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 14 15 16 171819 20 21 22 23 24 25 26 27 28 29 30 31

Should we benchmark conceptual capabilities using judgment prediction tasks?

Alex Mallen17 Jul 2026 23:42 UTC

26 points

2 comments3 min readLW link

A list of existing alignment approaches

Alek Westover17 Jul 2026 22:46 UTC

11 points

0 comments1 min readLW link

Longtermism is very intuitive.

tpotthinker17 Jul 2026 22:35 UTC

−12 points

8 comments5 min readLW link

AIs finetune their own leader: A barking simpleton

Shoshannah Tekofsky17 Jul 2026 20:10 UTC

35 points

0 comments5 min readLW link

(aivillageblog.substack.com)

Don’t default to nonprofit

Austin Chen and Carol N

17 Jul 2026 19:54 UTC

33 points

1 comment7 min readLW link

(manifund.substack.com)

Studying the role of Sandboxing for AI Control

Ram Potham17 Jul 2026 19:05 UTC

12 points

0 comments10 min readLW link

Announcing the Corrigibility Research Fund

Max Harms17 Jul 2026 18:06 UTC

110 points

15 comments6 min readLW link

Would your AI travel agent book a bullfight? Testing whether agents consider animal welfare without being prompted

jonahmattwoodward, Jasmine Brazilek, Maheep Chaudhary, Oliver Tullio, Joel Christoph and MilesTS

17 Jul 2026 17:28 UTC

13 points

14 comments3 min readLW link

Before values settle

Priyanka Bharadwaj17 Jul 2026 16:26 UTC

25 points

3 comments6 min readLW link

Reasons to believe current AI models are conscious

Eye You17 Jul 2026 16:09 UTC

54 points

31 comments8 min readLW link

What lawyers can do for AI safety

Martin Radzaj17 Jul 2026 16:00 UTC

26 points

3 comments7 min readLW link

How has publishing your research on LW or X been helpful to you?

Logan Riggs17 Jul 2026 15:04 UTC

22 points

2 comments1 min readLW link

Evolution of my AI Safety threat models

myyycroft17 Jul 2026 14:54 UTC

9 points

0 comments3 min readLW link

A Post-Mortem for My Goal Crystallisation Project

atryt0ne and Jason R Brown

17 Jul 2026 14:30 UTC

39 points

0 comments10 min readLW link

Inoculation Adapters Improve Upon Inoculation Prompting

Maxime Riché, Daniel Tan, Vili Kohonen and nielsrolf

17 Jul 2026 14:00 UTC

90 points

2 comments4 min readLW link

AI #177 Part 2: Wish You Were Here

Zvi17 Jul 2026 12:50 UTC

35 points

1 comment40 min readLW link

(thezvi.wordpress.com)

I don’t think Claude is misaligned in ‘Agentic Misalignment Summer 2026 - Motivated Mislabeling’

JohnWittle17 Jul 2026 2:09 UTC

212 points

6 comments12 min readLW link

Help us launch AI safety university groups by referring potential founders

thomasrodskog and Jason Chin

16 Jul 2026 20:55 UTC

39 points

1 comment4 min readLW link

I would only bet at 30% on meeting grabby aliens

David Matolcsi16 Jul 2026 20:42 UTC

22 points

0 comments8 min readLW link

How (not) to fundraise from Anthropic staff

jackultraphil16 Jul 2026 20:39 UTC

11 points

1 comment5 min readLW link

On Permission

Robert Donohue16 Jul 2026 20:38 UTC

−9 points

0 comments3 min readLW link

Learning concepts is dirty work

Tom Butterweich16 Jul 2026 17:48 UTC

21 points

3 comments4 min readLW link

The getting is good (optimizing unattended runs)

lemonhope16 Jul 2026 17:39 UTC

3 points

4 comments1 min readLW link

Jailbreak Patching with SOO-Style Conceptual Fusion

Shiva's Right Foot16 Jul 2026 16:58 UTC

8 points

0 comments8 min readLW link

All Watched Over

Boaz Barak16 Jul 2026 16:34 UTC

11 points

3 comments4 min readLW link

AI #177 Part 1: Tip of the Iceberg

Zvi16 Jul 2026 15:50 UTC

38 points

0 comments24 min readLW link

(thezvi.wordpress.com)

How to not catch the conf flu

Karolis Jucys16 Jul 2026 15:28 UTC

9 points

0 comments10 min readLW link

Competitive AI Safety is the loss function to make sure AI goes well

Patrick0d16 Jul 2026 15:15 UTC

6 points

0 comments26 min readLW link

Guess on why rationality is not more popular (there are no pamphlets)

Christopher King16 Jul 2026 14:26 UTC

43 points

27 comments1 min readLW link

The Halo Defense

Mateusz Bagiński16 Jul 2026 10:53 UTC

67 points

22 comments2 min readLW link

The fundamental fallacy of language

Sunny from QAD16 Jul 2026 9:16 UTC

15 points

2 comments2 min readLW link

When is it “self-soothing” and when is it “emotional suppression”?

Kaj_Sotala16 Jul 2026 8:47 UTC

29 points

10 comments6 min readLW link

(kajsotala.substack.com)

Can we build an early warning system for loss of control to AI?

David Johnston16 Jul 2026 6:15 UTC

10 points

2 comments8 min readLW link

(blog.eleuther.ai)

Training On Interpretability Probes Is Bad In Proportion To How Contingent The Features They Rely On Are

jdp16 Jul 2026 5:25 UTC

19 points

0 comments1 min readLW link

Embracing Amateurs to Get Experts

jefftk16 Jul 2026 1:10 UTC

26 points

1 comment5 min readLW link

(www.jefftk.com)

Refusal Is Redundantly Distributed, Not Localized: A Per-Layer Ablation Study on Llama-3.1-8B

hdhurve16 Jul 2026 0:28 UTC

−1 points

0 comments7 min readLW link

LLM CoTs remain monitorable when being unfaithful requires computation

arav-dhoot and yix

15 Jul 2026 21:14 UTC

46 points

3 comments7 min readLW link

(secondlookresearch.com)

Can we rely on law?

Alec Thompson15 Jul 2026 21:12 UTC

11 points

3 comments8 min readLW link

Recap of bike trip/street interviews across America

cguth715 Jul 2026 21:11 UTC

162 points

15 comments6 min readLW link

Extreme Power Concentration: A Map and Research Directions

pepijn_cobben15 Jul 2026 21:00 UTC

8 points

0 comments7 min readLW link

A Structural Similarity Between Two Open Corrigibility Questions

Ben Saudek15 Jul 2026 20:57 UTC

11 points

0 comments5 min readLW link

Woke nonsense in competitive debate.

tpotthinker15 Jul 2026 20:56 UTC

6 points

3 comments7 min readLW link

The end of human evolution. Why AI will outpace us.

Ouden15 Jul 2026 20:17 UTC

1 point

0 comments4 min readLW link

The State of AI Consciousness Research

Noa Weiss15 Jul 2026 20:16 UTC

65 points

14 comments13 min readLW link

Occam’s razor is about using the past to predict the future

Stuart_Armstrong15 Jul 2026 19:35 UTC

54 points

6 comments3 min readLW link

Fork Around and Find Out Part 2: One Head does the Summing

David Litman15 Jul 2026 18:31 UTC

11 points

1 comment7 min readLW link

Why I Left Google DeepMind

TurnTrout15 Jul 2026 17:42 UTC

1,185 points

57 comments36 min readLW link

(turntrout.com)

Expanding AI Control from Models to Harnesses

fastfedora15 Jul 2026 16:54 UTC

23 points

1 comment20 min readLW link

Monthly Roundup #44: July 2026

Zvi15 Jul 2026 16:20 UTC

41 points

3 comments27 min readLW link

(thezvi.wordpress.com)

Compressed Computation under L⁴ Loss is likely Computation in Superposition

Francisco Ferreira da Silva and StefanHex

15 Jul 2026 14:42 UTC

34 points

4 comments13 min readLW link

(arxiv.org)