Module 4 / Lesson 12

Management of x-Risk from AI

1 min read

Lesson 12: Management of x-risk from AI

Lesson 12 builds directly on the risk framings from Lesson 11, moving from why AI might be existentially risky to how we can measure and manage those risks. This lesson blends conceptual stress tests (like the AI 2027 scenario), governance debates, and practical approaches to x-risk management. It has three sub-lessons:

Lesson 12.1: Managing the AI Risk

Introduces different management strategies for AI existential risks.
Covers pause/delay strategies, alignment strategies, and information strategies.
Helps think critically about trade-offs in intervention and governance design.

Lesson 12.2: AI in 2027 (The AI 2027 Scenario)

Presents the AI 2027 thought experiment: what if AI becomes superhuman faster than expected?
Guides through the interactive scenario (AI-2027.com), focusing on branching points, assumptions, and outcomes.
Emphasizes critical thinking about acceleration, decision-making, and governance pressures.

Lesson 12.3: P-Doom Movie

- A comic Movie on P-Doom and Super Intelligence AI

Lesson 12.4: Debates & Policy

Frames debates around three perspectives:
- The Optimist (risks minimal, no intervention).
- The Worried Realist (risks high, but policies ineffective).
- The Worried Interventionist (risks high, policies necessary).

Key Takeaways

Risk perception influences management strategy: optimism vs. pessimism leads to very different policy choices.
Scenarios like AI 2027 are not predictions but stress tests that sharpen understanding of uncertainty.
Policy debates reflect not just values but also beliefs about feasibility, coordination, and global competition.
Measurement and management of AI risk are inseparable: what we believe about risk informs what we choose to do about it.

Your progress stays in this browser.