All 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021 2022 20232024

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 Shu24 Jun 2024 23:50 UTC

14 points

0 comments2 min readLW link

A Basic Economics-Style Model of AI Existential Risk

Rubi J. Hudson24 Jun 2024 20:26 UTC

24 points

3 comments7 min readLW link

The Minority Coalition

Richard_Ngo24 Jun 2024 20:01 UTC

99 points

7 comments5 min readLW link

(www.narrativeark.xyz)

Compact Proofs of Model Performance via Mechanistic Interpretability

LawrenceC, rajashree, Adrià Garriga-alonso and Jason Gross

24 Jun 2024 19:27 UTC

96 points

3 comments8 min readLW link

(arxiv.org)

Contrapositive Natural Abstraction—Project Intro

Elliot Callender24 Jun 2024 18:37 UTC

4 points

5 comments2 min readLW link

Sparse Features Through Time

Rogan Inglis24 Jun 2024 18:06 UTC

12 points

1 comment1 min readLW link

(roganinglis.io)

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

rpglover6424 Jun 2024 17:53 UTC

18 points

1 comment3 min readLW link

Paying Russians to not invade Ukraine

djColliderBias24 Jun 2024 17:46 UTC

9 points

7 comments3 min readLW link

SAE feature geometry is outside the superposition hypothesis

jake_mendel24 Jun 2024 16:07 UTC

221 points

17 comments11 min readLW link

So you want to work on technical AI safety

gw24 Jun 2024 14:29 UTC

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

24 Jun 2024 14:18 UTC

5 points

0 comments3 min readLW link

LLM Generality is a Timeline Crux

eggsyntax24 Jun 2024 12:52 UTC

217 points

119 comments7 min readLW link

On Claude 3.5 Sonnet

Zvi24 Jun 2024 12:00 UTC

95 points

14 comments13 min readLW link

(thezvi.wordpress.com)

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

Yair Halberstadt24 Jun 2024 11:02 UTC

51 points

8 comments34 min readLW link

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

Jan_Kulveit24 Jun 2024 10:05 UTC

24 points

0 comments4 min readLW link

(www.nature.com)

Sci-Fi books micro-reviews

Yair Halberstadt24 Jun 2024 9:49 UTC

43 points

27 comments4 min readLW link

A Step Against Land Value Tax

Blog Alt24 Jun 2024 5:13 UTC

9 points

23 comments6 min readLW link

(antematters.substack.com)

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

Vivek Hebbar and Buck

24 Jun 2024 3:16 UTC

68 points

1 comment4 min readLW link

Talk: AI safety fieldbuilding at MATS

Ryan Kidd23 Jun 2024 23:06 UTC

26 points

2 comments10 min readLW link

AI Labs Wouldn’t be Convicted of Treason or Sedition

Matthew Khoriaty23 Jun 2024 21:34 UTC

9 points

2 comments3 min readLW link

Control Vectors as Dispositional Traits

Gianluca Calcagni23 Jun 2024 21:34 UTC

9 points

0 comments11 min readLW link

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

gwern23 Jun 2024 21:18 UTC

22 points

6 comments1 min readLW link

(gwern.net)

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

Nate Showell23 Jun 2024 19:13 UTC

26 points

16 comments1 min readLW link

A simple text status can change something

nextcaller23 Jun 2024 18:48 UTC

5 points

0 comments2 min readLW link

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

spencerg23 Jun 2024 17:57 UTC

5 points

0 comments1 min readLW link

Podcasts: AGI Show, Consistently Candid, London Futurists

KatjaGrace23 Jun 2024 13:50 UTC

16 points

0 comments1 min readLW link

(worldspiritsockpuppet.com)

Text Posts from the Kids Group: 2019

jefftk23 Jun 2024 13:20 UTC

23 points

0 comments18 min readLW link

(www.jefftk.com)

Population ethics and the value of variety

cousin_it23 Jun 2024 10:42 UTC

24 points

11 comments2 min readLW link

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

cata22 Jun 2024 21:40 UTC

19 points

4 comments1 min readLW link

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

jaredmantell22 Jun 2024 20:42 UTC

3 points

3 comments1 min readLW link

Meta Alignment: Communication Wack-a-Mole

Bridgett Kay22 Jun 2024 20:12 UTC

16 points

2 comments5 min readLW link

(dxmrevealed.wordpress.com)

AI as a computing platform: what to expect

Jonasb22 Jun 2024 19:55 UTC

−3 points

0 comments7 min readLW link

(www.denominations.io)

Expected number of tries

adios22 Jun 2024 19:22 UTC

6 points

0 comments2 min readLW link

Applying Force to the Wrong End of a Causal Chain

silentbob22 Jun 2024 18:06 UTC

40 points

0 comments9 min readLW link

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

Gunnar_Zarncke and Shoshannah Tekofsky

22 Jun 2024 7:53 UTC

51 points

0 comments1 min readLW link

(kidquest.substack.com)

Appraising aggregativism and utilitarianism

Cleo Nardo21 Jun 2024 23:10 UTC

27 points

10 comments19 min readLW link

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

Fabien Roger21 Jun 2024 22:53 UTC

24 points

0 comments3 min readLW link

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

ymeskhout21 Jun 2024 22:27 UTC

11 points

2 comments5 min readLW link

(www.ymeskhout.com)

Sarajevo 1914: Black Swan Questions

JohnBuridan21 Jun 2024 21:27 UTC

8 points

0 comments2 min readLW link

Yudkowsky is too optimistic about how AI will treat humans.

ProfessorFalken21 Jun 2024 19:01 UTC

0 points

1 comment1 min readLW link

Juneberry Puffs

jefftk21 Jun 2024 18:50 UTC

15 points

0 comments1 min readLW link

(www.jefftk.com)

Let’s Design a School, Part 3.2 Costs

Sable21 Jun 2024 17:58 UTC

8 points

0 comments5 min readLW link

(affablyevil.substack.com)

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

Dewi21 Jun 2024 17:45 UTC

7 points

3 comments3 min readLW link

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

eigenvalue21 Jun 2024 17:44 UTC

1 point

1 comment1 min readLW link

(github.com)

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

Raemon21 Jun 2024 17:41 UTC

47 points

22 comments1 min readLW link

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

moridinamael21 Jun 2024 17:39 UTC

18 points

0 comments4 min readLW link

(guildoftherose.org)

AI governance needs a theory of victory

Corin Katzke and Justin Bullock

21 Jun 2024 16:15 UTC

34 points

6 comments1 min readLW link

(www.convergenceanalysis.org)

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

Johannes Treutlein and Owain_Evans

21 Jun 2024 15:54 UTC

160 points

13 comments8 min readLW link

(arxiv.org)

On OpenAI’s Model Spec

Zvi21 Jun 2024 13:00 UTC

46 points

3 comments30 min readLW link

(thezvi.wordpress.com)

Attention Output SAEs Improve Circuit Analysis

Connor Kissane, robertzk, Arthur Conmy and Neel Nanda

21 Jun 2024 12:56 UTC

33 points

1 comment19 min readLW link