AI-powered
podcast player
Listen to all your favourite podcasts with AI-powered features
59sec snip
Holly Elmore on pausing AI
The Inside View
Ensuring Alignment Between Capabilities and Exceeding Capabilities
There may be a challenge in achieving stable alignment between capabilities that exceed our own and those within our level of understanding. The speaker questions whether misalignments could grow too significant due to the vast differential in capabilities. The uncertainty remains whether dedicating substantial resources of money and time could solve this issue or if there is inherently no way to ensure safety in such situations. An interesting perspective from Paul Christiano suggests having 500 safety researchers for every physicist to address the alignment concerns, emphasizing the importance of allocating resources effectively.
00:00
Transcript
Episode notes
Get the Snipd
podcast app
Unlock the knowledge in podcasts with the podcast player of the future.
AI-powered
Discover
highlights
Listen to the best highlights from the podcasts you love and dive into the full episode
Save any
moment
Hear something you like? Tap your headphones to save it with AI-generated key takeaways
Share
& Export
Send highlights to Twitter, WhatsApp or export them to Notion, Readwise & more
AI-powered
podcast player
Listen to all your favourite podcasts with AI-powered features
Discover
highlights
Listen to the best highlights from the podcasts you love and dive into the full episode