All200520062007200820092010201120122013201420152016201720182019202020212022202320242025
AllJanFebMarAprMayJun
All12345678910111213141516171819202122232425262728293031

Claude 4, Op­por­tunis­tic Black­mail, and “Pleas”

Stephen MartinMay 22, 2025, 7:59 PM
24 points
1 comment2 min readLW link

Prob­lems in AI Align­ment: A Scale Model

Mickey MuldoonMay 22, 2025, 7:22 PM
−1 points
3 comments2 min readLW link
(muldoon.cloud)

Art Is Art: AI Is the Next Erotica

Charlie EdwardsMay 22, 2025, 6:04 PM
0 points
1 comment14 min readLW link

Re­ward but­ton alignment

Steven ByrnesMay 22, 2025, 5:36 PM
50 points
15 comments12 min readLW link

We’re Not Ad­ver­tis­ing Enough (Post 3 of 6 on AI Gover­nance)

Mass_DriverMay 22, 2025, 5:05 PM
107 points
10 comments28 min readLW link

Claude 4

Zach Stein-PerlmanMay 22, 2025, 5:00 PM
71 points
24 comments1 min readLW link
(www.anthropic.com)

Video and tran­script of talk on AI welfare

Joe CarlsmithMay 22, 2025, 4:15 PM
24 points
1 comment28 min readLW link
(joecarlsmith.substack.com)

What we can learn from af­ter­life myths

jchanMay 22, 2025, 3:49 PM
5 points
0 comments15 min readLW link

Policy recom­men­da­tions re­gard­ing re­pro­duc­tive technology

TsviBTMay 22, 2025, 2:49 PM
76 points
2 comments3 min readLW link

AI #117: OpenAI Buys De­vice Maker IO

ZviMay 22, 2025, 1:40 PM
37 points
9 comments62 min readLW link
(thezvi.wordpress.com)

Does BPC-157 work for heal­ing and tis­sue re­pair?

ChristianKlMay 22, 2025, 1:18 PM
17 points
0 comments5 min readLW link
(somaticsignals.jollyjoyjourney.com)

[Question] How load-bear­ing is KL di­ver­gence from a known-good base model in mod­ern RL?

faul_snameMay 22, 2025, 12:08 PM
12 points
3 comments4 min readLW link

Chris­ti­an­ity vs. Tantra vs. Sex – one spiritual path?

pchvykovMay 22, 2025, 11:15 AM
−2 points
0 comments24 min readLW link

Mir­ror Or­ganisms Are Not Im­mune to Predation

Matthias DellagoMay 22, 2025, 11:10 AM
27 points
5 comments1 min readLW link

How 2025 AI Fore­casts Fared So Far

Adam B, romeo and elifland
May 22, 2025, 9:42 AM
11 points
2 comments8 min readLW link
(theaidigest.org)

Con­tain and ver­ify: The endgame of US-China AI competition

sjadlerMay 22, 2025, 8:13 AM
5 points
6 comments2 min readLW link
(open.substack.com)

Laugencroissant

Martin SustrikMay 22, 2025, 6:30 AM
13 points
0 comments3 min readLW link
(250bpm.substack.com)

Google I/​O Day

ZviMay 21, 2025, 10:00 PM
49 points
0 comments20 min readLW link
(thezvi.wordpress.com)

Pod­cast: How not to waste a billion dol­lars (on your clini­cal trial), with Meri Beck­with on Devel­op­ment & Research

rossryMay 21, 2025, 9:27 PM
25 points
0 comments3 min readLW link
(developmentandresearch.bio)

Pod­cast: From molecule to medicine, with Ross Rhe­in­gans-Yoo on Com­plex Systems

rossryMay 21, 2025, 9:08 PM
15 points
0 comments5 min readLW link
(www.complexsystemspodcast.com)

The stakes of AI moral status

Joe CarlsmithMay 21, 2025, 6:20 PM
78 points
62 comments14 min readLW link
(joecarlsmith.substack.com)

[Question] Which AI Safety tech­niques will be in­effec­tive against diffu­sion mod­els?

Allen ThomasMay 21, 2025, 6:13 PM
1 point
0 comments1 min readLW link

Through The Look­ing Glasses: Is­sues & Solu­tions for Aug­mented Reality

claywrenMay 21, 2025, 6:11 PM
1 point
0 comments22 min readLW link

Root­ing for Mo­ments, Not Jerseys. Another Ap­proach to En­joy­ing Sports

Ahmed ElsayyadMay 21, 2025, 6:11 PM
1 point
0 comments3 min readLW link

Un­ex­ploitable search: block­ing mal­i­cious use of free parameters

Jacob Pfau and Geoffrey Irving
May 21, 2025, 5:23 PM
34 points
16 comments6 min readLW link

The Real AI Safety Risk Is a Con­cep­tual Ex­ploit: Anthropomorphism

Anthony FoxMay 21, 2025, 4:29 PM
−2 points
0 comments2 min readLW link

You Can’t Skip Ex­plo­ra­tion: Why un­der­stand­ing ex­per­i­men­ta­tion and taste is key to un­der­stand­ing AI

Oliver SourbutMay 21, 2025, 4:08 PM
18 points
0 comments11 min readLW link
(www.oliversourbut.net)

The Prob­lem and Op­por­tu­nity of Scale

belosMay 21, 2025, 3:52 PM
1 point
0 comments5 min readLW link
(bestofagreatlot.substack.com)

Sleep need re­duc­tion therapies

harsimonyMay 21, 2025, 3:22 PM
75 points
18 comments10 min readLW link
(splittinginfinity.substack.com)

Parental Guidance: Fram­ing Su­per­in­tel­li­gence

ejk64May 21, 2025, 3:01 PM
10 points
0 comments3 min readLW link

Why Aren’t Ra­tion­al­ists Win­ning (Again)

k64May 21, 2025, 2:46 PM
6 points
25 comments5 min readLW link

Can We Nat­u­ral­ize Mo­ral Episte­mol­ogy?

tylermjohn May 21, 2025, 2:25 PM
50 points
22 comments6 min readLW link

Units Have More Depth Than I Thought

MorpheusMay 21, 2025, 1:51 PM
31 points
5 comments1 min readLW link

Hu­mans are Inse­cure Pass­word Generators

Isaac KingMay 21, 2025, 5:58 AM
15 points
0 comments5 min readLW link

[Cross­post] An­thropic Shadow Geopolitics

akarlinMay 21, 2025, 4:50 AM
8 points
5 comments18 min readLW link

The Need for Poli­ti­cal Ad­ver­tis­ing (Post 2 of 6 on AI Gover­nance)

Mass_DriverMay 21, 2025, 12:44 AM
54 points
2 comments13 min readLW link

Notes from Dopamine Detoxing

Alice BlairMay 20, 2025, 11:43 PM
13 points
2 comments9 min readLW link

Re­vis­it­ing the ideas for non-neu­ralese architectures

StanislavKrymMay 20, 2025, 11:35 PM
2 points
0 comments1 min readLW link

Gem­ini Diffu­sion: watch this space

Yair HalberstadtMay 20, 2025, 7:29 PM
190 points
35 comments1 min readLW link
(deepmind.google)

A Sketch of Be­loc­racy: a new sys­tem of governance

belosMay 20, 2025, 6:30 PM
5 points
0 comments8 min readLW link
(bestofagreatlot.substack.com)

The Codex of Ul­ti­mate Vibing

ZviMay 20, 2025, 6:30 PM
45 points
2 comments11 min readLW link
(thezvi.wordpress.com)

Out­comes of the Geopoli­ti­cal Singularity

Nikola JurkovicMay 20, 2025, 6:09 PM
61 points
5 comments5 min readLW link

AISN #55: Trump Ad­minis­tra­tion Re­scinds AI Diffu­sion Rule, Allows Chip Sales to Gulf States

Corin Katzke and Dan H
20 May 2025 16:21 UTC
4 points
1 comment4 min readLW link
(forum.effectivealtruism.org)

Time isn’t real man

Jodie themathgenius20 May 2025 15:48 UTC
3 points
2 comments1 min readLW link
(jodie.website)

US Govt Whistle­blower guide (in­com­plete draft)

samuelshadrach20 May 2025 15:34 UTC
−3 points
15 comments24 min readLW link

If one sur­viv­ing civ­i­liza­tion can res­cue oth­ers, shouldn’t civ­i­liza­tions ran­dom­ize?

Knight Lee20 May 2025 15:26 UTC
−2 points
4 comments1 min readLW link

The Tox­i­c­ity of Me­ta­mod­ernism: A Public Ser­vice Announcement

henophilia20 May 2025 15:05 UTC
−19 points
4 comments3 min readLW link
(blog.hermesloom.org)

Pres­i­dent of Euro­pean Com­mis­sion ex­pects hu­man-level AI by 2026

sanyer20 May 2025 14:13 UTC
35 points
4 comments1 min readLW link
(ec.europa.eu)

Selec­tive reg­u­lariza­tion for al­ign­ment-fo­cused rep­re­sen­ta­tion engineering

Sandy Fraser20 May 2025 12:54 UTC
21 points
3 comments12 min readLW link

Short­est damn doom­splainer in world history

lemonhope20 May 2025 9:07 UTC
2 points
6 comments1 min readLW link