Our graduates work at
Unit 1
AI and the years ahead
How and why might we build future AI systems?
Unit 2
What is AI alignment?
What do we need to do to ensure AI systems do what we want, and why is this difficult?
Unit 3
Reinforcement learning from human (or AI) feedback
Why do AI systems today mostly do what we want?
Unit 4
Scalable oversight
How might we scale human feedback for more powerful and complex models?
Unit 5
Robustness unlearning and control
Can we prevent oversight methods from being gamed?
Unit 6
Mechanistic interpretability
How might we understand what’s going on inside an AI model?
Unit 7
Technical governance approaches
How might we measure and mitigate the risks of deploying AI models?
Unit 8
Contributing to AI safety
How can you contribute?
Unit 9
Rapidly testing your project
Unit 10
Developing your project
Unit 11
Further developing your project
Unit 12
Building in public
Put your ideas on paper and share them with the world
Analytics cookies help us improve our website and measure ad performance. Privacy Policy.