Friendly agent design principles

1. Positive AI Alignment vs Negative AI Alignment (Prompt Engineering Secrets)