← Home
Lex Fridman Podcast #452 · November 11, 2024 · 5h 22m
Anthropic CEO on Claude, AGI & the Future of AI & Humanity
Anthropic CEO Dario Amodei in the longest and most substantive AI safety conversation on any podcast. Covers Claude's personality design, scaling laws, interpretability research, the race to AGI, and whether AI systems can be meaningfully aligned with human values.
This summary was generated from show notes and public descriptions, not from a full transcript review. Details may contain inaccuracies.
Highlights
•
AI alignment is the most important problem humanity has ever faced
Amodei: if AI systems become more capable than humans (which current scaling trends suggest is possible), ensuring they're aligned with human values becomes the most consequential problem in history — because the downside is extinction.Misc
✧At 5h 22m, this is one of the longest podcast episodes ever recorded
✧Features Amanda Askell and Chris Olah alongside Amodei
Was this useful?