My sense is that the RL paradigm is a pretty good fit there. There's just going to be a ton of what didn't work and then very concretely what did work in terms of code as policy style API calls generated or what have you. In terms of like will RL just make the agents like fix the agent problem? My suspicion is no. So I think that probably the solution will actually be some form of feedback mechanism. Not necessarily human feedback but some way for the models to like heal their agents or heal their chains.

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