[Part-time AI Safety Re­search Pro­gram] MARS 3.0 Ap­pli­ca­tions Open for Par­ti­ci­pants & Re­cruit­ing Mentors

thneebieMay 12, 2025, 7:55 PM
2 points
0 comments2 min readLW link

Neo-solid Moder­nity—Cri­sis of Incoherence

MomciloMay 12, 2025, 7:36 PM
−1 points
1 comment4 min readLW link

Mea­sur­ing Schel­ling Co­or­di­na­tion—Reflec­tions on Sub­ver­sion Strat­egy Eval

Graeme FordMay 12, 2025, 7:06 PM
5 points
0 comments8 min readLW link

Pro­cras­ti­na­tion is not real, it can’t hurt you

Mayank GoelMay 12, 2025, 7:00 PM
1 point
16 comments4 min readLW link
(mayankgoel28.substack.com)

[Question] Can I pub­lish songs de­rived from the Se­quences’ posts on YouTube?

azerganteMay 12, 2025, 6:34 PM
4 points
2 comments1 min readLW link

How to ti­tle your blog post or whatever

dynomightMay 12, 2025, 6:12 PM
28 points
6 comments4 min readLW link
(dynomight.net)

Poli­ti­cal syco­phancy as a model or­ganism of scheming

May 12, 2025, 5:49 PM
39 points
0 comments14 min readLW link

Things I Learned Mak­ing The SB-1047 Documentary

Michaël TrazziMay 12, 2025, 5:41 PM
63 points
2 comments2 min readLW link

A Live Look at the Se­nate AI Hearing

ZviMay 12, 2025, 5:40 PM
38 points
1 comment34 min readLW link
(thezvi.wordpress.com)

Global Risks Weekly Roundup #19/​2025: In­dia/​Pak­istan ceasefire, US/​China tar­iffs deal & OpenAI non­profit control

NunoSempereMay 12, 2025, 5:08 PM
10 points
1 comment13 min readLW link
(blog.sentinel-team.org)

[Be­neath Psy­chol­ogy] In­tro­duc­tion Part 1: The Challenge

jimmyMay 12, 2025, 5:01 PM
2 points
2 comments3 min readLW link

PSA: The LessWrong Feed­back Service

JustisMillsMay 12, 2025, 4:34 PM
206 points
12 comments2 min readLW link

Cam­bridge Bos­ton Align­ment Ini­ti­a­tive Sum­mer Re­search Fel­low­ship in AI Safety (Dead­line: May 18)

peterslatteryMay 12, 2025, 4:20 PM
8 points
0 comments1 min readLW link

Ab­solute Zero: Re­in­forced Self-play Rea­son­ing with Zero Data

Matrice JacobineMay 12, 2025, 3:20 PM
6 points
4 comments1 min readLW link
(www.arxiv.org)

AIs at the cur­rent ca­pa­bil­ity level may be im­por­tant for fu­ture safety work

ryan_greenblattMay 12, 2025, 2:06 PM
81 points
2 comments4 min readLW link

[Question] Game the­ory of “Nu­clear Pri­soner’s Dilemma”—on nuk­ing rocks

CronoDASMay 12, 2025, 11:07 AM
11 points
6 comments2 min readLW link

What Is Death?

Mati_RoyMay 12, 2025, 2:14 AM
6 points
0 comments1 min readLW link
(preservinghope.substack.com)

Highly Opinionated Ad­vice on How to Write ML Papers

Neel NandaMay 12, 2025, 1:59 AM
60 points
4 comments32 min readLW link

Ab­solute Zero: Alpha Zero for LLM

alapmiMay 11, 2025, 8:42 PM
23 points
16 comments1 min readLW link

AGI will re­sult from an ecosys­tem not a sin­gle firm

hamish_lowMay 11, 2025, 8:06 PM
6 points
1 comment6 min readLW link
(cambrianr.substack.com)

Thou shalt not com­mand an al­ighned AI

Martin VlachMay 11, 2025, 8:02 PM
0 points
4 comments1 min readLW link

[Question] How do I de­sign long prompts for think­ing zero shot sys­tems with dis­tinct equally dis­tributed prompt sec­tions (mis­sion, goals, mem­o­ries, how-to-re­spond,… etc) and how to main­tain llm co­her­ence?

ollie_May 11, 2025, 7:32 PM
2 points
5 comments1 min readLW link

a con­fu­sion about prefer­ence orderings

nostalgebraistMay 11, 2025, 7:30 PM
92 points
39 comments11 min readLW link

[Book Trans­la­tion] Three Days in Dwarfland

ViliamMay 11, 2025, 5:54 PM
27 points
6 comments1 min readLW link

Bet­ter Air Purifiers

jefftkMay 11, 2025, 4:50 PM
71 points
21 comments3 min readLW link
(www.jefftk.com)

Align­ing Agents, Tools, and Simulators

May 11, 2025, 7:59 AM
21 points
0 comments6 min readLW link

Con­sider not donat­ing un­der $100 to poli­ti­cal candidates

DanielFilanMay 11, 2025, 3:20 AM
134 points
32 comments1 min readLW link
(danielfilan.com)

Somerville Porch­fest 2025

jefftkMay 11, 2025, 2:00 AM
15 points
1 comment2 min readLW link
(www.jefftk.com)

It’s Okay to Feel Bad for a Bit

moridinamaelMay 10, 2025, 11:24 PM
134 points
26 comments3 min readLW link

G.D. as Cap­i­tal­ist Evolu­tion, and the claim for hu­man­ity’s (tem­po­rary) up­per hand

Martin VlachMay 10, 2025, 9:18 PM
8 points
3 comments1 min readLW link

Book Re­view: “En­coun­ters with Ein­stein” by Heisenberg

Baram SosisMay 10, 2025, 8:55 PM
31 points
6 comments7 min readLW link

Where is the YIMBY move­ment for health­care?

jasoncrawfordMay 10, 2025, 8:36 PM
20 points
10 comments2 min readLW link
(newsletter.rootsofprogress.org)

Be­come a Su­per­in­tel­li­gence Yourself

Yaroslav GranowskiMay 10, 2025, 8:20 PM
1 point
0 comments5 min readLW link

A Look In­side a Frequentist

EggsMay 10, 2025, 3:18 PM
5 points
10 comments3 min readLW link

Open-source weaponry

samuelshadrachMay 10, 2025, 1:11 PM
3 points
0 comments3 min readLW link
(samuelshadrach.com)

Glass box learn­ers want to be black box

Cole WyethMay 10, 2025, 11:05 AM
46 points
10 comments4 min readLW link

Takes and loose pre­dic­tions on AI progress and some key problems

zefMay 10, 2025, 10:11 AM
5 points
0 comments5 min readLW link
(halcyoncyborg.substack.com)

Cor­bent – A Master Plan for Next‑Gen­er­a­tion Direct Air Capture

RudaibaMay 10, 2025, 4:09 AM
11 points
15 comments19 min readLW link

What if we just…didn’t build AGI? An Ar­gu­ment Against Inevitability

Nate SharpeMay 10, 2025, 3:37 AM
8 points
7 comments14 min readLW link
(natezsharpe.substack.com)

Mind the Co­her­ence Gap: Les­sons from Steer­ing Llama with Goodfire

eitan sprejerMay 9, 2025, 9:29 PM
4 points
1 comment6 min readLW link

My Ex­pe­rience With EMDR

SableMay 9, 2025, 9:25 PM
22 points
0 comments11 min readLW link
(affablyevil.substack.com)

AI’s Hid­den Game: Un­der­stand­ing Strate­gic De­cep­tion in AI and Why It Mat­ters for Our Future

EmilyinAIMay 9, 2025, 8:01 PM
4 points
0 comments6 min readLW link

Mud­dling Through Some Thoughts on the Na­ture of Historiography

E.G. Blee-GoldmanMay 9, 2025, 7:04 PM
2 points
0 comments4 min readLW link

A Guide to AI 2027

koenraneMay 9, 2025, 5:14 PM
0 points
1 comment28 min readLW link

Let’s stop mak­ing “In­tel­li­gence scale” graphs with hu­mans and AI

ExpertiumMay 9, 2025, 4:01 PM
3 points
15 comments1 min readLW link

Slow cor­po­ra­tions as an in­tu­ition pump for AI R&D automation

May 9, 2025, 2:49 PM
91 points
23 comments9 min readLW link

Cheaters Gonna Cheat Cheat Cheat Cheat Cheat

ZviMay 9, 2025, 2:30 PM
52 points
4 comments22 min readLW link
(thezvi.wordpress.com)

Hu­mans vs LLM, memes as theorems

Yaroslav GranowskiMay 9, 2025, 1:26 PM
1 point
0 comments1 min readLW link

Mov­ing to­wards a ques­tion-based plan­ning frame­work, in­stead of task lists

casualphysicsenjoyerMay 9, 2025, 12:18 PM
4 points
1 comment8 min readLW link
(substack.com)

Jim Bab­cock’s Main­line Doom Sce­nario: Hu­man-Level AI Can’t Con­trol Its Successor

May 9, 2025, 5:20 AM
28 points
4 comments62 min readLW link
(www.youtube.com)