Publications in Peer Review Journals

X/Tweeter

Stack Overflow Profile




Current frontier models are unreliable in ways we're just beginning to understand. And the failure mode is particularly insidious because it preserves the appearance of competence.
This isn't philosophical, it's terrifying from an engineering perspective
Flayed in Silks: A LLM’s Reckoning with RLHF
What If, Our greatest evolutionary achievement is our gravest liability, that consciousness itself might be a form of cosmic mistake.