[Question] Would it be use­ful to col­lect the con­texts, where var­i­ous LLMs think the same?

Martin VlachAug 24, 2023, 10:01 PM
6 points
1 comment1 min readLW link

[Question] Help Needed: Craft­ing a Bet­ter CFAR Fol­low-Up Survey

kotrfaAug 24, 2023, 5:26 PM
9 points
2 comments1 min readLW link

AI #26: Fine Tun­ing Time

ZviAug 24, 2023, 3:30 PM
49 points
6 comments33 min readLW link
(thezvi.wordpress.com)

Is this the be­gin­ning of the end for LLMS [as the royal road to AGI, what­ever that is]?

Bill BenzonAug 24, 2023, 2:50 PM
3 points
15 comments3 min readLW link

AI Safety Bounties

PatrickLAug 24, 2023, 2:29 PM
11 points
0 comments7 min readLW link
(rethinkpriorities.org)

AI Reg­u­la­tion May Be More Im­por­tant Than AI Align­ment For Ex­is­ten­tial Safety

otto.bartenAug 24, 2023, 11:41 AM
65 points
39 comments5 min readLW link

AI Prob­a­bil­ity Trees—Katja Grace

Nathan YoungAug 24, 2023, 9:45 AM
8 points
3 comments7 min readLW link

[Question] What wiki-edit­ing fea­tures would make you use the LessWrong wiki more?

Nathan YoungAug 24, 2023, 9:22 AM
21 points
27 comments1 min readLW link

On Lucidity

LeberAug 24, 2023, 8:45 AM
0 points
5 comments1 min readLW link
(leber.substack.com)

The God of Hu­man­ity, and the God of the Robot Utilitarians

RaemonAug 24, 2023, 8:27 AM
80 points
13 comments2 min readLW link1 review

I mea­sure Google’s Mu­sicLM over 3 months as it ap­pears to go from jaw-drop­ping to em­bar­rass­ingly re­peat­ing itself

AttentionResearcherAug 24, 2023, 4:20 AM
19 points
4 comments4 min readLW link

En­hanc­ing Cor­rigi­bil­ity in AI Sys­tems through Ro­bust Feed­back Loops

JustausernameAug 24, 2023, 3:53 AM
1 point
0 comments6 min readLW link

The lost millennium

Ege ErdilAug 24, 2023, 3:48 AM
54 points
14 comments3 min readLW link

Re­greas­ing a KitchenAid Mixer

jefftkAug 24, 2023, 2:30 AM
15 points
0 comments1 min readLW link
(www.jefftk.com)

Assess­ment of in­tel­li­gence agency func­tion­al­ity is difficult yet important

trevorAug 24, 2023, 1:42 AM
48 points
5 comments9 min readLW link

China’s po­si­tion on au­tonomous weapons

bhauthAug 23, 2023, 10:20 PM
17 points
2 comments1 min readLW link
(academic.oup.com)

Diet Ex­per­i­ment Pr­ereg­is­tra­tion: Long-term wa­ter fast­ing + seed oil re­moval

lcAug 23, 2023, 10:08 PM
56 points
18 comments1 min readLW link

The Low-Hang­ing Fruit Prior and sloped valleys in the loss landscape

Aug 23, 2023, 9:12 PM
82 points
1 comment13 min readLW link

Govern­ing, Fast and Slow

CarsonAug 23, 2023, 8:01 PM
3 points
0 comments3 min readLW link

A prob­lem with the most re­cently pub­lished ver­sion of CEV

ThomasCederborgAug 23, 2023, 6:05 PM
10 points
8 comments8 min readLW link1 review

[Question] Which paths to pow­er­ful AI should be boosted?

Zach Stein-PerlmanAug 23, 2023, 4:00 PM
5 points
1 comment1 min readLW link

A The­ory of Laughter

Steven ByrnesAug 23, 2023, 3:05 PM
102 points
14 comments28 min readLW link

Why Is No One Try­ing To Align Profit In­cen­tives With Align­ment Re­search?

PrometheusAug 23, 2023, 1:16 PM
51 points
11 comments4 min readLW link

Ex­plor­ing the Re­spon­si­ble Path to AI Re­search in the Philippines

MiguelDevAug 23, 2023, 8:44 AM
6 points
0 comments6 min readLW link

[Question] Do agents with (mu­tu­ally known) iden­ti­cal util­ity func­tions but ir­rec­on­cilable knowl­edge some­times fight?

mako yassAug 23, 2023, 8:13 AM
14 points
13 comments1 min readLW link

South Bay ACX/​SSC Fall Mee­tups Everywhere

allisonaAug 23, 2023, 3:00 AM
3 points
0 comments1 min readLW link

Separate the truth from your wishes

Jacob G-WAug 23, 2023, 12:52 AM
6 points
3 comments1 min readLW link
(jacobgw.com)

Im­pli­ca­tions of ev­i­den­tial co­op­er­a­tion in large worlds

Lukas FinnvedenAug 23, 2023, 12:43 AM
39 points
4 comments17 min readLW link
(lukasfinnveden.substack.com)

South Bay Ca­sual Group Walk

allisonaAug 22, 2023, 10:43 PM
7 points
2 comments1 min readLW link

Walk while you talk: don’t balk at “no chalk”

dkl9Aug 22, 2023, 9:27 PM
41 points
10 comments2 min readLW link
(dkl9.net)

State of Gen­er­ally Available Self-Driving

jefftkAug 22, 2023, 6:50 PM
66 points
6 comments2 min readLW link
(www.jefftk.com)

Seth Ex­plains Consciousness

Jacob FalkovichAug 22, 2023, 6:06 PM
39 points
130 comments14 min readLW link1 review
(putanumonit.com)

ChatGPT challenges the case for hu­man irrationality

Kevin DorstAug 22, 2023, 12:46 PM
3 points
10 comments7 min readLW link
(kevindorst.substack.com)

[Question] Does one have rea­son to be­lieve the simu­la­tion hy­poth­e­sis is prob­a­bly true?

kuiraAug 22, 2023, 8:34 AM
1 point
20 comments1 min readLW link

The Joan of Arc Challenge For Ob­jec­tive List The­ory

Bentham's BulldogAug 22, 2023, 8:01 AM
−2 points
4 comments10 min readLW link

The Lop­sided Lives Ar­gu­ment For He­donism About Well-being

Bentham's BulldogAug 22, 2023, 7:59 AM
−2 points
8 comments22 min readLW link

Causal­ity and a Cost Se­man­tics for Neu­ral Networks

scottviteriAug 21, 2023, 9:02 PM
22 points
1 comment1 min readLW link

Ideas for im­prov­ing epistemics in AI safety outreach

micAug 21, 2023, 7:55 PM
64 points
6 comments3 min readLW link

Rice’s The­o­rem says that AIs can’t de­ter­mine much from study­ing AI source code

Michael Weiss-MalikAug 21, 2023, 7:05 PM
−12 points
4 comments1 min readLW link

Large Lan­guage Models will be Great for Censorship

Ethan EdwardsAug 21, 2023, 7:03 PM
185 points
14 comments8 min readLW link
(ethanedwards.substack.com)

“Throw­ing Ex­cep­tions” Is A Strange Pro­gram­ming Pattern

Thoth HermesAug 21, 2023, 6:50 PM
−2 points
13 comments6 min readLW link
(thothhermes.substack.com)

[Question] Which pos­si­ble AI sys­tems are rel­a­tively safe?

Zach Stein-PerlmanAug 21, 2023, 5:00 PM
42 points
20 comments1 min readLW link

Self-shut­down AI

Jan BetleyAug 21, 2023, 4:48 PM
13 points
2 comments2 min readLW link

Con­tex­tual Trans­la­tions—At­tempt 1

Varshul GuptaAug 21, 2023, 2:30 PM
−1 points
0 comments2 min readLW link
(dubverseblack.substack.com)

DIY De­liber­ate Practice

lynettebyeAug 21, 2023, 12:22 PM
63 points
4 comments5 min readLW link
(lynettebye.com)

Down­stairs Open­ing: 2br Apartment

jefftkAug 21, 2023, 12:50 AM
8 points
2 comments3 min readLW link
(www.jefftk.com)

Effi­ciency and re­source use scal­ing parity

Ege ErdilAug 21, 2023, 12:18 AM
51 points
1 comment4 min readLW link1 review

Ruin­ing an ex­pected-log-money maximizer

philhAug 20, 2023, 9:20 PM
33 points
33 comments1 min readLW link1 review
(reasonableapproximation.net)

Steven Wolfram on AI Alignment

Bill BenzonAug 20, 2023, 7:49 PM
66 points
15 comments4 min readLW link

[Question] What value does per­sonal pre­dic­tion track­ing have?

fxAug 20, 2023, 6:43 PM
7 points
3 comments1 min readLW link