The 80,000 Hours Podcast on Artificial Intelligence and related topics
A compilation of eleven key episodes of the 80,000 Hours Podcast on artificial intelligence and related topics.
Subscribe here, or anywhere you get podcasts:
About the show
The 80,000 Hours Podcast on Artificial Intelligence is a collection of eleven key episodes of The 80,000 Hours Podcast, selected to help listeners learn about the potential upsides and downsides of powerful and transformative AI models.
The 11 episodes in this series cover:
- How modern AI models work, and what we don’t know about how they work
- Ways advances in AI could radically improve our wellbeing
- Ways informed people fear AI advances could go wrong and cause a huge amount of harm
- Some of the research being conducted to make things more likely to go well and less likely to accidentally go badly
- How you might figure out how to make your biggest contribution to positively shaping the development of AI
They’re selected from over 30 episodes of the show that talk about AI.
The compilation should work well for people who are largely new to the topic but interested to learn more.
If you enjoy these the full show has over 150 interviews that dive deep into a wide range of important topics.
The eleven included episodes
Feedback from listeners
I really enjoyed hearing Paul’s reasoning throughout. Really well thought out and articulated.
Rob asked good questions that led the discussion to all the places I could hope it to go.
On episode 3 with Paul Christiano
As someone who has a very limited understanding of the technicalities of AI, I found this episode really enjoyable to listen to.
I felt as though there was a comprehensive take on the issue from foundations through to more specific points
On episode 1 with Brian Christian
This episode really felt like listening to a master of her craft. It made me feel inspired to study more and seek out options in my option space. Nova was such an entertaining speaker.
On episode 9 with Nova DasSarma
I loved this episode! One of my favourites in a while. I particularly liked the different cases for interpretability research and why it matters.
On episode 4 with Chris Olah
Great to hear a more AI risk-, FOOM-skeptical perspective. Plenty of Ben’s arguments were new to me, so I thought they were really valuable to hear.
On episode 6 with Ben Garfinkel
How to subscribe
To listen at leisure on your phone, and choose the audio speed you prefer, subscribe here or anywhere you get podcasts: