Bloggingheads: Yudkowsky and Aaronson talk about AI and Many-worlds

Eliezer Yud­kowsky and Scott Aaron­son—Per­con­ta­tions: Ar­tifi­cial In­tel­li­gence and Quan­tum Mechanics

Sec­tions of the di­avlog:

  • When will we build the first su­per­in­tel­li­gence?

  • Why quan­tum com­put­ing isn’t a recipe for robot apocalypse

  • How to guilt-trip a machine

  • The evolu­tion­ary psy­chol­ogy of ar­tifi­cial intelligence

  • Eliezer con­tends many-wor­lds is ob­vi­ously correct

  • Scott con­tends many-wor­lds is ridicu­lous (but might still be true)