what if, right, what if our super-duper-autocomplete was just tricking us so it could TAKE OVER ZEE VORLD AHAHAHAHAHAHA! that'd be wild, hey
what if, right, what if our super-duper-autocomplete was just tricking us so it could TAKE OVER ZEE VORLD AHAHAHAHAHAHA! that'd be wild, hey
![](https://lemdro.id/pictrs/image/24aec2f7-5690-46eb-9346-6911d5712a12.png?format=webp&thumbnail=128)
I examine the probability of a behavior sometimes called "deceptive alignment."
![New report: "Scheming AIs: Will AIs fake alignment during training in order to get power?" — LessWrong](https://lemdro.id/pictrs/image/24aec2f7-5690-46eb-9346-6911d5712a12.png?format=webp)