Right here is my dialog with Dario Amodei, CEO of Anthropic.
Dario is hilarious and has fascinating takes on what these fashions are doing, why they scale so properly, and what it can take to align them.
Transcript: https://www.dwarkeshpatel.com/dario-amodei
Apple Podcasts: https://apple.co/3rZOzPA
Spotify: https://spoti.fi/3QwMXXU
Comply with me on Twitter: https://twitter.com/dwarkesh_sp
—
I’m working an experiment on this episode.
I’m not doing an advert.
As an alternative, I’m simply going to ask you to pay for no matter worth you’re feeling you personally received out of this dialog.
Pay right here: https://bit.ly/3ONINtp
—
(00:00:00) – Introduction
(00:01:00) – Scaling
(00:15:46) – Language
(00:22:58) – Financial Usefulness
(00:38:05) – Bioterrorism
(00:43:35) – Cybersecurity
(00:47:19) – Alignment & mechanistic interpretability
(00:57:43) – Does alignment analysis require scale?
(01:05:30) – Misuse vs misalignment
(01:09:06) – What if AI goes properly?
(01:11:05) – China
(01:15:11) – How to consider alignment
(01:31:31) – Is fashionable safety adequate?
(01:36:09) – Inefficiencies in coaching
(01:45:53) – Anthropic’s Lengthy Time period Profit Belief
(01:51:18) – Is Claude acutely aware?
(01:56:14) – Holding a low profile
