Loading activity…
@hendrycks
https://eigenism.org
AI
Goodfire CEO Eric Ho argues that understanding models’ internal mechanisms is necessary to shape what they learn during training and verify what they’ve learned afterward.
AI
A post questions whose values guide frontier labs’ AI alignment. A reply argues that some users want to do dangerous things.
AI
The launch announcement says frontier agents still cheat frequently, and describes an evaluation spanning math, coding, knowledge work, visual tasks and more.
AI
CBS Mornings reports that current and former Anthropic employees warned AI could kill all humans. Critics worry exaggerated forecasts could undermine trust, while a defender of numerical forecasts argues they convey the seriousness of the concern.
AI
The author asks: If AI becomes smarter than humans, why would it choose to keep us around?
AI
OpenAI researcher shares his assessment of Hendrycks' contributions to alignment research.
AI
Argues concrete risks from uncontrolled programs warrant concern without picking camps.
