AI-powered
podcast player
Listen to all your favourite podcasts with AI-powered features
5min chapter
David Krueger–Coordination, Alignment, Academia
The Inside View
The Importance of Safety Performance Tradeoffs in Existential Safety
I think there will always be some amount of safety performance trade off no matter how good of like technical progress we make on alignment so I don't do it as something that can be solved perfectly at anytime soon. One of the things that we can trade off for is you can just like test your system more before you deploy it and that means that it takes you longer to deploy the system. And another one is keeping a human in the loop so in general a lot of times you know having a human sort of overseeing the systems behavior and saying like oh this doesn't look safe let's shut it down or let's not let it take that action whatever could really harm the performance of the system.
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