6min chapter

80,000 Hours Podcast cover image

#92 – Brian Christian on the alignment problem

80,000 Hours Podcast

CHAPTER

Can I Learn to Do Things That We Can't Do?

Inverse reinforcement learning is a two stage thing. First figure out the goal, then figure out the behavior that gets the goalg. So in many cases, right, we just want to let the system do that optimization itself. It's not really a cycle per se, but it is a two-stage thing. I did a cleveration back in two thousand 17 on inverse t reenforcement learning. And hev got a paper that will link to. Do you know, how excited is the field about inverse reinforce learning as a way of alining a i with kind of complex values? Yes. One of the major thrusts of a i safety research. Stuart russell's group

00:00

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