All200520062007200820092010201120122013201420152016201720182019202020212022202320242025
AllJanFebMarAprMayJunJulAugSepOctNovDec
All12345678910111213141516171819202122232425262728293031

The Mask Comes Off: At What Price?

ZviOct 21, 2024, 11:50 PM
72 points

28 votes

Overall karma indicates overall quality.

16 comments8 min readLW link
(thezvi.wordpress.com)

Dist­in­guish­ing ways AI can be “con­cen­trated”

Matthew BarnettOct 21, 2024, 10:21 PM
34 points

12 votes

Overall karma indicates overall quality.

2 comments4 min readLW link

Jailbreak­ing ChatGPT and Claude us­ing Web API Con­text Injection

Jaehyuk LimOct 21, 2024, 9:34 PM
4 points

7 votes

Overall karma indicates overall quality.

0 comments3 min readLW link

How to Teach Your Brain to Hate Procrastination

10xyzOct 21, 2024, 8:12 PM
3 points

4 votes

Overall karma indicates overall quality.

0 comments2 min readLW link

Paus­ing for what?

MountainPathOct 21, 2024, 8:12 PM
0 points

4 votes

Overall karma indicates overall quality.

1 comment1 min readLW link

What is au­ton­omy? Why bound­aries are nec­es­sary.

Chris LakinOct 21, 2024, 5:56 PM
8 points

6 votes

Overall karma indicates overall quality.

1 comment1 min readLW link
(chrislakin.blog)

Could ran­domly choos­ing peo­ple to serve as rep­re­sen­ta­tives lead to bet­ter gov­ern­ment?

John HuangOct 21, 2024, 5:10 PM
76 points

42 votes

Overall karma indicates overall quality.

13 comments10 min readLW link

There aren’t enough smart peo­ple in biol­ogy do­ing some­thing boring

Abhishaike MahajanOct 21, 2024, 3:52 PM
28 points

15 votes

Overall karma indicates overall quality.

13 comments10 min readLW link

Au­toma­tion collapse

Geoffrey Irving, Tomek Korbak and Benjamin Hilton
Oct 21, 2024, 2:50 PM
72 points

25 votes

Overall karma indicates overall quality.

9 comments7 min readLW link

What AI com­pa­nies should do: Some rough ideas

Zach Stein-PerlmanOct 21, 2024, 2:00 PM
33 points

11 votes

Overall karma indicates overall quality.

10 comments5 min readLW link

[Question] What should OpenAI do that it hasn’t already done, to stop their va­can­cies from be­ing ad­ver­tised on the 80k Job Board?

WitheringWeightsOct 21, 2024, 1:57 PM
22 points

8 votes

Overall karma indicates overall quality.

0 comments1 min readLW link

A Rocket–In­ter­pretabil­ity Analogy

plexOct 21, 2024, 1:55 PM
155 points

63 votes

Overall karma indicates overall quality.

31 comments1 min readLW link

Tokyo AI Safety 2025: Call For Papers

BlaineOct 21, 2024, 8:43 AM
24 points

5 votes

Overall karma indicates overall quality.

0 comments3 min readLW link
(www.tais2025.cc)

OpenAI defected, but we can take hon­est actions

RemmeltOct 21, 2024, 8:41 AM
17 points

16 votes

Overall karma indicates overall quality.

16 comments2 min readLW link

Slightly More Than You Wanted To Know: Preg­nancy Length Effects

JustisMillsOct 21, 2024, 1:26 AM
63 points

22 votes

Overall karma indicates overall quality.

4 comments5 min readLW link
(justismills.substack.com)

In­for­ma­tion vs Assurance

johnswentworthOct 20, 2024, 11:16 PM
187 points

102 votes

Overall karma indicates overall quality.

18 comments2 min readLW link

Liquid vs Illiquid Ca­reers

vaishnav92Oct 20, 2024, 11:03 PM
35 points

24 votes

Overall karma indicates overall quality.

7 comments7 min readLW link
(vaishnavsunil.substack.com)

AI Can be “Gra­di­ent Aware” Without Do­ing Gra­di­ent hack­ing.

SodiumOct 20, 2024, 9:02 PM
21 points

10 votes

Overall karma indicates overall quality.

1 comment2 min readLW link

A brief the­ory of why we think things are good or bad

David JohnstonOct 20, 2024, 8:31 PM
7 points

5 votes

Overall karma indicates overall quality.

10 comments4 min readLW link

Think­ing in 2D

sarahconstantinOct 20, 2024, 7:30 PM
27 points

7 votes

Overall karma indicates overall quality.

0 comments8 min readLW link
(sarahconstantin.substack.com)

Pod­cast dis­cussing Han­son’s Cul­tural Drift Argument

vaishnav92 and regan.arntz.gray
Oct 20, 2024, 5:58 PM
3 points

3 votes

Overall karma indicates overall quality.

0 comments1 min readLW link
(moralmayhem.substack.com)

Ad­vice on Com­mu­ni­cat­ing Concisely

EvolutionByDesignOct 20, 2024, 4:45 PM
3 points

3 votes

Overall karma indicates overall quality.

9 comments1 min readLW link

Am­bi­gui­ties or the is­sues we face with AI in medicine

Thehumanproject.aiOct 20, 2024, 4:45 PM
2 points

2 votes

Overall karma indicates overall quality.

0 comments5 min readLW link

The Per­sonal Im­pli­ca­tions of AGI Realism

xiznebOct 20, 2024, 4:43 PM
7 points

9 votes

Overall karma indicates overall quality.

8 comments5 min readLW link

Safety tax functions

owencbOct 20, 2024, 2:08 PM
31 points

8 votes

Overall karma indicates overall quality.

0 comments6 min readLW link
(strangecities.substack.com)

Ex­plor­ing the Pla­tonic Rep­re­sen­ta­tion Hy­poth­e­sis Beyond In-Distri­bu­tion Data

rokosbasiliskOct 20, 2024, 8:40 AM
12 points

5 votes

Overall karma indicates overall quality.

2 comments1 min readLW link

Elec­toral Systems

RedFishBlueFishOct 20, 2024, 3:25 AM
1 point

3 votes

Overall karma indicates overall quality.

0 comments14 min readLW link

Over­com­ing Bias Anthology

Arjun PanicksseryOct 20, 2024, 2:01 AM
169 points

61 votes

Overall karma indicates overall quality.

14 comments2 min readLW link
(overcoming-bias-anthology.com)

D/​acc AI Se­cu­rity Salon

Allison DuettmannOct 19, 2024, 10:17 PM
19 points

6 votes

Overall karma indicates overall quality.

0 comments1 min readLW link

Who Should Have Been Killed, and Con­tains Neato? Who Else Could It Be, but that Villain Mag­neto!

Ace DelgadoOct 19, 2024, 8:39 PM
−16 points

8 votes

Overall karma indicates overall quality.

0 comments1 min readLW link

If far-UV is so great, why isn’t it ev­ery­where?

Austin ChenOct 19, 2024, 6:56 PM
71 points

28 votes

Overall karma indicates overall quality.

23 comments9 min readLW link
(strainhardening.substack.com)

What if AGI was already ac­ci­den­tally cre­ated in 2019? [Fic­tional story]

Alice WanderlandOct 19, 2024, 9:17 AM
−3 points

4 votes

Overall karma indicates overall quality.

2 comments15 min readLW link
(aliceandbobinwanderland.substack.com)

[Question] What ac­tual bad out­come has “ethics-based” RLHF AI Align­ment already pre­vented?

RokoOct 19, 2024, 6:11 AM
7 points

4 votes

Overall karma indicates overall quality.

16 comments1 min readLW link

[Question] What’s a good book for a tech­ni­cally-minded 11-year old?

Martin SustrikOct 19, 2024, 6:05 AM
10 points

3 votes

Overall karma indicates overall quality.

32 comments1 min readLW link

Method­ol­ogy: Con­ta­gious Beliefs

James Stephen BrownOct 19, 2024, 3:58 AM
3 points

3 votes

Overall karma indicates overall quality.

0 comments7 min readLW link

AI Prej­u­dices: Prac­ti­cal Implications

PeterMcCluskeyOct 19, 2024, 2:19 AM
12 points

4 votes

Overall karma indicates overall quality.

0 comments5 min readLW link
(bayesianinvestor.com)

Start an Up­per-Room UV In­stal­la­tion Com­pany?

jefftkOct 19, 2024, 2:00 AM
44 points

16 votes

Overall karma indicates overall quality.

9 comments1 min readLW link
(www.jefftk.com)

How I’d like al­ign­ment to get done (as of 2024-10-18)

TristanTrimOct 18, 2024, 11:39 PM
11 points

6 votes

Overall karma indicates overall quality.

4 comments4 min readLW link

Sab­o­tage Eval­u­a­tions for Fron­tier Models

David Duvenaud, Joe Benton, Sam Bowman, evhub, mishajw, Eric Christiansen, HoldenKarnofsky, Ethan Perez and Buck
Oct 18, 2024, 10:33 PM
95 points

36 votes

Overall karma indicates overall quality.

56 comments6 min readLW link
(assets.anthropic.com)

D&D Sci Coli­seum: Arena of Data

aphyerOct 18, 2024, 10:02 PM
42 points

16 votes

Overall karma indicates overall quality.

23 comments4 min readLW link

the Day­di­ca­tion technique

chaosmageOct 18, 2024, 9:47 PM
31 points

14 votes

Overall karma indicates overall quality.

0 comments2 min readLW link

[Linkpost] Hawk­ish na­tion­al­ism vs in­ter­na­tional AI power and benefit sharing

jakub_krys and Naci Cankaya
Oct 18, 2024, 6:13 PM
7 points

6 votes

Overall karma indicates overall quality.

5 comments1 min readLW link
(nacicankaya.substack.com)

LLM Psy­cho­met­rics and Prompt-In­duced Psychopathy

Korbinian K.Oct 18, 2024, 6:11 PM
12 points

7 votes

Overall karma indicates overall quality.

2 comments10 min readLW link

A short pro­ject on Mamba: grokking & interpretability

Alejandro TlaieOct 18, 2024, 4:59 PM
21 points

8 votes

Overall karma indicates overall quality.

0 comments6 min readLW link

LLMs can learn about them­selves by introspection

Felix J Binder and Owain_Evans
Oct 18, 2024, 4:12 PM
109 points

43 votes

Overall karma indicates overall quality.

38 comments9 min readLW link

[Question] Are there more than 12 paths to Su­per­in­tel­li­gence?

p4rziv4lOct 18, 2024, 4:05 PM
−3 points

3 votes

Overall karma indicates overall quality.

0 comments1 min readLW link

Low Prob­a­bil­ity Es­ti­ma­tion in Lan­guage Models

Gabriel WuOct 18, 2024, 3:50 PM
50 points

18 votes

Overall karma indicates overall quality.

0 comments10 min readLW link
(www.alignment.org)

The Mys­te­ri­ous Trump Buy­ers on Polymarket

AnnapurnaOct 18, 2024, 1:26 PM
52 points

31 votes

Overall karma indicates overall quality.

10 comments2 min readLW link
(jorgevelez.substack.com)

On In­ten­tion­al­ity, or: Towards a More In­clu­sive Con­cept of Lying

Cornelius DybdahlOct 18, 2024, 10:37 AM
8 points

4 votes

Overall karma indicates overall quality.

0 comments4 min readLW link

Species as Canon­i­cal Refer­ents of Su­per-Organisms

Yudhister KumarOct 18, 2024, 7:49 AM
16 points

7 votes

Overall karma indicates overall quality.

8 comments2 min readLW link
(www.yudhister.me)