Funny Anec­dote of Eliezer From His Sister

Noah BirnbaumApr 22, 2024, 10:05 PM
207 points
6 comments2 min readLW link

How LLMs Work, in the Style of The Economist

utilistrutilApr 22, 2024, 7:06 PM
0 points
0 comments2 min readLW link

Mea­sur­ing Co­her­ence and Goal-Direct­ed­ness in RL Policies

dx26Apr 22, 2024, 6:26 PM
10 points
0 comments7 min readLW link

AI Reg­u­la­tion is Unsafe

Maxwell TabarrokApr 22, 2024, 4:37 PM
40 points
41 comments4 min readLW link
(www.maximum-progress.com)

Pri­ors and Prejudice

MathiasKBApr 22, 2024, 3:00 PM
151 points
31 comments7 min readLW link

For­get Every­thing (Statis­ti­cal Me­chan­ics Part 1)

J BostockApr 22, 2024, 1:33 PM
41 points
7 comments3 min readLW link

On Llama-3 and Dwarkesh Pa­tel’s Pod­cast with Zuckerberg

ZviApr 22, 2024, 1:10 PM
63 points
4 comments47 min readLW link
(thezvi.wordpress.com)

Mo­ti­va­tion gaps: Why so much EA crit­i­cism is hos­tile and lazy

titotalApr 22, 2024, 11:49 AM
70 points
5 commentsLW link
(titotal.substack.com)

Should we break up Google Deep­Mind?

Hauke HillebrandtApr 22, 2024, 9:16 AM
3 points
0 commentsLW link

What should our con­tain­ers do?

Richard HenageApr 22, 2024, 6:17 AM
1 point
1 comment2 min readLW link

Goal ori­ented cog­ni­tion in “a sin­gle for­ward pass”

Apr 22, 2024, 5:03 AM
20 points
15 comments26 min readLW link

Time com­plex­ity for de­ter­minis­tic string machines

alcatalApr 21, 2024, 10:35 PM
21 points
2 comments21 min readLW link

Trans­fer Learn­ing in Humans

niplavApr 21, 2024, 8:49 PM
61 points
1 comment13 min readLW link

I cre­ated an Asi Align­ment Tier List

TimeGoatApr 21, 2024, 6:44 PM
−6 points
0 comments1 min readLW link

The los­ing iden­tity of Twitter

Itay DreyfusApr 21, 2024, 1:43 PM
20 points
1 comment12 min readLW link
(productidentity.co)

Good Bings copy, great Bings steal

dr_sApr 21, 2024, 9:52 AM
31 points
6 comments9 min readLW link

Paper: “The Ethics of Ad­vanced AI As­sis­tants” -Google DeepMind

Tristan WegnerApr 21, 2024, 6:45 AM
20 points
0 comments1 min readLW link
(storage.googleapis.com)

Con­tra Chord Simplification

jefftkApr 21, 2024, 2:30 AM
9 points
0 comments1 min readLW link
(www.jefftk.com)

A cou­ple pro­duc­tivity tips for overthinkers

Steven ByrnesApr 20, 2024, 4:05 PM
79 points
13 comments4 min readLW link

“You’re the most beau­tiful girl in the world” and Wittgen­stei­nian Lan­guage Games

Chris_LeongApr 20, 2024, 2:54 PM
5 points
18 comments1 min readLW link

Past Tense Features

CanApr 20, 2024, 2:34 PM
12 points
0 comments4 min readLW link

Thoughts on seed oil

dynomightApr 20, 2024, 12:29 PM
357 points
129 comments17 min readLW link
(dynomight.net)

How to know whether you are an ideal­ist or a phys­i­cal­ist/​materialist

JackOfAllTradesApr 20, 2024, 11:53 AM
−3 points
2 comments1 min readLW link

How I Think, Part Four: Money is Weird

Richard HenageApr 20, 2024, 6:21 AM
0 points
3 comments5 min readLW link

The power of finite and the weak­ness of in­finite bi­nary point numbers

AxiomWriterApr 20, 2024, 6:03 AM
−3 points
6 comments2 min readLW link

WISDOMISM A Mo­ral The­ory for the Age of Information

Peter lawless Apr 19, 2024, 11:06 PM
2 points
0 comments9 min readLW link

In­duc­ing Un­prompted Misal­ign­ment in LLMs

Apr 19, 2024, 8:00 PM
38 points
7 comments16 min readLW link

Introspection

A*Apr 19, 2024, 7:10 PM
7 points
0 comments1 min readLW link

[Full Post] Progress Up­date #1 from the GDM Mech In­terp Team

Apr 19, 2024, 7:06 PM
79 points
10 comments8 min readLW link

[Sum­mary] Progress Up­date #1 from the GDM Mech In­terp Team

Apr 19, 2024, 7:06 PM
72 points
0 comments3 min readLW link

Daniel Den­nett has died (1942-2024)

kaveApr 19, 2024, 4:17 PM
150 points
5 comments1 min readLW link
(dailynous.com)

Events Book­ing New Callers?

jefftkApr 19, 2024, 3:50 PM
9 points
0 comments1 min readLW link
(www.jefftk.com)

[Question] What is the best way to talk about prob­a­bil­ities you ex­pect to change with ev­i­dence/​ex­per­i­ments?

Will_PearsonApr 19, 2024, 3:35 PM
14 points
11 comments1 min readLW link

CTMU in­sight: maybe con­scious­ness *can* af­fect quan­tum out­comes?

zhukeepaApr 19, 2024, 3:23 PM
13 points
11 comments5 min readLW link

De­mon­strate and eval­u­ate risks from AI to so­ciety at the AI x Democ­racy re­search hackathon

Esben KranApr 19, 2024, 2:46 PM
5 points
0 commentsLW link
(www.apartresearch.com)

[Question] How to Model the Fu­ture of Open-Source LLMs?

Joel BurgetApr 19, 2024, 2:28 PM
25 points
9 comments1 min readLW link

What’s up with all the non-Mor­mons? Weirdly spe­cific uni­ver­sal­ities across LLMs

mwatkinsApr 19, 2024, 1:43 PM
40 points
13 comments27 min readLW link

[Question] If digi­tal goods in vir­tual wor­lds in­crease GDP, do we ac­tu­ally be­come richer?

No77eApr 19, 2024, 10:06 AM
10 points
14 comments1 min readLW link

Ex­per­i­ment on re­peat­ing choices

KatjaGraceApr 19, 2024, 4:20 AM
56 points
1 comment3 min readLW link
(worldspiritsockpuppet.com)

Effec­tive Altru­ists and Ra­tion­al­ists Views & The case for us­ing mar­ket­ing to high­light AI risks.

gilchApr 19, 2024, 4:16 AM
6 points
1 comment1 min readLW link
(youtu.be)

Co­he­sion and busi­ness problems

Adam ZernerApr 19, 2024, 12:45 AM
12 points
8 comments4 min readLW link

The Ther­mo­dy­nam­ics of Death

Peter lawless Apr 19, 2024, 12:36 AM
6 points
0 comments10 min readLW link

Back­yard Office

jefftkApr 19, 2024, 12:31 AM
13 points
0 comments1 min readLW link
(www.jefftk.com)

hy­dro­gen tube transport

bhauthApr 18, 2024, 10:47 PM
34 points
12 comments5 min readLW link
(www.bhauth.com)

LessOn­line Fes­ti­val Up­dates Thread

Ben PaceApr 18, 2024, 9:55 PM
59 points
26 comments1 min readLW link

A Re­view of In-Con­text Learn­ing Hy­pothe­ses for Au­to­mated AI Align­ment Research

alamertonApr 18, 2024, 6:29 PM
25 points
4 comments16 min readLW link

I’m open for pro­jects (sort of)

cousin_itApr 18, 2024, 6:05 PM
46 points
13 comments1 min readLW link

Blessed in­for­ma­tion, garbage in­for­ma­tion, cursed information

tailcalledApr 18, 2024, 4:56 PM
23 points
8 comments3 min readLW link

[Fic­tion] A Confession

Arjun PanicksseryApr 18, 2024, 4:28 PM
38 points
2 comments5 min readLW link
(arjunpanickssery.substack.com)

Discrim­i­nat­ing Be­hav­iorally Iden­ti­cal Clas­sifiers: a model prob­lem for ap­ply­ing in­ter­pretabil­ity to scal­able oversight

Sam MarksApr 18, 2024, 4:17 PM
113 points
10 comments12 min readLW link