All 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021 2022 202320242025

All Jan Feb Mar Apr MayJunJul Aug Sep Oct Nov Dec

All 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 242526 27 28 29 30

Being hella lost as rationality practice

Rachel ShuJun 24, 2024, 11:50 PM

14 points

0 comments2 min readLW link

A Basic Economics-Style Model of AI Existential Risk

Rubi J. HudsonJun 24, 2024, 8:26 PM

24 points

3 comments7 min readLW link

The Minority Coalition

Richard_NgoJun 24, 2024, 8:01 PM

103 points

9 comments5 min readLW link

(www.narrativeark.xyz)

Compact Proofs of Model Performance via Mechanistic Interpretability

LawrenceC, rajashree, Adrià Garriga-alonso and Jason Gross

Jun 24, 2024, 7:27 PM

97 points

4 comments8 min readLW link

(arxiv.org)

Contrapositive Natural Abstraction—Project Intro

Elliot CallenderJun 24, 2024, 6:37 PM

4 points

5 comments2 min readLW link

Sparse Features Through Time

Rogan InglisJun 24, 2024, 6:06 PM

12 points

1 comment1 min readLW link

(roganinglis.io)

PSA: Consider alternatives to AUROC when reporting classifier metrics for alignment

rpglover64Jun 24, 2024, 5:53 PM

18 points

1 comment3 min readLW link

Paying Russians to not invade Ukraine

djColliderBiasJun 24, 2024, 5:46 PM

9 points

7 comments3 min readLW link

SAE feature geometry is outside the superposition hypothesis

jake_mendelJun 24, 2024, 4:07 PM

228 points

17 comments11 min readLW link

So you want to work on technical AI safety

gwJun 24, 2024, 2:29 PM

51 points

3 comments14 min readLW link

The Future of Work: How Can Policymakers Prepare for AI’s Impact on Labor Markets?

davidconrad, Arturs and Tillman Schenk

Jun 24, 2024, 2:18 PM

5 points

0 comments3 min readLW link

LLM Generality is a Timeline Crux

eggsyntaxJun 24, 2024, 12:52 PM

218 points

119 comments7 min readLW link

On Claude 3.5 Sonnet

ZviJun 24, 2024, 12:00 PM

95 points

14 comments13 min readLW link

(thezvi.wordpress.com)

Book Review: Righteous Victims—A History of the Zionist-Arab Conflict

Yair HalberstadtJun 24, 2024, 11:02 AM

53 points

8 comments34 min readLW link

The Living Planet Index: A Case Study in Statistical Pitfalls

Jan_KulveitJun 24, 2024, 10:05 AM

24 points

0 comments4 min readLW link

(www.nature.com)

Sci-Fi books micro-reviews

Yair HalberstadtJun 24, 2024, 9:49 AM

44 points

27 comments4 min readLW link

A Step Against Land Value Tax

Blog AltJun 24, 2024, 5:13 AM

9 points

23 comments6 min readLW link

(antematters.substack.com)

Different senses in which two AIs can be “the same”

Vivek Hebbar and Buck

Jun 24, 2024, 3:16 AM

69 points

2 comments4 min readLW link

Talk: AI safety fieldbuilding at MATS

Ryan KiddJun 23, 2024, 11:06 PM

26 points

2 comments10 min readLW link

AI Labs Wouldn’t be Convicted of Treason or Sedition

Matthew KhoriatyJun 23, 2024, 9:34 PM

9 points

2 comments3 min readLW link

Control Vectors as Dispositional Traits

Gianluca CalcagniJun 23, 2024, 9:34 PM

10 points

0 comments11 min readLW link

“On the Impossibility of Superintelligent Rubik’s Cube Solvers”, Claude 2024 [humor]

gwernJun 23, 2024, 9:18 PM

22 points

6 comments1 min readLW link

(gwern.net)

[Question] How are you preparing for the possibility of an AI bust?

Nate ShowellJun 23, 2024, 7:13 PM

26 points

16 comments1 min readLW link

A simple text status can change something

nextcallerJun 23, 2024, 6:48 PM

5 points

0 comments2 min readLW link

35 Interactive Learning Modules Relevant to EAs / Effective Altruism (that are all free)

spencergJun 23, 2024, 5:57 PM

5 points

0 comments LW link

Podcasts: AGI Show, Consistently Candid, London Futurists

KatjaGraceJun 23, 2024, 1:50 PM

16 points

0 comments1 min readLW link

(worldspiritsockpuppet.com)

Text Posts from the Kids Group: 2019

jefftkJun 23, 2024, 1:20 PM

23 points

0 comments18 min readLW link

(www.jefftk.com)

Population ethics and the value of variety

cousin_itJun 23, 2024, 10:42 AM

24 points

11 comments2 min readLW link

[Question] Karma votes: blind to or accounting for score?

cataJun 22, 2024, 9:40 PM

19 points

4 comments1 min readLW link

[Question] Should effective altruism be more “cool”?

jaredmantellJun 22, 2024, 8:42 PM

3 points

3 comments1 min readLW link

Meta Alignment: Communication Wack-a-Mole

Bridgett KayJun 22, 2024, 8:12 PM

16 points

2 comments5 min readLW link

(dxmrevealed.wordpress.com)

AI as a computing platform: what to expect

JonasbJun 22, 2024, 7:55 PM

−3 points

0 comments7 min readLW link

(www.denominations.io)

Expected number of tries

adiosJun 22, 2024, 7:22 PM

6 points

0 comments2 min readLW link

Applying Force to the Wrong End of a Causal Chain

silentbobJun 22, 2024, 6:06 PM

41 points

0 comments9 min readLW link

Bed Time Quests & Dinner Games for 3-5 year olds

Gunnar_Zarncke and Shoshannah Tekofsky

Jun 22, 2024, 7:53 AM

51 points

0 comments1 min readLW link

(kidquest.substack.com)

Appraising aggregativism and utilitarianism

Cleo NardoJun 21, 2024, 11:10 PM

27 points

10 comments19 min readLW link

Best-of-n with misaligned reward models for Math reasoning

Fabien RogerJun 21, 2024, 10:53 PM

25 points

0 comments3 min readLW link

No really, the Sticker Shortcut fallacy is indeed a fallacy

ymeskhoutJun 21, 2024, 10:27 PM

11 points

2 comments5 min readLW link

(www.ymeskhout.com)

Sarajevo 1914: Black Swan Questions

SebastianG Jun 21, 2024, 9:27 PM

8 points

0 comments2 min readLW link

Yudkowsky is too optimistic about how AI will treat humans.

ProfessorFalkenJun 21, 2024, 7:01 PM

0 points

1 comment1 min readLW link

Juneberry Puffs

jefftkJun 21, 2024, 6:50 PM

15 points

0 comments1 min readLW link

(www.jefftk.com)

Let’s Design a School, Part 3.2 Costs

SableJun 21, 2024, 5:58 PM

8 points

0 comments5 min readLW link

(affablyevil.substack.com)

2022 AI Alignment Course: 5→37% working on AI safety

DewiJun 21, 2024, 5:45 PM

7 points

3 comments3 min readLW link

Some Thoughts on AI Alignment: Using AI to Control AI

eigenvalueJun 21, 2024, 5:44 PM

1 point

1 comment1 min readLW link

(github.com)

What distinguishes “early”, “mid” and “end” games?

RaemonJun 21, 2024, 5:41 PM

48 points

22 comments1 min readLW link

Nuclear War, Map and Territory, Values | Guild of the Rose Newsletter, May 2024

moridinamaelJun 21, 2024, 5:39 PM

18 points

0 comments4 min readLW link

(guildoftherose.org)

AI governance needs a theory of victory

Corin Katzke and Justin Bullock

Jun 21, 2024, 4:15 PM

45 points

8 comments LW link

(www.convergenceanalysis.org)

Connecting the Dots: LLMs can Infer & Verbalize Latent Structure from Training Data

Johannes Treutlein and Owain_Evans

Jun 21, 2024, 3:54 PM

163 points

13 comments8 min readLW link

(arxiv.org)

On OpenAI’s Model Spec

ZviJun 21, 2024, 1:00 PM

47 points

4 comments30 min readLW link

(thezvi.wordpress.com)

Attention Output SAEs Improve Circuit Analysis

Connor Kissane, robertzk, Arthur Conmy and Neel Nanda

Jun 21, 2024, 12:56 PM

33 points

3 comments19 min readLW link