The opposite post (reasons not to worry) could be good as well. e.g.
No sign of automation accelerating in general
Maybe extreme lack of consensus about basic questions, mechanisms, solutions.
Mid-term safety work is chasing a uselessly fast-moving target—who was thinking about language model alignment even 3 years ago?
The opposite post (reasons not to worry) could be good as well. e.g.
No sign of automation accelerating in general
Maybe extreme lack of consensus about basic questions, mechanisms, solutions.
Mid-term safety work is chasing a uselessly fast-moving target—who was thinking about language model alignment even 3 years ago?