Complete AI Training
Sign inGet my AI kit

Your job's AI kit

Get your AI kit

Tell us who you are and what you do. We show you your kit right away and email you the link: skills, prompts, AI agents, MCP servers and courses for your job.

500+ jobs ready, and we make a kit for any other job. No payment needed to look.

Share

Video course · 15 chapters · 514 min · certificate

AI Safety – Full Course by Safe.AI Founder on Machine Learning & Ethics (Center for AI Safety)

Dan Hendrycks' ML safety course from the Center for AI Safety: deep learning review, risk decomposition and accident models, black swans, adversarial and long-tail robustness, anomaly detection, interpretable uncertainty, transparency and trojans, emergent behaviour and honesty, machine ethics, ML for decisions and cyberdefence, cooperative AI, existential risk, AI and evolution, and the safety-capabilities balance.

Become a memberAlready a member? Sign in

What you'll learn

  • Apply hazard analysis to ML systems
  • Explain robustness to adversaries and black swans
  • Detect anomalies and calibrate uncertainty
  • Understand transparency, trojans and emergent behaviour
  • Discuss machine ethics and cooperative AI
  • Reason about existential risk and safety strategy

Chapters

15 chapters · 514:20
  1. 11:09 01Intro Members Introduction Why ML safety.
  2. 41:32 02Review Members Deep learning review Foundations.
  3. 46:41 03Hazards Members Risk decomposition and accident models Hazard analysis.
  4. 19:23 04Tail risk Members Black swans Extreme events.
  5. 30:55 05Adversarial Members Adversarial robustness Attacks.
  6. 23:16 06Long tail Members Black swan robustness Rare inputs.
  7. 42:36 07Anomalies Members Anomaly detection Unknown unknowns.
  8. 47:20 08Monitoring Members Uncertainty, transparency and trojans Know what models know.
  9. 37:14 09Honesty Members Emergent behaviour and honest models Capability spikes.
  10. 52:02 10Ethics Members Machine ethics Ethical behaviour.
  11. 32:52 11Applications Members ML for decisions and cyberdefence Systemic safety.
  12. 33:33 12Cooperation Members Cooperative AI Emerging area.
  13. 14:43 13X-risk Members Existential risk overview Possible hazards.
  14. 49:52 14Evolution Members AI and evolution Selfish agents.
  15. 31:12 15Strategy Members Safety-capabilities balance and conclusion Steer safely.