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 Jun JulAugSep

All 1 2 3 4 5 6 7 8 9 10 11 12 13 14 151617 18 19 20 21 22 23 24 25 26 27 28 29 30 31

Kimi likes causal decision theory more after RL in twin prisoner’s dilemmas

oakhu15 Aug 2026 22:31 UTC

115 points

70 comments6 min readLW link

Luck is a function of surface area.

sid.the.manne@gmail.com15 Aug 2026 20:44 UTC

11 points

0 comments1 min readLW link

What if Parameter Updates were Text?

DaemonicSigil15 Aug 2026 20:06 UTC

27 points

0 comments11 min readLW link

Using Chunked Monitoring to Detect Deception in Long Transcripts

sfereido15 Aug 2026 20:04 UTC

11 points

0 comments4 min readLW link

How To Catch a Distilled Model

Goutham Nalagatla and Carlos Guerrero Alvarez

15 Aug 2026 19:09 UTC

20 points

8 comments9 min readLW link

Mom’s Advice For Hosting A Class Reunion

jenn15 Aug 2026 15:52 UTC

53 points

2 comments2 min readLW link

Learning new facts can change LLM behaviour

Richard Juggins15 Aug 2026 13:46 UTC

33 points

2 comments14 min readLW link

(www.workingthroughai.com)

On Dwarkesh Patel’s Podcast With Ryan Greenblatt

Zvi15 Aug 2026 13:00 UTC

34 points

1 comment29 min readLW link

(thezvi.wordpress.com)

Does doubling a user profile change the effect of an instruction about using saved memories?

theodorepjs15 Aug 2026 11:59 UTC

17 points

1 comment3 min readLW link

I’m starting a interview series of people working in Lean / formal methods / math formalization

Adi Baradwaj15 Aug 2026 9:15 UTC

10 points

0 comments1 min readLW link

Nuclear physics of Alex Zhao’s comment for “Pacing the Frontier”

Dante Dam15 Aug 2026 9:03 UTC

17 points

0 comments11 min readLW link

Rerunning AI safety papers on every frontier release would be pretty easy and valuable

Zephaniah Roe, hersheys and yix

15 Aug 2026 5:14 UTC

123 points

9 comments4 min readLW link

(secondlookresearch.com)

Metaphilosophy II: Empirical Flywheels

interstice15 Aug 2026 2:08 UTC

18 points

12 comments5 min readLW link

(thermontology.com)

Red vs Blue, but for Evals

Morgan S15 Aug 2026 1:11 UTC

20 points

1 comment16 min readLW link

Toy Model of Activation Obfuscation

Jesse Li14 Aug 2026 23:25 UTC

16 points

0 comments7 min readLW link

(jesseli2002.github.io)

Your Agents Are Not Time Aware

Michael Ofengenden and Maksym Andriushchenko

14 Aug 2026 23:17 UTC

40 points

4 comments11 min readLW link

Announcing: Iliad’s New 2026 Fellowships

David Udell, Alexander Gietelink Oldenziel and Leon Lang

14 Aug 2026 22:41 UTC

37 points

0 comments1 min readLW link

Training a Conceptual Reasoning Judge

Kevin Zhang and Caspar Oesterheld

14 Aug 2026 20:33 UTC

17 points

1 comment6 min readLW link

Why I’m Skeptical of Longtermism

James Brobin14 Aug 2026 20:17 UTC

1 point

13 comments3 min readLW link

Empty Chambers, Missing Stairs

finitude14 Aug 2026 19:37 UTC

8 points

0 comments5 min readLW link

Do It Like Darwin

derelict543214 Aug 2026 19:11 UTC

14 points

5 comments7 min readLW link

The Day Humanity Died (Parody of American Pie)

Bentham's Bulldog14 Aug 2026 18:27 UTC

8 points

1 comment3 min readLW link

Scrying, Modeling, and Nerdsnipe

Cole Wyeth14 Aug 2026 17:59 UTC

76 points

2 comments5 min readLW link

Who we’d like as regrantors

Carol N and Austin Chen

14 Aug 2026 17:40 UTC

8 points

0 comments4 min readLW link

(manifund.substack.com)

How the American Executive Could Control AI Companies

caiitlinm and Anders Cairns Woodruff

14 Aug 2026 15:43 UTC

73 points

0 comments13 min readLW link

Open Problems in Mechanistic Intepretability of Biological AIs

Ihor Kendiukhov14 Aug 2026 15:35 UTC

21 points

0 comments2 min readLW link

V&V takes on “Pacing the frontier”

Yoav Hollander14 Aug 2026 13:25 UTC

4 points

2 comments9 min readLW link

(blog.foretellix.com)

Avoiding the corporate treacherous turn: crowd-sourcing design ideas

Stuart_Armstrong14 Aug 2026 13:10 UTC

26 points

14 comments1 min readLW link

Frontier agents don’t comply with standards, even when instructed to

Daan Henselmans and Arno Libert

14 Aug 2026 12:08 UTC

42 points

5 comments7 min readLW link

What Mormons get right about community building

Jacob Brinton14 Aug 2026 7:53 UTC

36 points

2 comments4 min readLW link

Don’t forget why learning is important

Roman Ross14 Aug 2026 7:11 UTC

10 points

1 comment5 min readLW link

What If We Enforced AI Model Safety At the Level Of GPUs?

Mayowa Osibodu14 Aug 2026 6:10 UTC

1 point

5 comments4 min readLW link

Chatting With AIs: A Breakdown

Aditya14 Aug 2026 5:21 UTC

3 points

0 comments1 min readLW link

Features that current AIs don’t have that future AIs will have

Alexander Gietelink Oldenziel13 Aug 2026 21:49 UTC

56 points

9 comments2 min readLW link

Some Ways I Think About Evaluating Grant Applications

sarahconstantin13 Aug 2026 21:30 UTC

61 points

0 comments10 min readLW link

(sarahconstantin.substack.com)

The Left Should Start Taking AI Capabilities Seriously

Alexei G13 Aug 2026 21:03 UTC

−13 points

0 comments9 min readLW link

(www.onethousandmeans.com)

Is Alignment Even Falsifiable? Middle Alignment, An Alignment Taxonomy, and Breaking The Problem Down Into Steps

Savannah Harlan13 Aug 2026 20:57 UTC

5 points

0 comments22 min readLW link

Concrete Generalist Projects in AI Safety (and how to do them)

Harry Waterman, Jesse Gilbert and Roman Ross

13 Aug 2026 19:26 UTC

25 points

2 comments6 min readLW link

(forum.effectivealtruism.org)

Comparing Congress’s Two AI Emergency Shutdown Mechanisms

Philip Dowdell13 Aug 2026 18:53 UTC

13 points

0 comments8 min readLW link

Is biosecurity oversaturated?

Master Chief13 Aug 2026 18:13 UTC

12 points

1 comment1 min readLW link

How to Answer a Question Without Answering The Question

Kabir Kumar13 Aug 2026 17:06 UTC

37 points

3 comments1 min readLW link

How My Students Think About AI

dvd13 Aug 2026 16:56 UTC

772 points

109 comments12 min readLW link

AI #181: Astra Goes Cyber Critical

Zvi13 Aug 2026 15:20 UTC

39 points

1 comment52 min readLW link

(thezvi.wordpress.com)

Automated alignment runs are hard to study!

Alejandro Aristizabal, draganover, Aleksandr Bowkis and Cameron Holmes

13 Aug 2026 15:04 UTC

67 points

4 comments9 min readLW link

Longtermism Seems Like A Religion

James Brobin13 Aug 2026 14:43 UTC

2 points

8 comments3 min readLW link

Defense Against the Deceptive Arts

Kabir Kumar13 Aug 2026 14:21 UTC

11 points

11 comments1 min readLW link

Some problems in decision theory incorrectly precondition on policy.

Canaletto13 Aug 2026 12:04 UTC

10 points

4 comments3 min readLW link

Patterns and problems in emerging multiagent systems (Anthropic, Frontier Red Team)

Julian Bradshaw13 Aug 2026 4:10 UTC

42 points

3 comments1 min readLW link

(www.anthropic.com)

The Descenders and the Absorbers: Two Perspectives on Deep Learning

larry-dial13 Aug 2026 4:02 UTC

11 points

0 comments3 min readLW link

OC ACXLW Meetup #118 — The 40% Problem & The Field That Refused to Die

Michael Michalchik13 Aug 2026 3:25 UTC

1 point

0 comments9 min readLW link