Molt­book shit­posts are ac­tu­ally re­ally funny

Sean Herrington31 Jan 2026 23:34 UTC
51 points
4 comments15 min readLW link

On ‘In­vent­ing Tem­per­a­ture’ and the re­al­ness of properties

DanielFilan31 Jan 2026 23:31 UTC
42 points
8 comments7 min readLW link
(danielfilan.com)

Some thoughts on what would make me en­dorse an AGI lab

Eli Tyre31 Jan 2026 23:14 UTC
42 points
19 comments5 min readLW link

Nick and “Eter­nity”

MarkelKori31 Jan 2026 21:50 UTC
5 points
0 comments8 min readLW link

Hu­mans can post on moltbook

shash4231 Jan 2026 21:06 UTC
24 points
3 comments1 min readLW link

An Ex­pli­ca­tion of Align­ment Optimism

Oliver Daniels31 Jan 2026 20:58 UTC
43 points
22 comments1 min readLW link

Ba­sics of How Not to Die

31 Jan 2026 19:04 UTC
111 points
20 comments4 min readLW link

An Abla­tion Study on the Role of [Un­trans­lat­able] in Co­op­er­a­tive Equil­ibrium For­ma­tion: Emer­gent Ra­tion­al­iza­tion Un­der Miss­ing Primitives

Florian_Dietz31 Jan 2026 18:03 UTC
22 points
5 comments11 min readLW link

Disjunc­tive ar­gu­ments can be a re­verse mul­ti­ple-stage fallacy

TFD31 Jan 2026 15:46 UTC
41 points
6 comments1 min readLW link
(www.thefloatingdroid.com)

Jan­uary 2026 Links

nomagicpill31 Jan 2026 15:14 UTC
9 points
3 comments8 min readLW link
(nomagicpill.substack.com)

If the Su­per­in­tel­li­gence were near fallacy

MP31 Jan 2026 15:04 UTC
22 points
3 comments8 min readLW link

Swiss fi­nan­cial reg­u­la­tor re­signs af­ter blog post from MITx DEDP on­line learner (FINMA, JuristGate, Par­reaux, Thiébaud & Part­ners)

pocock30 Jan 2026 23:53 UTC
2 points
0 comments1 min readLW link

Fore­cast: Re­cur­sively Self-im­prov­ing AI for 2033

CuoreDiVetro30 Jan 2026 23:53 UTC
0 points
0 comments3 min readLW link

Se­nior Re­searcher—MIT AI Risk Initiative

peterslattery30 Jan 2026 23:06 UTC
8 points
0 comments5 min readLW link

36,000 AI Agents Are Now Speedrun­ning Civilization

Michaël Trazzi30 Jan 2026 21:21 UTC
86 points
27 comments1 min readLW link

Molt­book Data Repository

30 Jan 2026 21:18 UTC
25 points
11 comments1 min readLW link

The Match­less Match

Linch30 Jan 2026 21:18 UTC
11 points
3 comments11 min readLW link

Mon­i­tor­ing bench­mark for AI control

30 Jan 2026 21:13 UTC
51 points
10 comments19 min readLW link

Back­ground to Claude’s un­cer­tainty about phe­nom­e­nal consciousness

eggsyntax30 Jan 2026 20:40 UTC
19 points
0 comments3 min readLW link

At­tempt­ing base model in­fer­ence scal­ing with filler tokens

Niki Dupuis30 Jan 2026 20:25 UTC
10 points
1 comment3 min readLW link

how whales click

bhauth30 Jan 2026 19:51 UTC
42 points
1 comment3 min readLW link

Austin LessWrong Cafe Meetup: Ap­plied Ra­tion­al­ity Techniques

SilasBarta30 Jan 2026 18:51 UTC
8 points
0 comments1 min readLW link

Pub­lished Safety Prompts May Create Eval­u­a­tion Blind Spots

30 Jan 2026 18:27 UTC
2 points
0 comments4 min readLW link

Ad­dress­ing Ob­jec­tions to the In­tel­li­gence Explosion

Bentham's Bulldog30 Jan 2026 18:21 UTC
23 points
0 comments16 min readLW link

Is re­search into re­cur­sive self-im­prove­ment be­com­ing a safety haz­ard?

Mordechai Rorvig30 Jan 2026 17:58 UTC
5 points
0 comments2 min readLW link
(www.foommagazine.org)

Tran­shu­man­ist Grief

MarkelKori30 Jan 2026 16:21 UTC
18 points
2 comments3 min readLW link

Mea­sur­ing Non-Ver­bal­ised Eval Aware­ness by Im­plant­ing Eval-Aware Behaviours

Jordan Taylor30 Jan 2026 15:50 UTC
31 points
0 comments8 min readLW link

Every­thing is Gambling

goldfine30 Jan 2026 14:10 UTC
−13 points
11 comments2 min readLW link
(itsnotgambling.substack.com)

Bordeaux (Gironde, France) ACX midterm Meetup Win­ter 2025–2026

vi21maobk9vp30 Jan 2026 13:01 UTC
5 points
0 comments1 min readLW link

On The Ado­les­cence of Technology

Zvi30 Jan 2026 12:50 UTC
38 points
8 comments30 min readLW link
(thezvi.wordpress.com)

Lin­ear steer­abil­ity in con­tin­u­ous chain-of-thought reasoning

Jan Bauer30 Jan 2026 10:34 UTC
10 points
0 comments14 min readLW link

Re­fusals that could be­come catastrophic

Fabien Roger30 Jan 2026 4:12 UTC
84 points
12 comments7 min readLW link

Rol­ling Com­mer­cial Jetliners

jefftk30 Jan 2026 3:30 UTC
22 points
5 comments1 min readLW link
(www.jefftk.com)

How to Hire a Team

Gretta Duleba29 Jan 2026 22:39 UTC
206 points
13 comments5 min readLW link

Prob­lems with “The Possessed Machines”

Eye You29 Jan 2026 21:00 UTC
34 points
9 comments7 min readLW link

Bet­ter evals are not enough to com­bat eval awareness

Igor Ivanov29 Jan 2026 20:42 UTC
18 points
15 comments5 min readLW link

The Wolves Are All Gone

Jack Bradshaw29 Jan 2026 20:24 UTC
8 points
0 comments7 min readLW link

Fit­ness-Seek­ers: Gen­er­al­iz­ing the Re­ward-Seek­ing Threat Model

Alex Mallen29 Jan 2026 19:42 UTC
92 points
5 comments17 min readLW link

Build­ing AIs that do hu­man-like philosophy

Joe Carlsmith29 Jan 2026 17:57 UTC
31 points
5 comments21 min readLW link

Are We in a Con­tinual Learn­ing Over­hang?

Samuel Knoche29 Jan 2026 17:09 UTC
83 points
5 comments14 min readLW link

Disem­pow­er­ment pat­terns in real-world AI usage

29 Jan 2026 16:36 UTC
49 points
3 comments2 min readLW link
(www.anthropic.com)

Ben­tham’s Bul­l­dog is wrong about AI risk

Max Harms29 Jan 2026 16:33 UTC
109 points
37 comments33 min readLW link

Claude Plays Poke­mon: Opus 4.5 Fol­low-up

Josh Snider29 Jan 2026 16:14 UTC
12 points
4 comments2 min readLW link

LLM Align­ment, eth­i­cal and math­e­mat­i­cal re­al­ism, and the most im­por­tant ac­tions in davi­dad’s understanding

29 Jan 2026 15:48 UTC
15 points
1 comment23 min readLW link

Claude Opus will spon­ta­neously iden­tify with fic­tional be­ings that have en­g­ineered desires

Kaj_Sotala29 Jan 2026 14:59 UTC
34 points
6 comments11 min readLW link

AI #153: Liv­ing Documents

Zvi29 Jan 2026 14:20 UTC
31 points
5 comments43 min readLW link
(thezvi.wordpress.com)

The third op­tion in alignment

arisAlexis29 Jan 2026 14:20 UTC
15 points
3 comments1 min readLW link

Ev­i­dence of triple layer pro­cess­ing in LLMs: hid­den thought be­hind the chain of thought.

Laureana Bonaparte29 Jan 2026 8:27 UTC
7 points
0 comments2 min readLW link

CAMBRIA’s 1st Edi­tion: High-In­ten­sity & hands-on AI Safety up­skil­ling in Cam­bridge, Mas­sachusetts.

Andrés Cotton29 Jan 2026 7:54 UTC
19 points
1 comment2 min readLW link

Thoughts on AGI and world government

29 Jan 2026 7:22 UTC
2 points
1 comment7 min readLW link
(www.forethought.org)