All 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021 2022 2023 20242025

All Jan Feb Mar Apr May Jun JulAugSep Oct

All1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31

How Does A Blind Model See The Earth?

henry11 Aug 2025 19:58 UTC

473 points

38 comments7 min readLW link

(outsidetext.substack.com)

Four ways learning Econ makes people dumber re: future AI

Steven Byrnes21 Aug 2025 17:52 UTC

359 points

49 comments6 min readLW link

(x.com)

AI Induced Psychosis: A shallow investigation

Tim Hua26 Aug 2025 20:03 UTC

357 points

43 comments26 min readLW link

The Problem

Rob Bensinger, tanagrabeast, yams, So8res, Eliezer Yudkowsky and Gretta Duleba

5 Aug 2025 21:40 UTC

313 points

218 comments26 min readLW link

How anticipatory cover-ups go wrong

Kaj_Sotala8 Aug 2025 10:26 UTC

295 points

25 comments6 min readLW link

Banning Said Achmiz (and broader thoughts on moderation)

habryka22 Aug 2025 23:02 UTC

244 points

395 comments30 min readLW link

Church Planting: When Venture Capital Finds Jesus

Elizabeth16 Aug 2025 19:40 UTC

226 points

23 comments16 min readLW link

(acesounderglass.com)

An epistemic advantage of working as a moderate

Buck20 Aug 2025 17:47 UTC

217 points

96 comments4 min readLW link

Emotions Make Sense

DaystarEld3 Aug 2025 7:03 UTC

207 points

40 comments21 min readLW link

(daystareld.com)

Hyperbolic model fits METR capabilities estimate worse than exponential model

gjm19 Aug 2025 15:12 UTC

201 points

9 comments4 min readLW link

Will Any Crap Cause Emergent Misalignment?

J Bostock27 Aug 2025 18:20 UTC

192 points

37 comments3 min readLW link

Should you make stone tools?

Alex_Altair14 Aug 2025 0:15 UTC

190 points

48 comments3 min readLW link

Before LLM Psychosis, There Was Yes-Man Psychosis

johnswentworth25 Aug 2025 17:47 UTC

186 points

20 comments3 min readLW link

Many prediction markets would be better off as batched auctions

William Howard2 Aug 2025 12:04 UTC

173 points

21 comments5 min readLW link

(antidiluvian.substack.com)

Somebody invented a better bookmark

Alex_Altair14 Aug 2025 17:57 UTC

173 points

22 comments2 min readLW link

My AGI timeline updates from GPT-5 (and 2025 so far)

ryan_greenblatt20 Aug 2025 16:11 UTC

163 points

14 comments4 min readLW link

Underdog bias rules everything around me

Richard_Ngo17 Aug 2025 19:21 UTC

159 points

53 comments7 min readLW link

(www.mindthefuture.info)

Open Global Investment as a Governance Model for AGI

Nick Bostrom27 Aug 2025 17:42 UTC

152 points

47 comments39 min readLW link

(nickbostrom.com)

My Interview With Cade Metz on His Reporting About Lighthaven

Zack_M_Davis17 Aug 2025 2:30 UTC

151 points

15 comments5 min readLW link

Re: recent Anthropic safety research

Eliezer Yudkowsky6 Aug 2025 22:52 UTC

145 points

22 comments5 min readLW link

(x.com)

METR’s Evaluation of GPT-5

GradientDissenter7 Aug 2025 22:17 UTC

141 points

15 comments20 min readLW link

(metr.github.io)

The Inkhaven Residency

Ben Pace2 Aug 2025 18:51 UTC

134 points

35 comments3 min readLW link

Training a Reward Hacker Despite Perfect Labels

ariana_azarbal, vgillioz and TurnTrout

14 Aug 2025 23:57 UTC

132 points

45 comments4 min readLW link

SB-1047 Documentary: The Post-Mortem

Michaël Trazzi1 Aug 2025 21:42 UTC

130 points

0 comments5 min readLW link

(∃ Stochastic Natural Latent) Implies (∃ Deterministic Natural Latent)

johnswentworth and David Lorell

22 Aug 2025 21:46 UTC

126 points

8 comments9 min readLW link

Towards Alignment Auditing as a Numbers-Go-Up Science

Sam Marks4 Aug 2025 22:30 UTC

123 points

15 comments6 min readLW link

Agent foundations: not really math, not really science

Alex_Altair17 Aug 2025 5:48 UTC

114 points

25 comments5 min readLW link

The Egyptian Mamluks as case study for AI take-over

Buddenbroke19 Aug 2025 16:46 UTC

113 points

6 comments7 min readLW link

The Bone-Chilling Evil of Factory Farming

Bentham's Bulldog12 Aug 2025 18:02 UTC

109 points

11 comments6 min readLW link

Why Latter-day Saints Have Strong Communities

Jeffrey Heninger17 Aug 2025 4:20 UTC

102 points

29 comments9 min readLW link

METR Research Update: Algorithmic vs. Holistic Evaluation

David Rein13 Aug 2025 22:47 UTC

101 points

7 comments1 min readLW link

(metr.org)

[Question] Inscrutability was always inevitable, right?

Steven Byrnes6 Aug 2025 21:57 UTC

99 points

33 comments2 min readLW link

Attaching requirements to model releases has serious downsides (relative to a different deadline for these requirements)

ryan_greenblatt27 Aug 2025 17:04 UTC

99 points

2 comments3 min readLW link

Von Neumann’s Fallacy and You

incident-recipient28 Aug 2025 15:52 UTC

98 points

29 comments4 min readLW link

Yudkowsky on “Don’t use p(doom)”

Raemon22 Aug 2025 23:44 UTC

98 points

39 comments4 min readLW link

Summary of our Workshop on Post-AGI Outcomes

David Duvenaud, Raymond Douglas, Nora_Ammann and Jan_Kulveit

29 Aug 2025 17:14 UTC

96 points

3 comments3 min readLW link

Generalized Coming Out Of The Closet

johnswentworth12 Aug 2025 21:38 UTC

92 points

51 comments4 min readLW link

Aesthetic Preferences Can Cause Emergent Misalignment

Anders Woodruff26 Aug 2025 18:41 UTC

90 points

16 comments3 min readLW link

Steve Petersen seeking funding

abramdemski1 Aug 2025 17:03 UTC

87 points

0 comments1 min readLW link

[Anthropic] A hacker used Claude Code to automate ransomware

bohaska27 Aug 2025 14:57 UTC

86 points

25 comments3 min readLW link

(www.anthropic.com)

A Comprehensive Guide to Running

Declan Molony25 Aug 2025 15:12 UTC

85 points

24 comments16 min readLW link

Briefly on MAPLE, and the broader community

herschel19 Aug 2025 19:45 UTC

83 points

38 comments6 min readLW link

Debugging for Mid Coders

Raemon16 Aug 2025 22:32 UTC

82 points

41 comments7 min readLW link

The Collider Bias Theory of (Not Quite) Everything

Jack_S16 Aug 2025 16:53 UTC

82 points

3 comments10 min readLW link

The Dark Arts As A Scaffolding Skill For Rationality

Screwtape1 Aug 2025 17:12 UTC

82 points

25 comments7 min readLW link

Shorter Tokens Are More Likely

Brendan Long24 Aug 2025 0:22 UTC

81 points

19 comments5 min readLW link

(www.brendanlong.com)

Arguments About AI Consciousness Seem Highly Motivated And At Best Overconfident

Zvi25 Aug 2025 13:20 UTC

81 points

5 comments25 min readLW link

(thezvi.wordpress.com)

My Least Libertarian Opinion: Ban Exclusivity Deals*

Brendan Long10 Aug 2025 21:41 UTC

78 points

17 comments2 min readLW link

(www.brendanlong.com)

On closed-door AI safety research

richbc18 Aug 2025 21:59 UTC

76 points

11 comments15 min readLW link

Mech Interp Wiki Page and Why You Should Edit Wikipedia

Noah Birnbaum and JoNeedsSleep

12 Aug 2025 17:28 UTC

75 points

16 comments1 min readLW link