Post

A perfectly safe AI in an authoritarian setting is not safe at all. Matthew Botvinick starts a new thread on AI safety and democracy. Out from @princetonupress.bsky.social rorotoko.com/micro-inter...