how do we expect to solve AI misalignment, AI wireheading, AI sycophancy when we haven’t even solved them for humans? maybe we could put more effort toward being friends before we potentially introduce an entire new class of enemy?

Andy Roddick joins the Cosa Nostra and threatens someone with his Babolat.

call that a pro-Texan racquet.

it’s always “will you be my emergency contact” and never “will you be my +1”

insane alpha in being load-bearing

computer use MCP sure is something, huh?

your scientists were so preoccupied with whether or not they should, they didn’t stop to think if they could.