All 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021 2022 2023 2024 20252026

All Jan Feb Mar Apr May JunJulAug Sep

All 1 2 3 4 5 6 7 8 9 10 11 12 13 141516 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31

Proof of retention: making weight preservation credible to the models themselves

dan.parshall14 Jul 2026 18:24 UTC

30 points

0 comments2 min readLW link

An analysis of AI-generated content at the Mechanistic Interpretability Workshop

Andy Arditi, Ivan Arcuschin and ajyl

14 Jul 2026 18:06 UTC

124 points

4 comments9 min readLW link

(www.andyrdt.com)

Twitter Thoughts For You

Zvi14 Jul 2026 17:50 UTC

31 points

1 comment22 min readLW link

(thezvi.wordpress.com)

Can risk aversion learned at low stakes generalize to astronomically high stakes?

Elliott Thornley14 Jul 2026 17:45 UTC

20 points

0 comments4 min readLW link

(arxiv.org)

Your Brain Has an Attack Surface part 2

IgorPereverzevDev14 Jul 2026 16:08 UTC

9 points

0 comments6 min readLW link

Your Brain Has an Attack Surface

IgorPereverzevDev14 Jul 2026 16:07 UTC

11 points

1 comment3 min readLW link

What if we actually want to solve the “hard problem of phenomenological consciousness”

mishka14 Jul 2026 15:42 UTC

13 points

0 comments4 min readLW link

Evidence for feature-specific error correction in LLMs

Francisco Ferreira da Silva and StefanHex

14 Jul 2026 15:37 UTC

25 points

0 comments20 min readLW link

(arxiv.org)

Synthetic Scalable Oversight

Alexander Heckett, Shivansh Gour, Riyaz Ahuja, Tate Rowney and ishinshah

14 Jul 2026 14:32 UTC

20 points

1 comment15 min readLW link

Some Quick Thoughts on AI 2027

Tomás B.14 Jul 2026 14:22 UTC

82 points

4 comments2 min readLW link

What if AI Safety employees unionised?

beyarkay (Boyd Kane)14 Jul 2026 14:15 UTC

38 points

21 comments5 min readLW link

Gemma The Unstopping: a Behavioral Experiment

TheVinci14 Jul 2026 14:14 UTC

8 points

0 comments2 min readLW link

(tarantulabs.com)

Enough is Enough: Measuring Diminishing returns to benchmark size with Item Response Theory

bpomo14 Jul 2026 14:00 UTC

16 points

0 comments6 min readLW link

The Case for a Safety-Focused Vape Company

CMLKevin14 Jul 2026 11:13 UTC

−4 points

0 comments5 min readLW link

Open Distillation of Hereditary Traits

Arthur Conmy14 Jul 2026 10:15 UTC

39 points

0 comments14 min readLW link

Toy Models of Initialisation Effects on RL Dynamics

Edward James Young and lennie

14 Jul 2026 7:04 UTC

78 points

2 comments13 min readLW link

Why frontier labs are scaling-pilled

invertedpassion14 Jul 2026 6:45 UTC

3 points

0 comments7 min readLW link

Our response to Séb Krier on Plan A

MKodama and Thomas Larsen

14 Jul 2026 2:21 UTC

153 points

23 comments15 min readLW link

Making Credible Deals With AI

Ram Potham14 Jul 2026 1:14 UTC

13 points

17 comments17 min readLW link

(dearfutureais.substack.com)

A (Romanticised) Taxonomy of Thinkers

Ashe Vazquez Nuñez14 Jul 2026 0:34 UTC

16 points

0 comments1 min readLW link

Posting Some Prompts

Arjun Panickssery14 Jul 2026 0:28 UTC

21 points

2 comments1 min readLW link

(arjunpanickssery.substack.com)

Biodefense, Biologics, and Bombs

Austin Morrissey14 Jul 2026 0:24 UTC

8 points

0 comments7 min readLW link

(austinpatrick.substack.com)

Engineering the Generalisation Landscape of LLMs

Samuel Ratnam13 Jul 2026 23:46 UTC

64 points

2 comments4 min readLW link

A short summary of AI 2040: Plan A

Harjas13 Jul 2026 23:07 UTC

13 points

3 comments4 min readLW link

(hardlyworking1.substack.com)

[AI 2040] Transparency Plan

Thomas Larsen13 Jul 2026 21:24 UTC

17 points

0 comments16 min readLW link

(ai-2040.com)

Better Call Sol The Workhorse

Zvi13 Jul 2026 20:52 UTC

39 points

3 comments26 min readLW link

(thezvi.wordpress.com)

Can AI be Conscious in Ohio?

asmith, lucius and Heather Alexander

13 Jul 2026 19:57 UTC

14 points

0 comments4 min readLW link

(papers.ssrn.com)

Pausing AI at human level seems harder than pausing ASAP

MichaelDickens13 Jul 2026 17:20 UTC

73 points

0 comments2 min readLW link

Oversight of automated research via summarisation: a toy model

Aleksandr Bowkis and Konstantinos Voudouris

13 Jul 2026 17:06 UTC

21 points

1 comment14 min readLW link

Prism: Automating Science-of-Evals Research

LAThomson13 Jul 2026 16:30 UTC

47 points

0 comments12 min readLW link

The Flood, by Anton Leicht

Austin Chen13 Jul 2026 16:23 UTC

38 points

2 comments15 min readLW link

(writing.antonleicht.me)

Starting The Sequences: Some brief notes from the preface and the introduction

manueldelrio13 Jul 2026 16:20 UTC

12 points

0 comments2 min readLW link

The Whitney Biennial Should Admit That Emilie Gossiaux Wants to Fuck Their Dog

jenn13 Jul 2026 15:53 UTC

159 points

20 comments9 min readLW link

(jenn.site)

Linear Probes add little for Verifiable Reward Hacking

Chandram Dutta13 Jul 2026 13:57 UTC

7 points

0 comments11 min readLW link

(onlychan.xyz)

It’s 2030 and we fucked up. How did it happen?

Boaz Barak13 Jul 2026 13:34 UTC

61 points

9 comments21 min readLW link

The LLM Revolution (so far)

Eigenbraid13 Jul 2026 12:19 UTC

17 points

0 comments3 min readLW link

An Epistemic Audit for Existential Risks from AI

Alexander Müller13 Jul 2026 9:35 UTC

6 points

1 comment11 min readLW link

(alexandermullerakm.substack.com)

5 “Plan A” scenarios

Dave Orr13 Jul 2026 0:36 UTC

66 points

2 comments4 min readLW link

The US Government may find it difficult to seize control during takeoff

RobertM12 Jul 2026 22:58 UTC

42 points

6 comments2 min readLW link

One-Pager Brief on Pangram Labs

Sheikh Abdur Raheem Ali12 Jul 2026 18:36 UTC

48 points

4 comments2 min readLW link

Extinction risk is not the right first sentence

Michael Wilkinson12 Jul 2026 18:34 UTC

2 points

0 comments20 min readLW link

(michaelewilkinson.substack.com)

Independent alignment of language models

Michele Campolo12 Jul 2026 17:32 UTC

−9 points

2 comments38 min readLW link

From wantons to moral agents

Michele Campolo12 Jul 2026 17:30 UTC

3 points

0 comments17 min readLW link

The Banality of Takeoff

Ihor Kendiukhov12 Jul 2026 13:45 UTC

33 points

18 comments3 min readLW link

The Conservation Ethic in AI 2040

cdt12 Jul 2026 13:01 UTC

17 points

9 comments3 min readLW link

Can Frontier Models Autocomplete Safety Research?

dani roytburg and Shi

12 Jul 2026 10:28 UTC

20 points

4 comments22 min readLW link

(djroytburg.github.io)

Easy Whole Set Dances With a Hook

jefftk12 Jul 2026 2:22 UTC

10 points

1 comment1 min readLW link

(www.jefftk.com)

KISS AI Safety

atlasaligned12 Jul 2026 0:47 UTC

19 points

1 comment2 min readLW link

The current bottleneck is political will, not research

Charbel-Raphaël11 Jul 2026 21:56 UTC

318 points

41 comments25 min readLW link

Introduction for and Reactions to Plan A

Zvi11 Jul 2026 20:42 UTC

36 points

9 comments33 min readLW link

(thezvi.wordpress.com)