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 121314 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31

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)

Theories of Deep Learning

astle dsa11 Jul 2026 17:21 UTC

5 points

0 comments7 min readLW link

Measuring Is Not Enough Anymore

Lennart Finke11 Jul 2026 3:21 UTC

11 points

1 comment5 min readLW link

(fi-le.net)

Don’t bring an AI detector to a deepfake fight: proving reality through multimodal provenance

Julien Despois11 Jul 2026 1:58 UTC

10 points

0 comments9 min readLW link

A Simple Model of AI “Psychosis”

Adele Lopez11 Jul 2026 1:53 UTC

42 points

1 comment6 min readLW link

The Termination Circuit (how reasoning models stop thinking).

Chandram Dutta11 Jul 2026 1:37 UTC

12 points

0 comments3 min readLW link

(onlychan.xyz)

Notes on Tony Parkes’ “Contra Dance Calling”

jefftk11 Jul 2026 0:50 UTC

10 points

0 comments4 min readLW link

(www.jefftk.com)

Plan A: Suggestions For Further Work

Thomas Larsen10 Jul 2026 23:02 UTC

52 points

7 comments7 min readLW link

Freeing Thucydides

djbinder10 Jul 2026 22:12 UTC

44 points

2 comments4 min readLW link

(defensesindepth.bio)

An Induction Head in Disguise: Chasing Grammar in a Character-Level Transformer

Ameya Panchal10 Jul 2026 22:05 UTC

−1 points

0 comments5 min readLW link

(ameya-bit.github.io)

The standard can be to admit the existence of a standard

Firinn10 Jul 2026 21:22 UTC

44 points

1 comment36 min readLW link

Persona Cartography: Charting Language Model Personality Traits in Weight Space

antonghawthorne, Mariia Koroliuk, Irakli Shalibashvili, sidbaines, Clément Dumas, Konstantinos Voudouris and David Africa

10 Jul 2026 18:54 UTC

44 points

0 comments18 min readLW link

(arxiv.org)

The easiest pathway to control is through executive power

djbinder10 Jul 2026 18:48 UTC

111 points

3 comments6 min readLW link

(defensesindepth.bio)

The Human Substitution Test as a Sanity Check for AI Evaluations

VojtaKovarik, Tomáš Gavenčiak and Mateusz Bagiński

10 Jul 2026 17:27 UTC

31 points

5 comments8 min readLW link

(limits-of-evaluation.org)

Cap-and-trade question: AI-2040

kapedalex10 Jul 2026 16:54 UTC

0 points

2 comments4 min readLW link

Does AI reasoning imply responsibility?

Hippocleides10 Jul 2026 16:44 UTC

1 point

3 comments2 min readLW link

Plan A’s problem with dry tinder

Tom Davidson10 Jul 2026 15:28 UTC

81 points

6 comments8 min readLW link

AI #176 Part 2: Plan B

Zvi10 Jul 2026 12:40 UTC

26 points

1 comment36 min readLW link

(thezvi.wordpress.com)

Beliefs and position mid 2026

RussellThor10 Jul 2026 11:43 UTC

7 points

2 comments6 min readLW link

Models of Society Are Built on Models of Agents

Jonas Hallgren10 Jul 2026 9:08 UTC

18 points

0 comments10 min readLW link

(equilibria1.substack.com)

Value generalisation: value correction

Stuart_Armstrong10 Jul 2026 7:56 UTC

25 points

3 comments6 min readLW link

Don’t normalize a permanent underclass (even a rich one)

hadad10 Jul 2026 6:40 UTC

39 points

7 comments5 min readLW link

How robust are natural language autoencoders to initialization?

michaelzhang and TurnTrout

10 Jul 2026 0:40 UTC

82 points

3 comments13 min readLW link

(turntrout.com)

A genealogy of AI safety: how directions are born, and how they die (2005-2026)

Elena Ericheva10 Jul 2026 0:35 UTC

23 points

0 comments29 min readLW link

Reading into VLM hallucinations using the Jacobian lens

Hawrani10 Jul 2026 0:31 UTC

8 points

0 comments6 min readLW link

Are We Guarding Against Backdoors Or Failing To Notice Them? (Part 1 / 6)

SakshamSingh10 Jul 2026 0:31 UTC

9 points

0 comments6 min readLW link

Toward A Public Science of Model Behavior

Daniel D. Johnson and Sarah Schwettmann

10 Jul 2026 0:27 UTC

24 points

1 comment9 min readLW link

(transluce.org)

AI Safety Policy Needs to train Legal Practitioners

Katalina Hernandez10 Jul 2026 0:13 UTC

55 points

29 comments7 min readLW link

(stresstestingreality.substack.com)

What would it take for AI to discover penicillin?

bosoncutter10 Jul 2026 0:00 UTC

11 points

0 comments5 min readLW link

(bosoncutter.substack.com)

Natural Language Autoencoders are summarizers, but do they have to be?

Andrey Anurin9 Jul 2026 19:54 UTC

10 points

1 comment22 min readLW link

Where Do LLM Values Come From?

lilysun004, Arthur Conmy and Josh Engels

9 Jul 2026 19:54 UTC

17 points

1 comment18 min readLW link

Selective Optimism: a critique of AI 2040

Richard_Ngo9 Jul 2026 19:43 UTC

231 points

13 comments8 min readLW link

(www.mindthefuture.info)

Rogue ASI Can’t Stay Aligned to Itself

absenteewarlord9 Jul 2026 19:38 UTC

8 points

2 comments1 min readLW link

Criticism against “unembedded FDT” doesn’t apply to FDT

Fernand09 Jul 2026 18:35 UTC

2 points

5 comments3 min readLW link

Your Prompt-Injection Defense Metric Might Be Lying to You

sahilraut9 Jul 2026 17:53 UTC

5 points

0 comments8 min readLW link

When is misalignment just a bug?

Yoav Hollander9 Jul 2026 17:07 UTC

16 points

5 comments11 min readLW link

(blog.foretellix.com)

What is the computational substance of the axiom of choice?

tailcalled9 Jul 2026 16:53 UTC

26 points

2 comments6 min readLW link

Skeptical of the TESCREAL Acronym? Read This.

philosophytorres9 Jul 2026 16:27 UTC

−41 points

12 comments20 min readLW link

AI 2040: Plan A

Daniel Kokotajlo, elifland, Thomas Larsen, romeo, brendanhalstead and ryan_greenblatt

9 Jul 2026 16:25 UTC

568 points

135 comments1 min readLW link

(www.ai-2040.com)