All200520062007200820092010201120122013201420152016201720182019202020212022202320242025
AllJanFebMarAprMayJunJulAugSepOctNovDec
All12345678910111213141516171819202122232425262728293031

The Army of Jakoths (a parable)

MikkWMay 21, 2023, 10:48 PM
−6 points
0 comments1 min readLW link

A&I (Rihanna ‘S&M’ par­ody lyrics)

nahojMay 21, 2023, 10:34 PM
−2 points
0 comments2 min readLW link

Four Bat­tle­grounds: Power in the Age of Ar­tifi­cial In­tel­li­gence (Book re­view)

PeterMcCluskeyMay 21, 2023, 9:19 PM
25 points
0 comments4 min readLW link
(bayesianinvestor.com)

Gen­der Vec­tors in ROME’s La­tent Space

XodarapMay 21, 2023, 6:46 PM
14 points
2 comments3 min readLW link

Weight by Impact

VaniverMay 21, 2023, 2:37 PM
29 points
1 comment3 min readLW link

[out­dated] My cur­rent the­ory of change to miti­gate ex­is­ten­tial risk by mis­al­igned ASI

mesaoptimizerMay 21, 2023, 1:46 PM
32 points
8 comments6 min readLW link
(mesaoptimizer.com)

Bab­ble on grow­ing trust

qbolecMay 21, 2023, 1:19 PM
13 points
1 comment5 min readLW link

Ele­va­tor Positioning

jefftkMay 21, 2023, 11:30 AM
15 points
1 comment1 min readLW link
(www.jefftk.com)

Trans­former Ar­chi­tec­ture Choice for Re­sist­ing Prompt In­jec­tion and Jail-Break­ing Attacks

RogerDearnaleyMay 21, 2023, 8:29 AM
9 points
1 comment4 min readLW link

Jeff Clune ad­ver­tis­ing a post­doc on twit­ter...and ask­ing where he should tar­get his posts

Joyee ChenMay 21, 2023, 1:02 AM
4 points
0 comments1 min readLW link

Run­ning Sound for Yourself

jefftkMay 20, 2023, 10:10 PM
11 points
0 comments2 min readLW link
(www.jefftk.com)

Job Open­ing: SWE to help build sig­na­ture vet­ting sys­tem for AI-re­lated petitions

Ethan Ashkie and Andrew_Critch
May 20, 2023, 7:02 PM
52 points
0 comments1 min readLW link

My Kind of Pragmatism

Nora BelroseMay 20, 2023, 6:58 PM
37 points
11 comments3 min readLW link

Colors Ap­pear To Have Al­most-Univer­sal Sym­bolic Associations

Thoth HermesMay 20, 2023, 6:40 PM
−33 points
4 comments7 min readLW link
(thothhermes.substack.com)

Twiblings, four-par­ent ba­bies and other re­pro­duc­tive technology

GeneSmithMay 20, 2023, 5:11 PM
191 points
33 comments6 min readLW link

P-zom­bies, Com­pres­sion and the Si­mu­la­tion Hypothesis

RussellThorMay 20, 2023, 11:36 AM
5 points
0 comments5 min readLW link

The pos­si­ble shared Craft of de­liber­ate Lex­i­co­ge­n­e­sis

TsviBTMay 20, 2023, 5:56 AM
56 points
5 comments5 min readLW link

Buy­ing Tall-Poppy-Cut­ting Offsets

trevorMay 20, 2023, 3:59 AM
23 points
4 comments2 min readLW link
(www.overcomingbias.com)

See­ing Ghosts by GPT-4

Christopher KingMay 20, 2023, 12:11 AM
−13 points
0 comments1 min readLW link

[Question] What’s the best way to stream­line two-party sale ne­go­ti­a­tions be­tween real hu­mans?

Isaac KingMay 19, 2023, 11:30 PM
15 points
21 comments1 min readLW link

Trust de­vel­ops grad­u­ally via mak­ing bids and set­ting boundaries

Richard_NgoMay 19, 2023, 10:16 PM
134 points
12 comments4 min readLW link

Con­fu­sions and up­dates on STEM AI

Eleni AngelouMay 19, 2023, 9:34 PM
23 points
0 comments3 min readLW link

GPT as an “In­tel­li­gence Fork­lift.”

boazbarakMay 19, 2023, 9:15 PM
49 points
27 comments3 min readLW link

Idea: med­i­cal hy­pothe­ses app for mys­te­ri­ous chronic illnesses

riceissaMay 19, 2023, 8:49 PM
64 points
8 comments3 min readLW link

A flaw in the A.G.I. Ruin Argument

Cole WyethMay 19, 2023, 7:40 PM
1 point
7 comments3 min readLW link
(colewyeth.com)

We are mis­al­igned: the sad­den­ing idea that most of hu­man­ity doesn’t in­trin­si­cally care about x-risk, even on a per­sonal level

Christopher KingMay 19, 2023, 4:12 PM
3 points
5 comments2 min readLW link

Do Dead­lines Make Us Less Creative?

lynettebyeMay 19, 2023, 3:41 PM
44 points
6 comments4 min readLW link

Two Axes of Con­tra Bands

jefftkMay 19, 2023, 2:20 PM
2 points
0 comments1 min readLW link
(www.jefftk.com)

Is Effec­tive Vol­un­teer­ing Pos­si­ble?

David BravoMay 19, 2023, 12:41 PM
13 points
2 comments9 min readLW link

Mr. Meeseeks as an AI ca­pa­bil­ity tripwire

Eric ZhangMay 19, 2023, 11:33 AM
37 points
17 comments2 min readLW link

The Com­pleat Cybornaut

ukc10014, Jozdien and NicholasKees
May 19, 2023, 8:44 AM
66 points
2 comments16 min readLW link

[Question] What if we’re not the first AI-ca­pa­ble civ­i­liza­tion on Earth?

RomanSMay 19, 2023, 7:50 AM
−14 points
8 comments1 min readLW link

Re­solv­ing in­ter­nal con­flicts re­quires listen­ing to what parts want

Richard_NgoMay 19, 2023, 12:04 AM
71 points
0 comments4 min readLW link

[Question] How could I mea­sure the nootropic benefits testos­terone in­jec­tions may have?

shapeshifterMay 18, 2023, 9:40 PM
10 points
3 comments1 min readLW link

In­ves­ti­gat­ing Fabrication

LoganStrohlMay 18, 2023, 5:46 PM
112 points
14 comments16 min readLW link

Microsoft and Google us­ing LLMs for Cybersecurity

PhosphorousMay 18, 2023, 5:42 PM
6 points
0 comments5 min readLW link

The Benev­olent Billion­aire (a pla­gia­rized prob­lem)

Ivan OrdonezMay 18, 2023, 5:39 PM
8 points
11 comments4 min readLW link

Notes from the LSE Talk by Raghu­ram Ra­jan on Cen­tral Bank Balance Sheet Expansions

PixelatedPenguinMay 18, 2023, 5:34 PM
1 point
0 comments2 min readLW link

We Shouldn’t Ex­pect AI to Ever be Fully Rational

OneManyNoneMay 18, 2023, 5:09 PM
19 points
31 comments6 min readLW link

Rel­a­tive Value Func­tions: A Flex­ible New For­mat for Value Estimation

ozziegooenMay 18, 2023, 4:39 PM
20 points
0 commentsLW link

Some back­ground for rea­son­ing about dual-use al­ign­ment research

Charlie SteinerMay 18, 2023, 2:50 PM
126 points
22 comments9 min readLW link1 review

The Un­ex­pected Clanging

Chris_LeongMay 18, 2023, 2:47 PM
14 points
22 comments2 min readLW link

AI #12:The Quest for Sane Regulations

ZviMay 18, 2023, 1:20 PM
77 points
12 comments64 min readLW link
(thezvi.wordpress.com)

[Cross­post] A re­cent write-up of the case for AI (ex­is­ten­tial) risk

TimseyMay 18, 2023, 1:13 PM
6 points
0 comments19 min readLW link

Deon­tolog­i­cal Norms are Unimportant

Bentham's BulldogMay 18, 2023, 9:33 AM
−15 points
8 comments10 min readLW link

Col­lec­tive Identity

NicholasKees, ukc10014 and Garrett Baker
May 18, 2023, 9:00 AM
59 points
12 comments8 min readLW link

Ac­ti­va­tion ad­di­tions in a sim­ple MNIST network

Garrett BakerMay 18, 2023, 2:49 AM
26 points
0 comments2 min readLW link

[Question] What are the limits of the weak man?

ymeskhoutMay 18, 2023, 12:50 AM
9 points
2 comments4 min readLW link

What Yann LeCun gets wrong about al­ign­ing AI (video)

blake8086May 18, 2023, 12:02 AM
0 points
0 comments1 min readLW link
(www.youtube.com)

Let’s use AI to harden hu­man defenses against AI manipulation

Tom DavidsonMay 17, 2023, 11:33 PM
35 points
7 comments24 min readLW link