AI Safety Fundamentals

I completed the course on AI Safety Fundamentals as part of a reading group at Imperial College London. Each week we worked through the reading leadings and discussed our thoughts. Below I have consolidated my notes on each of the readings.

Contents

• Week 1 - Artificial General Intelligence
• Week 2 - Goals and Misalignment
• Week 3 - Threat Models and Types of Solutions
• Week 4 - Learning from Humans
• Week 5 - Decomposing Tasks for Better Supervision
• Week 6 - Interpretability
• Week 7 - Agent Foundations, AI Governance