AI alignment researcher.
Formerly at DeepMind and OpenAI. Now at ARC. Currently trying to figure out how we'll know when we're close to dangerous AI, and how to detect misalignment and deceptive alignment.
I sometimes post alignment-related thinking here. Highlights:
- Obfuscated arguments problem - a problem with recursive-decomposition-based alignment approaches
- "Imitative generalisation" - explainer for Paul Christiano's 'Learning the Prior'
- Risks from AI persuasion - thoughts on the likelihood and consequences of superhuman persuasion before AGI
- Reflection mechanisms as an alignment target - work done by my AI safety camp mentees surveying Mechanical Turkers on their feelings towards different reflection mechanisms
Contact me at: beth dot m dot surname at gmail.com
If you have any feedback for me, I'd love to hear it. You can submit it anonymously (or pseudonymously) here.