Alicorn’s disavowal is due to fear of something learning about her through a series of weak inferences; however, ironically, her disavowal/vow to avoid revealing data useful for weak inference is itself grist for the weak inference mill. (And hopefully my example inferences show that the weak inferences may not be all that weak.)
That something learning about her might be entirely neutral. On the other hand, it might be an unFriendly AI programmed by Black Panthers who were picked on as kids by nerds and are irritated by Ryan North’s longboard stylings.
Alicorn’s disavowal is due to fear of something learning about her through a series of weak inferences; however, ironically, her disavowal/vow to avoid revealing data useful for weak inference is itself grist for the weak inference mill. (And hopefully my example inferences show that the weak inferences may not be all that weak.)
That something learning about her might be entirely neutral. On the other hand, it might be an unFriendly AI programmed by Black Panthers who were picked on as kids by nerds and are irritated by Ryan North’s longboard stylings.