Burnout, de­pres­sion, and AI safety: some con­crete men­tal health strategies

KatWoods26 Dec 2025 19:52 UTC
44 points
2 comments4 min readLW link

How hard should I pri­ori­tize hav­ing kids?

Recurrented26 Dec 2025 19:29 UTC
11 points
7 comments1 min readLW link

The moral critic of the AI in­dus­try—a Q&A with Holly Elmore

Mordechai Rorvig26 Dec 2025 17:49 UTC
7 points
0 comments2 min readLW link
(www.foommagazine.org)

Ap­ply for Align­ment Men­tor­ship from TurnTrout and Alex Cloud

26 Dec 2025 17:20 UTC
40 points
0 comments2 min readLW link
(turntrout.com)

Mea­sur­ing no CoT math time hori­zon (sin­gle for­ward pass)

ryan_greenblatt26 Dec 2025 16:37 UTC
212 points
18 comments3 min readLW link

Whole Brain Emu­la­tion as an An­chor for AI Welfare

sturb26 Dec 2025 14:45 UTC
52 points
13 comments6 min readLW link

Child­hood and Ed­u­ca­tion #16: Let­ting Kids Be Kids

Zvi26 Dec 2025 13:50 UTC
55 points
3 comments18 min readLW link
(thezvi.wordpress.com)

Re­gres­sion by Composition

Anders_H26 Dec 2025 12:18 UTC
13 points
0 comments1 min readLW link
(rss.org.uk)

Un­known Knowns: Five Ideas You Can’t Unsee

Linch25 Dec 2025 23:28 UTC
76 points
37 comments6 min readLW link
(linch.substack.com)

There’s Room in the Manger

Celer25 Dec 2025 18:00 UTC
20 points
0 comments2 min readLW link
(keller.substack.com)

Call for Science of Eval Aware­ness (+ Re­search Direc­tions)

Igor Ivanov25 Dec 2025 17:26 UTC
29 points
23 comments5 min readLW link

AI #148: Christ­mas Break

Zvi25 Dec 2025 14:00 UTC
30 points
4 comments39 min readLW link
(thezvi.wordpress.com)

Clip­board Normalization

jefftk25 Dec 2025 13:50 UTC
105 points
9 comments1 min readLW link
(www.jefftk.com)

The In­tel­li­gence Axis: A Func­tional Ty­pol­ogy

Anurag 25 Dec 2025 12:18 UTC
3 points
0 comments5 min readLW link

Honor­able AI

Kaarel24 Dec 2025 21:20 UTC
37 points
23 comments41 min readLW link

Catch-Up Al­gorith­mic Progress Might Ac­tu­ally be 60× per Year

Aaron_Scher24 Dec 2025 21:03 UTC
92 points
16 comments10 min readLW link

The Ones who Feed their Children

xhnk7jwvqj-max24 Dec 2025 19:15 UTC
10 points
2 comments3 min readLW link

[Book Re­view] “Real­ity+” by David Chalmers

lsdev24 Dec 2025 19:14 UTC
4 points
0 comments2 min readLW link

Kids and Space

jefftk24 Dec 2025 15:30 UTC
73 points
5 comments3 min readLW link
(www.jefftk.com)

Zvi’s 2025 In Movies

Zvi24 Dec 2025 13:30 UTC
27 points
1 comment11 min readLW link
(thezvi.wordpress.com)

[Question] Acausal com­mu­ni­ca­tion be­tween iso­lated uni­verses through simulation

Horosphere24 Dec 2025 11:46 UTC
13 points
14 comments1 min readLW link

Method­olog­i­cal con­sid­er­a­tions in mak­ing ma­lign ini­tial­iza­tions for con­trol research

Alek Westover24 Dec 2025 1:18 UTC
10 points
0 comments13 min readLW link

Im­mun­odefi­ciency to Par­a­sitic AI

Andrii Shportko24 Dec 2025 0:17 UTC
4 points
1 comment2 min readLW link

An in­tro­duc­tion to mod­u­lar in­duc­tion and some at­tempts to solve it

Thomas Kehrenberg23 Dec 2025 22:35 UTC
12 points
1 comment18 min readLW link

Rules clar­ifi­ca­tion for the Write like lsusr competition

Isusr23 Dec 2025 21:12 UTC
8 points
2 comments2 min readLW link

Hu­man Values

Maitreya23 Dec 2025 21:08 UTC
32 points
1 comment3 min readLW link

Align­ment Fellowship

rich_anon23 Dec 2025 20:29 UTC
58 points
14 comments1 min readLW link

Iter­a­tive Ma­trix Steer­ing: Forc­ing LLMs to “Ra­tion­al­ize” Hal­lu­ci­na­tions via Sub­space Alignment

Artem Herasymenko23 Dec 2025 20:13 UTC
9 points
2 comments4 min readLW link

Un­pack­ing Geo­met­ric Rationality

MorgneticField23 Dec 2025 20:10 UTC
2 points
0 comments33 min readLW link

Dream­ing Vec­tors: Gra­di­ent-de­scented steer­ing vec­tors from Ac­ti­va­tion Or­a­cles and us­ing them to Red-Team AOs

ceselder23 Dec 2025 19:28 UTC
22 points
4 comments12 min readLW link

The Cen­ter for Re­duc­ing Suffer­ing wants in­put from the suffer­ing re­duc­tion community

Zoé23 Dec 2025 18:27 UTC
1 point
0 comments1 min readLW link
(centerforreducingsuffering.org)

It’s Good To Create Happy Peo­ple: A Com­pre­hen­sive Case

Bentham's Bulldog23 Dec 2025 16:43 UTC
1 point
5 comments33 min readLW link

I Died on DMT

Rebecca Dai23 Dec 2025 16:15 UTC
12 points
2 comments7 min readLW link
(rebeccadai.substack.com)

Open Source is a Nor­mal Term

jefftk23 Dec 2025 15:40 UTC
24 points
4 comments1 min readLW link
(www.jefftk.com)

Don’t Trust Your Brain

silentbob23 Dec 2025 15:06 UTC
37 points
5 comments4 min readLW link

The ML drug dis­cov­ery startup try­ing re­ally, re­ally hard to not cheat

Abhishaike Mahajan23 Dec 2025 14:48 UTC
86 points
2 comments19 min readLW link
(www.owlposting.com)

Keep­ing Up Against the Jone­ses: Balsa’s 2025 Fundraiser

Zvi23 Dec 2025 14:40 UTC
49 points
1 comment6 min readLW link
(thezvi.wordpress.com)

Does 1025 mod­ulo 57 equal 59?

Jan Betley23 Dec 2025 13:00 UTC
33 points
3 comments2 min readLW link

What Can Wittgen­stein Teach Us About LLM Safety Re­search?

Manqing Liu23 Dec 2025 4:14 UTC
8 points
0 comments4 min readLW link

Job List­ing (CLOSED): CBAI Re­search Managers

23 Dec 2025 4:03 UTC
1 point
0 comments1 min readLW link

Ground­ing Value Learn­ing in Evolu­tion­ary Psy­chol­ogy: an Alter­na­tive Pro­posal to CEV

RogerDearnaley23 Dec 2025 3:40 UTC
40 points
25 comments20 min readLW link

The Benefits of Med­i­ta­tion Come From Tel­ling Peo­ple That You Meditate

ThirdEyeJoe (cousin of CottonEyedJoe)23 Dec 2025 1:48 UTC
35 points
5 comments2 min readLW link

The fu­ture of al­ign­ment if LLMs are a bubble

Stuart_Armstrong23 Dec 2025 0:08 UTC
47 points
13 comments5 min readLW link

Un­su­per­vised Agent Discovery

Gunnar_Zarncke22 Dec 2025 22:01 UTC
24 points
0 comments6 min readLW link

An­nounc­ing Gemma Scope 2

22 Dec 2025 21:56 UTC
94 points
1 comment2 min readLW link

[Ad­vanced In­tro to AI Align­ment] 0. Overview and Foundations

Towards_Keeperhood22 Dec 2025 21:20 UTC
15 points
0 comments5 min readLW link

$500 Write like lsusr competition

lsusr22 Dec 2025 20:09 UTC
29 points
43 comments3 min readLW link

Ap­pen­dices: Su­per­vised fine­tun­ing on low-harm re­ward hack­ing gen­er­al­ises to high-harm re­ward hacking

22 Dec 2025 19:33 UTC
17 points
0 comments1 min readLW link

Su­per­vised fine­tun­ing on low-harm re­ward hack­ing gen­er­al­ises to high-harm re­ward hacking

22 Dec 2025 19:32 UTC
14 points
0 comments30 min readLW link

Re­cent LLMs can use filler to­kens or prob­lem re­peats to im­prove (no-CoT) math performance

ryan_greenblatt22 Dec 2025 17:21 UTC
152 points
18 comments7 min readLW link