AI-powered
podcast player
Listen to all your favourite podcasts with AI-powered features
2min snip
John Schulman (OpenAI Cofounder) - Reasoning, RLHF, & Plan for 2027 AGI
Dwarkesh Podcast
**Pre-training and post-training **
Pre-training involves training a model to imitate all the content on the internet and maximize likelihood by predicting the next token given the previous tokens. This results in a model that can generate content similar to random web pages and assign probabilities to everything. On the other hand, post-training aims to target a narrower range of behavior, such as creating a helpful chat assistant persona. The objective of post-training is to produce outputs that humans will like and find useful, instead of just imitating raw content from the web.
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