AI Safety at UCLA Intro Fellowship: Governance Track
Table of Contents
Part 1: Introduction to AI Safety
- Week 1: Artificial Intelligence — How it Works and What it Can Achieve
- Week 2: Loss of Control
- Week 3: Concentration of Power & Gradual Disempowerment
Part 2: Introduction to AI Governance
- Week 4: Domestic AI Governance
- Week 5: International AI Governance
- Week 6: Technical AI Governance & Careers in AI Governance
Part 1: Introduction to AI Safety
Week 1: Artificial Intelligence — How it Works and What it Can Achieve
Core Content (120 min):
- Andrej Karpathy - Intro to Large Language Models (60 min)
- Machines of Loving Grace (60 min)
Optional Additional Content:
AI History
- Visualizing the Deep Learning Revolution by Richard Ngo
- Nvidia: The chip maker that became an AI superpower
- Compute Trends Across Three Eras of Machine Learning
How AI Works
- AI, Machine Learning, and Deep Learning
- Gradient Descent: How Neural Networks Learn | Chapter 2, Deep Learning
- What is Self Supervised Learning?
AI Futures
- The economic potential of generative AI: The next productivity frontier
- Forecasting Transformative AI, Part 1: What Kind of AI?
- The Transformative Potential of AGI – and When It Might Arrive by Shane Legg and Chris Anderson
Week 2: Loss of Control
Core Content (120 min):
- AGI Safety from First Principles (60 min)
- AI 2027, race ending (60 min)
Optional Additional Content:
First Principles AI Safety
- Existential Risk from Power-Seeking AI
- Why Would AI Want to do Bad Things? Instrumental Convergence
- Why AI Alignment Could Be Hard with Modern Deep Learning by Ajeya Cotra
Surveys of AI Risks
Concrete Scenarios
- What Failure Looks Like by Paul Christiano
- Auto-GPT and AI Race Acceleration by The AI Beat
- The True Story of How GPT-2 Became Maximally Lewd
Week 3: Concentration of Power and Gradual Disempowerment
Core Content (120 min):
- AI Enabled Coups: How a Small Group Could Use AI to Seize Power (60 min)
- Gradual Disempowerment: Systemic Existential Risks from Incremental AI Development (60 min)
Optional Additional Content:
- How an AI company CEO could quietly take over the world
- Extreme power concentration
- Gradual Disempowerment
Part 2: Introduction to AI Governance
Week 4: Domestic AI Governance
Core Content (100 min):
- AI Safety, Ethics, and Society: Sections 8.1 (Governance), 8.3 (Distribution), 8.4 (Corporate Governance), and 8.5 (National Governance) (45 min)
- Frontier AI Regulation: Managing Emerging Risks to Public Safety: Executive Summary, Sections 3 and 4 (45 min)
- California’s SB 53: The First Frontier AI Law, Explained (10 min)
Optional Additional Content:
- The AI Triad and What It Means for National Security Strategy
- The Policy Playbook: Building a Systems-Oriented Approach to Technology and National Security Policy
- Primer on Safety Standards and Regulations for Industrial-Scale AI Development
Week 5: International AI Governance
Core Content (120 min):
- Historical case studies of technology governance and international agreements (20 min)
- AI Safety, Ethics, and Society: Section 8.6 (International Governance) (20 min)
- Analysis of Global AI Governance Strategies (40 min)
- High-Level Summary of the EU AI Act (10 min)
Optional Additional Content:
- Artificial Intelligence Index Report 2025 section 6.1 - pgs 326-335
- Strengthening Resilience to AI Risk: A Guide for UK Policymakers
- China’s AI Regulations and How They Get Made
- International Institutions for Advanced AI
- The Bletchley Declaration
- OECD AI Principles
Week 6: Technical AI Governance and Careers in AI Governance
Core Content (100 min):
- Computing Power and the Governance of AI (10 min)
- Open Problems in Technical AI Governance: Sections 1, 2, and one section from 3-8 (40 min)
- AI governance and policy | Career review (50 min)
Optional Additional Content:
Technical AI Governance
- Choking Off China’s Access to the Future of AI
- Primer on AI Chips and AI Governance
- An International Agreement to Prevent the Premature Creation of Artificial Superintelligence
Careers in AI Governance