This repository contains simulation models and evaluation tools from the ANGEL Project — a biologically grounded initiative for AI alignment and human-centered safety. The focus here is on evaluating systemic risk, stress propagation, and survivability metrics within economic and technological systems.
We use machine learning tools to simulate alignment risk in real-world systems such as:
- Pharmaceutical industry transitions
- Profit vs. sustainability dynamics
- Species-level survivability modeling
The goal is to demonstrate how economic decisions affect collective well-being and how aligned AI systems can help redirect outcomes.
- Python 3.10
- Jupyter Notebooks
- Pandas, NumPy, Matplotlib
- TruLens for response evaluation
- GitHub for version control and collaboration
/pharma/ → Industry transition models and datasets
├── pharma_assets_table.csv
└── pharma_transition_notebook.ipynb
README.md → Project overview
requirements.txt → Package dependencies
.env.example → Template for environment variables
- Transitioning pharma from profit-centric to regenerative medicine
- Economic stress propagation through supply chains
- Dissonance in public health outcomes under misaligned incentives
Install dependencies and run the notebooks locally:
pip install -r requirements.txt
jupyter notebook
If you're using API keys or private configuration, duplicate the .env.example file and rename it to .env. Then insert your own keys as needed.
Prepared by: Robin Macomber
GitHub: Oberon245