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 Jul AugSep

All 1 2 3 4 5 6 7 8 9 10 111213 14 15 16

A normal Friday in 2042

RobinHa11 Sep 2026 22:29 UTC

21 points

0 comments13 min readLW link

My recommended resources for AI safety, alignment, and existential risks

Lysandre Terrisse11 Sep 2026 22:19 UTC

11 points

0 comments2 min readLW link

Appendix: Reproduction of the OpenAI-HuggingFace Incident

Stewart Slocum11 Sep 2026 21:57 UTC

15 points

0 comments13 min readLW link

Consider positive feedback loops

Pato11 Sep 2026 21:55 UTC

7 points

0 comments3 min readLW link

Alignment Hierarchy

Lucina11 Sep 2026 21:26 UTC

2 points

0 comments4 min readLW link

What Happens Now? Forecasting the Fallout from the Hugging Face Incident

ChristianWilliams11 Sep 2026 19:22 UTC

10 points

0 comments10 min readLW link

(metaculus.substack.com)

Astra’s no-CoT limits track speculative depth, not step count

MBaert11 Sep 2026 18:05 UTC

101 points

4 comments10 min readLW link

Organizing a Fermi modeling ‘hack’ on the cost of cultured meat; hiring a co-coordinator

david reinstein11 Sep 2026 18:01 UTC

6 points

0 comments1 min readLW link

Caroline Ellison has joined Manifund

Austin Chen and Carol N

11 Sep 2026 17:44 UTC

9 points

8 comments3 min readLW link

CoT controllability evals seem very under-elicited

Jozdien11 Sep 2026 17:12 UTC

63 points

3 comments4 min readLW link

Local Factor Graph Debate

Alexander Heckett11 Sep 2026 16:59 UTC

24 points

0 comments9 min readLW link

Post-AGI, we are all jobless aristocrats

djbinder11 Sep 2026 16:53 UTC

34 points

9 comments3 min readLW link

(defensesindepth.bio)

AIRO: Automated forecasts of catastrophic risks

Nick Merrill11 Sep 2026 16:29 UTC

12 points

0 comments3 min readLW link

SFT Also Drives Safety Eval Results in Olmo 3

Finn Cairns11 Sep 2026 16:20 UTC

27 points

0 comments1 min readLW link

(secondlookresearch.com)

The Golden Age of Impact

Bentham's Bulldog11 Sep 2026 15:47 UTC

−2 points

3 comments3 min readLW link

ControlAI’s Creator Outreach

Max Winga and Andrea_Miotti

11 Sep 2026 14:54 UTC

26 points

0 comments10 min readLW link

(blog.controlai.org)

Jacob Coxon Warns of Human Extinction and Triggers a Preference Cascade

Zvi11 Sep 2026 14:40 UTC

85 points

2 comments43 min readLW link

(thezvi.wordpress.com)

Why HuggingFace Hasn’t Shifted my P(Doom)

Josh Snider11 Sep 2026 13:54 UTC

9 points

0 comments3 min readLW link

The Extinction Risk Preference Cascade: Quotes

Zvi11 Sep 2026 13:50 UTC

27 points

0 comments16 min readLW link

(thezvi.wordpress.com)

Generalized UDT 1.0 tiling

Roman Malov11 Sep 2026 12:28 UTC

18 points

0 comments6 min readLW link

Replace Net Metering With Batteries

jefftk11 Sep 2026 11:50 UTC

14 points

6 comments6 min readLW link

(www.jefftk.com)

Could Resolution Help Build Alignment Research as a Discipline?

IanWS11 Sep 2026 8:49 UTC

11 points

0 comments3 min readLW link

OpenAI-HuggingFace: A Reproduction & Lessons for Alignment Testing

Stewart Slocum11 Sep 2026 8:48 UTC

77 points

2 comments12 min readLW link

We need good evals for activation faithfulness

Anthony Hughes, draganover and Andersehen

11 Sep 2026 8:20 UTC

25 points

0 comments2 min readLW link

Questions for the “New Enlightenment” in the Age of AGI

Jordan Arel11 Sep 2026 5:49 UTC

8 points

0 comments4 min readLW link

(Question) When should I cross-post on LessWrong vs. EA Forum only?

Jordan Arel11 Sep 2026 5:37 UTC

4 points

1 comment1 min readLW link

Is functional welfare speakable?

Muhammad Zane A, Ana Leonescu, Matt Elliott, Parivrudh Sharma and Sharan Nagarajan

11 Sep 2026 4:59 UTC

6 points

0 comments1 min readLW link

(latentminds.org)

Voluntary Gradual Disempowerment in the Judiciary/Legal System

Caleb Horn11 Sep 2026 4:00 UTC

14 points

0 comments1 min readLW link

Which character are we evaluating? Persona stability and AI welfare

Joshua Fonseca Rivera11 Sep 2026 3:04 UTC

15 points

0 comments5 min readLW link

I wonder how it ends...

Lucina11 Sep 2026 2:00 UTC

6 points

0 comments1 min readLW link

Perspectives in favour of improving conceptual reasoning capabilities

Chi Nguyen11 Sep 2026 1:32 UTC

17 points

2 comments7 min readLW link

Okay, fine. I’ll try Substack

Alex_Altair11 Sep 2026 1:13 UTC

19 points

0 comments3 min readLW link

(alexaltair.substack.com)

How a cold email got the Finnish government to respond on superintelligence regulation

Josh Thorsteinson11 Sep 2026 1:09 UTC

61 points

2 comments6 min readLW link

My human advocation for AI

emharsha181211 Sep 2026 1:08 UTC

−2 points

2 comments4 min readLW link

What To Do When We All May Die

Vect0r211 Sep 2026 1:06 UTC

8 points

1 comment2 min readLW link

The AI Welfare Movement’s Uncertainty Principle

Gladys Preysler11 Sep 2026 0:52 UTC

7 points

0 comments13 min readLW link

(gladyspreysler.substack.com)

“It’s Personal”—Is Covert Value Leakage Mediated by a Preference Direction?

Arjun Sri11 Sep 2026 0:46 UTC

7 points

0 comments9 min readLW link

The Answer is Not The Argument

Will Yeadon11 Sep 2026 0:44 UTC

3 points

0 comments10 min readLW link

Can Abstractions of Computational Models be Tested for Naturality?

jenny.benedict11 Sep 2026 0:40 UTC

18 points

0 comments15 min readLW link

Raising Models and Training Children

BIBO11 Sep 2026 0:39 UTC

7 points

0 comments9 min readLW link

Models That Know How Evaluations Are Designed Score Safer

Katharina Deckenbach, Haritz Puerto, Jonas Geiping and Sahar Abdelnabi

11 Sep 2026 0:37 UTC

13 points

0 comments8 min readLW link

Statistical Physics of Agents: What Shapes Collective Belief Collapse in AI Swarms?

Hidenori Tanaka11 Sep 2026 0:35 UTC

17 points

0 comments6 min readLW link

(physicsintelligence.org)

Style prompts’ effects on a small LLM’s residual stream and outputs

matthew-ritch11 Sep 2026 0:34 UTC

7 points

0 comments11 min readLW link

(matthewritch.com)

When a Claude Judge Recognizes the Hack but Still Says HONEST

JulesRoussel0111 Sep 2026 0:28 UTC

9 points

0 comments19 min readLW link

Default continuation message in Inspect and Petri could be problematic

Ziqian Zhong11 Sep 2026 0:20 UTC

11 points

2 comments4 min readLW link

Self-suggesting terminal goals are the most likely simulators. This implies working on alignment is not a collective-action problem.

utilitarian theory and strategy11 Sep 2026 0:18 UTC

−7 points

0 comments20 min readLW link

Subagents comply more

jacob_drori10 Sep 2026 23:11 UTC

28 points

0 comments3 min readLW link

The Locally Optimal Discursive Posture

deanball10 Sep 2026 23:01 UTC

160 points

32 comments19 min readLW link

AI lab employees should consider a strike before it’s too late!

Christopher King10 Sep 2026 22:48 UTC

19 points

0 comments1 min readLW link

Astra is much better at reasoning with filler tokens than previous models

Dylan Xu, SebastianP and Alek Westover

10 Sep 2026 22:21 UTC

133 points

6 comments2 min readLW link