A sim­ple rule for causation

Vivek Hebbar24 Feb 2026 23:14 UTC
37 points
2 comments3 min readLW link

SWE-Bench Pro is even worse

Jonathan Gabor24 Feb 2026 22:51 UTC
24 points
0 comments1 min readLW link
(jonathanpgabor.substack.com)

We are all le­gal re­al­ists now

TFD24 Feb 2026 21:51 UTC
−12 points
1 comment4 min readLW link
(www.thefloatingdroid.com)

Re­spon­si­ble Scal­ing Policy v3

HoldenKarnofsky24 Feb 2026 20:20 UTC
179 points
82 comments36 min readLW link

[Question] What was the most effec­tive team you’ve ever been on, and what made it ex­cel­lent?

Eli Tyre24 Feb 2026 20:18 UTC
77 points
7 comments2 min readLW link

Why At­tack Suc­cess Rate Gives a False Pic­ture of Back­door Removal

Geoffrey Voyer24 Feb 2026 20:02 UTC
3 points
0 comments12 min readLW link

How I Started Be­ing Productive

atomic24 Feb 2026 19:49 UTC
8 points
0 comments10 min readLW link

Solv­ing The RAISE Act Like a (fic­tional) New York Detective

Josephine Schwab24 Feb 2026 19:35 UTC
3 points
1 comment6 min readLW link

Ex­clu­sive: Hegseth gives An­thropic un­til Fri­day to back down on AI safeguards

Matrice Jacobine24 Feb 2026 19:19 UTC
95 points
9 comments3 min readLW link
(www.axios.com)

Ci­garette Ads for Ba­bies from Microsoft Bing Image Generator

Edd Schneider24 Feb 2026 19:06 UTC
−4 points
1 comment4 min readLW link

Real­is­tic Eval­u­a­tions Will Not Prevent Eval­u­a­tion Awareness

Adam Karvonen24 Feb 2026 17:51 UTC
37 points
9 comments6 min readLW link

The Easiest Route to Se­cret Loy­alty May Be Hi­jack­ing the Model’s Chain of Command

Joe Kwon24 Feb 2026 17:47 UTC
16 points
1 comment5 min readLW link

Large-Scale On­line Deanonymiza­tion with LLMs

24 Feb 2026 17:02 UTC
69 points
5 comments4 min readLW link
(simonlermen.substack.com)

Open sourc­ing a browser ex­ten­sion that shows when peo­ple are wrong on the internet

lc24 Feb 2026 16:36 UTC
227 points
34 comments2 min readLW link
(github.com)

Ras­cal’s Wager

corticalcircuitry24 Feb 2026 16:13 UTC
3 points
2 comments3 min readLW link
(sergey.substack.com)

Citrini’s Sce­nario Is A Great But Deeply Flawed Thought Experiment

Zvi24 Feb 2026 15:40 UTC
37 points
6 comments22 min readLW link
(thezvi.wordpress.com)

Ob­ser­va­tions from Run­ning an Agent Collective

williawa24 Feb 2026 15:34 UTC
45 points
2 comments10 min readLW link

What is a species?

David Goodman24 Feb 2026 14:23 UTC
49 points
15 comments26 min readLW link

Mo­ral pub­lic goods are a big deal for whether we get a good future

24 Feb 2026 14:14 UTC
12 points
0 comments18 min readLW link
(www.forethought.org)

Two memos from 2024

Richard_Ngo24 Feb 2026 7:19 UTC
38 points
0 comments7 min readLW link

What is com­pu­ta­tional me­chan­ics? An ex­plainer

Leo Cymbalista24 Feb 2026 6:09 UTC
16 points
0 comments15 min readLW link

Mon­day AI Radar #14

Against Moloch24 Feb 2026 5:34 UTC
4 points
0 comments6 min readLW link
(againstmoloch.com)

The ML on­tol­ogy and the al­ign­ment ontology

Richard_Ngo24 Feb 2026 4:39 UTC
110 points
9 comments4 min readLW link

[USA To­day op-ed]: No, AI isn’t in­evitable. We should stop it while we can.

David Scott Krueger24 Feb 2026 2:05 UTC
17 points
0 comments1 min readLW link
(www.usatoday.com)

Bioan­chors 2: Elec­tric Bacilli

TsviBT24 Feb 2026 1:07 UTC
38 points
1 comment7 min readLW link

Sin­gle Stack LLMs are Split-Brain Pa­tients.

niceminus1924 Feb 2026 0:04 UTC
5 points
0 comments3 min readLW link

Us­ing fic­tion to imag­ine a path­way to friendlyAGI

Rick Moss23 Feb 2026 23:48 UTC
3 points
0 comments2 min readLW link

When Bench­marks Lie: Eval­u­at­ing Mal­i­cious Prompt Clas­sifiers Un­der True Distri­bu­tion Shift

Max Fomin23 Feb 2026 23:44 UTC
1 point
2 comments6 min readLW link

The per­sona se­lec­tion model

Sam Marks23 Feb 2026 22:56 UTC
176 points
53 comments43 min readLW link
(alignment.anthropic.com)

Agenda Reflec­tion: Test­ing Au­to­mated Align­ment

Ariel_23 Feb 2026 21:53 UTC
11 points
0 comments2 min readLW link
(zenodo.org)

Claude Son­net 4.6 Gives You Flexibility

Zvi23 Feb 2026 20:30 UTC
29 points
1 comment9 min readLW link
(thezvi.wordpress.com)

Se­crets of the LessWrong RSS Feed

Brendan Long23 Feb 2026 20:12 UTC
36 points
6 comments4 min readLW link

Which ques­tions can’t we punt?

Lizka23 Feb 2026 19:17 UTC
39 points
2 comments15 min readLW link

Ex­po­nen­tial GDP growth from lin­ear growth in va­ri­ety of goods

Will_Howard23 Feb 2026 18:50 UTC
4 points
2 comments5 min readLW link
(open.substack.com)

Pre-train­ing data poi­son­ing likely makes in­stal­ling se­cret loy­alties easier

Joe Kwon23 Feb 2026 18:12 UTC
12 points
0 comments4 min readLW link

The 2028 Global In­tel­li­gence Cri­sis—a fi­nance-ori­ented vignette

Rasool23 Feb 2026 17:12 UTC
50 points
13 comments1 min readLW link
(www.citriniresearch.com)

AI Im­pact Sum­mit 2026 : A Field Report

23 Feb 2026 16:58 UTC
38 points
1 comment9 min readLW link

The map of the map is not the map

jimmy23 Feb 2026 16:54 UTC
18 points
3 comments9 min readLW link

Fact-check­ing an AI op­ti­mist ar­ti­cle in The Economist

ToSummarise23 Feb 2026 13:56 UTC
41 points
3 comments4 min readLW link
(www.tosummarise.com)

Re­view: “We can’t dis­agree for­ever”

Martin Randall23 Feb 2026 13:17 UTC
15 points
0 comments3 min readLW link

Why I Think Pause is Impossible

E.G. Blee-Goldman23 Feb 2026 11:58 UTC
1 point
4 comments6 min readLW link

Can Aha Mo­ments be Fake? Iden­ti­fy­ing True and Dec­o­ra­tive Think­ing Steps in CoT

Jiachen Zhao23 Feb 2026 11:51 UTC
24 points
0 comments10 min readLW link
(arxiv.org)

A World Without Vio­let: Pe­cu­liar Con­se­quences of Grant­ing Mo­ral Sta­tus to Ar­tifi­cial Intelligences

Sever Topan23 Feb 2026 7:23 UTC
17 points
8 comments4 min readLW link
(severtopan.substack.com)

Was It Owl a Dream?

Yovel Rom23 Feb 2026 5:07 UTC
17 points
4 comments4 min readLW link
(yovelrom.substack.com)

In­nate Immunity

joec23 Feb 2026 5:00 UTC
23 points
2 comments6 min readLW link

Why I Tran­si­tioned: A Third (FtM) Perspective

Character#273623 Feb 2026 4:39 UTC
22 points
6 comments14 min readLW link

The power of a sim­ple 3-way truth scale

Bruce Lewis23 Feb 2026 2:41 UTC
4 points
2 comments2 min readLW link

Stor­ing Food

jefftk23 Feb 2026 1:40 UTC
77 points
9 comments2 min readLW link
(www.jefftk.com)

Old SUNY Dorm Logic is not helping ru­ral pop­u­la­tion col­lapse in NY.

Edd Schneider23 Feb 2026 1:28 UTC
9 points
4 comments3 min readLW link

Chang­ing the world for the worse

mingyuan22 Feb 2026 23:55 UTC
129 points
17 comments3 min readLW link
(mingyuan.substack.com)