Agency First AI (AFAI) — Core Repository

Welcome to the central hub for Agency First AI, a framework and demonstration for prioritizing human agency in AI systems.

---

Libertas Infinitum

---

About

Agency First AI (AFAI) is a design paradigm that ensures AI systems preserve, measure, and protect human agency above all else. This repository provides the tools and documentation to understand, implement, and audit AI systems using the AFAI framework.

This repo contains:

• The AFAI Manifesto: Outlines the core concepts, metrics, and operational principles.
• The afai_lib Python Library: A library for calculating and auditing the Agency Preservation Score (APS).
• Domain Adapters: Examples of how to apply the AFAI framework to specific domains, such as recommendation systems.
• A Live Demo: A web-based demonstration of the AFAI framework in action.
• Supporting Assets: Diagrams, documentation, and other resources.

---

Website

This project has a companion website built with GitHub Pages.

Live Website: https://AgencyFirstAI.github.io/afai-manifesto/

How to Set Up the Website

To enable the GitHub Pages website for your fork of this repository, follow these steps:

1. Go to your repository's Settings page.
2. In the "Code and automation" section of the sidebar, click on Pages.
3. Under "Build and deployment", for the Source, select Deploy from a branch.
4. Under "Branch", select main (or your default branch) and choose the /docs folder.
5. Click Save.

Your website should be live at the URL provided above within a few minutes.

---

Getting Started

To explore the simulation and develop locally, you will need Python 3.7+.

1. Clone the repository:

   git clone https://github.com/AgencyFirstAI/afai-manifesto.git
   cd afai-manifesto

2. Set up a virtual environment:

   python -m venv .venv
   source .venv/bin/activate  # On Windows, use `.venv\Scripts\activate`

3. Install dependencies:

   # (No external dependencies are currently required)

4. Run the demo:

   python demo/run_demo.py

---

Contributing

We welcome contributions of all kinds, from documentation improvements to new domain adapters. All contributions are subject to our Code of Conduct and Contribution Guidelines.

Please follow the standard workflow:

1. Fork the repository and create a new feature branch.
2. Make your changes, adhering to PEP 8 style guidelines.
3. Add or update tests to validate your changes.
4. Submit a pull request with a clear description of the "why" behind your contribution.

Changes to the core APS metric are subject to a more rigorous public review process outlined in CONTRIBUTING.md.

---

License

This project is licensed under the MIT License.