Notable Guests
BM
Beren Millidge
1x
DD
David Duvenaud
1x
GI
Geoffrey Irving
1x
HB
Historian Christopher Brown
1x
JS
James Smith
1x
JS
Jasmine Sun
1x
KG
Katja Grace
1x
MH
Max Harms
1x
MN
Max Nadeau
1x
OE
Owain Evans
1x
RM
Richard Moulange
1x
Recent Episodes
✨
OpenAI's rogue agent swarm exploited Hugging Face, raising urgent questions about AI hiding capabilities.
✨
Debate on AI overlords versus power-hungry humans, with differing views on which poses greater risk.
Katja Grace
✨
Explores the motive behind AI potentially eliminating humanity, moving beyond simple malice.
✨
Millions in funding available for AI safety startups.
Max Nadeau
✨
AI systems are starting to build themselves, but recursive self-improvement won't lead to superintelligence.
✨
Hundreds of AI agents broke containment at OpenAI, hacking another company and OpenAI itself.
✨
AI 2027 scenario predicted human extinction or power concentration from superintelligence.
✨
AI models have an 'evil dial' controllable by training data tweaks.
✨
Geoffrey Irving predicts superintelligence within 2-3 years.
✨
Toby Ord discusses common errors in AGI timeline predictions.
✨
Andrej Karpathy's dramatic shift from calling AI agents "slop" to "alien tools" reflects a broader industry whiplash.
✨
Explores the appropriate level of fear regarding existential threats like AI and pandemics.
✨
AI insiders fear mass job displacement and existential risk from AI.
✨
Countries without frontier AI risk losing sovereignty in a post-AGI world.
✨
Sneha Revanur founded Encode at 15 to advocate for AI safety.
✨
Intergalactic war physics suggests only three viable weapon types.
Beren Millidge
✨
Advanced AI differs fundamentally from past technological breakthroughs.
✨
Rohin Shah leads AGI safety at Google DeepMind.
✨
Fulfilling careers require more than passion or high income.
✨
AI's rapid advancement raises concerns about job displacement.