AI Safety Fundamentals: Alignment

83 Episodes

1. Biological Anchors: A Trick That Might Or Might Not Work

   Published: 13/05/2023
2. Intelligence Explosion: Evidence and Import

   Published: 13/05/2023
3. On the Opportunities and Risks of Foundation Models

   Published: 13/05/2023
4. Visualizing the Deep Learning Revolution

   Published: 13/05/2023
5. Future ML Systems Will Be Qualitatively Different

   Published: 13/05/2023
6. A Short Introduction to Machine Learning

   Published: 13/05/2023
7. AGI Safety From First Principles

   Published: 13/05/2023
8. What Failure Looks Like

   Published: 13/05/2023
9. Specification Gaming: The Flip Side of AI Ingenuity

   Published: 13/05/2023
10. Deceptively Aligned Mesa-Optimizers: It’s Not Funny if I Have to Explain It

    Published: 13/05/2023
11. The Alignment Problem From a Deep Learning Perspective

    Published: 13/05/2023
12. The Easy Goal Inference Problem Is Still Hard

    Published: 13/05/2023
13. Learning From Human Preferences

    Published: 13/05/2023
14. Superintelligence: Instrumental Convergence

    Published: 13/05/2023
15. ML Systems Will Have Weird Failure Modes

    Published: 13/05/2023
16. Thought Experiments Provide a Third Anchor

    Published: 13/05/2023
17. Goal Misgeneralisation: Why Correct Specifications Aren’t Enough for Correct Goals

    Published: 13/05/2023
18. Is Power-Seeking AI an Existential Risk?

    Published: 13/05/2023
19. Where I Agree and Disagree with Eliezer

    Published: 13/05/2023
20. AGI Ruin: A List of Lethalities

    Published: 13/05/2023