This chapter explores the concept of automated feedback in AI systems, where the systems evaluate their own outputs or are evaluated by other AI systems to provide feedback for training. It discusses the challenges of aligning these systems and the potential risks of training AI models. The chapter also discusses the importance of better evaluations in AI safety and proposes the idea of iterated distill amplification as a potential solution.

Get the Snipd
podcast app

Unlock the knowledge in podcasts with the podcast player of the future.
App store bannerPlay store banner

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

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