EIEpisodeIndex
From podcastDwarkesh PodcastFrom guestPaul Christiano

Episode #56

Paul Christiano — Preventing an AI takeover

Paul Christiano is the world’s leading AI safety researcher.

October 31, 20233h 7m11 chaptersPaul Christiano
More from Dwarkesh Podcast
What this episode covers

Paul Christiano is the world’s leading AI safety researcher.

Where to start

Will this revolutionize theoretical CS and math?

Will this revolutionize theoretical CS and math?

Start at 2:35:01

What do we want post-AGI world to look like?

What do we want post-AGI world to look like?

Start at 0:00

Disagreements with Carl Shulman

Disagreements with Carl Shulman

Start at 2:55:10

People and topics
Key takeaways
  • What do we want post-AGI world to look like?
  • Timelines
  • Evolution vs gradient descent
  • Misalignment and takeover
  • Is alignment dual-use?
All moments
Paul Christiano — Preventing an AI takeover podcast chapters, timestamps & summary | EpisodeIndex