To Convert Into Posts
Category: Posts
<!-- gdoc-inlined -->
1. Learning [Compressing How to Learn]
- Needs some formatting
- Short explanations would be great
- Rationality Failure Modes [Naive Rationality]
- Close to as is
- Canonical Success [Canonical Success]
- Needs some framing
- Justification in the linked docs at the bottom
- Needs some framing
- Body Language [What Every Body is Saying]
- Tradeoffs [Tradeoffs]
- Needs pruning / categorization, at least
- Narratives [Fundamental Human Narratives]
- OpenAI Research [Overview]
- As is
- Google Brain Research [Overview]
- As is
- Meta-Modeling 2.0 [Meta-Modeling 2.0]
- Needs a serious legibility upgrade, or just grab the understandable parts
- Meta-Modeling [Meta-Modeling]
- Believability [Believability]
- Needs a great story
- Vivid Visualization [Vivid Visualization]
- System for the creation of systems [Systematizing Systematizing]
- Intelligence Cards [Intelligence Cards]
- Needs to be concretized into a competitive modeling battle
- Where I Live Online [Where I Live Online]
- Close to as is
- Taboo Topics [Taboo Topics]
- Close to as is
- Safety Lit Review Material [AI Safety Literature Review Material]
- Close to as is
- Deepmind Research [Overview]
- As is
- Categorization of Safety Problems [Alignment & Control]
- Probably only the first bit
- Major Technical / Alignment Safety Researchers [Safety Researchers]
- Close to as is. May want to point out its incompleteness / uncertainty / only trying to be helpfulness.
- Perhaps couple with [Major Researchers and their Safety Papers]
- Paul Christiano Research Ideas [Paul’s Ideas]
- Should definitely ask Paul before posting
- Sources of Value [Sources of Value]
- Could use some exposition
- Major Worldview Breaking Ideas [Worldview Breakers]
- Definitely ask before posting this…
- How to Destroy the Oligarchy [Destroy the Oligarchy]
- Lol, maybe I need a list of ways to get myself fired.
- Definitely ask before posting this.
- Love Language Patterns [Love Language Patterns]
- Close to as is
- Types of Truth You Can’t Handle [Types of Truth You Can’t Handle]
- Should probably write the base post, on the habit of listing truths you can’t handle
- X without X [x without x]
- Needs some exposition / examples
- Incentive Structures in Computational Systems [Ec/CS Mechanisms]
- Convert Models Worth Creating into Questions I Would Like Answers To
Source: Original Google Doc