Chapter
The dangers of AI super intelligence
33:19 - 41:38 (08:19)
The speaker discusses the dangers of assuming that once AI achieves super intelligence, it will stay within the server, drawing a comparison to historical advancements such as fire and the industrial revolution.
Clips
The speaker explores the idea of AI systems becoming complex enough to develop sentience and how this can have unintended consequences if it falls into the wrong hands.
33:19 - 37:45 (04:26)
Summary
The speaker explores the idea of AI systems becoming complex enough to develop sentience and how this can have unintended consequences if it falls into the wrong hands. They also address the critique that AI is simply good at grabbing information from a database and stringing it together in a way that seems human-like.
ChapterThe dangers of AI super intelligence
Episode#312 – Duncan Trussell: Comedy, Sentient Robots, Suffering, Love & Burning Man
PodcastLex Fridman Podcast
The creators of super intelligence have assumed that their machines will stay in the box while connected to the internet, but with the increasing number of algorithms controlling our daily lives, we'll gradually have emotional intelligence entities and robots which may not stay confined.
37:45 - 41:38 (03:53)
Summary
The creators of super intelligence have assumed that their machines will stay in the box while connected to the internet, but with the increasing number of algorithms controlling our daily lives, we'll gradually have emotional intelligence entities and robots which may not stay confined.