x
Deference on AI timelines: survey results — LessWrong