Trust, reward hacking, and recursive self-improvement

The Don’t Worry About the Vase Podcast is a listener-supported podcast. To receive new posts and support the cost of creation, consider becoming a free or paid subscriber.

* 00:00:00 - Introduction

* 00:00:35 - Table of Contents

* 00:02:17 - Levels of Friction

* 00:06:42 - The Art Of The Jailbreak

* 00:08:17 - The Quest for Sane Regulations

* 00:14:20 - People Really Hate AI

* 00:20:47 - A Very Bad Paper

* 00:28:16 - Rhetorical Innovation

* 00:35:40 - The Most Forbidden Technique

* 00:37:03 - Everyone Is Or Should Be Confused About Morality

* 00:39:01 - Aligning a Smarter Than Human Intelligence is Difficult

* 00:48:35 - We’ll Just Call It Something Else

* 00:51:10 - Vulnerable World Hypothesis

* 00:56:10 - Autonomous Killer Robots

* 00:58:05 - People Will Hand Over Power To The AIs

* 01:02:09 - People Are Worried About AI Killing Everyone

* 01:04:20 - Other People Are Not Worried About AI Killing Everyone

* 01:05:35 - The Lighter Side

https://open.substack.com/pub/thezvi/p/ai-156-part-2-errors-in-rhetoric?utm_campaign=post-expanded-share&utm_medium=web