, enabling verifiable agent actions.
• Safety & Governance: resource ceilings for iterations/parallelism, evaluation pipelines, and human-in-the-loop review.
• AEGNT Ontology: Algorithms Evolving Generative Neural Thresholds—a linguistic correction and intent statement for how this field should be framed.

---

How to Cite

Computational Amplification Through Aegntic AI — Cooper, M. (2025). Version 1.0. CC-BY-SA 4.0.
Aegntic Whitepaper Notice — Cooper, M. (2025). Language/ontology context for “AEGNT.”

Note: The Computational Amplification paper is released under Creative Commons CC-BY-SA 4.0 (attribution + share-alike). Please attribute appropriately when quoting or remixing.

---

License

• Computational Amplification: CC-BY-SA 4.0 (see paper footer).
• Aegntic Whitepaper Notice: see the document for attribution details.

---

Contact

• Aegntic Foundation — contact details included within the PDFs.
• Author: Mattae Cooper (Lead AI Systems Integrity Researcher).

---

Related

• ae-co-system — AEGNTIC’s open-source ecosystem for agentic development (docs/orchestration/MCP)