The pur­pose of the (Mo­saic) law

mruwnikSep 4, 2023, 11:38 PM
7 points
5 comments6 min readLW link

Against the Open Source /​ Closed Source Di­chotomy: Reg­u­lated Source as a Model for Re­spon­si­ble AI Development

alex.herwixSep 4, 2023, 8:25 PM
4 points
12 comments6 min readLW link
(forum.effectivealtruism.org)

Notes on nukes, IR, and AI from “Arse­nals of Folly” (and other books)

tlevinSep 4, 2023, 7:02 PM
11 points
0 comments6 min readLW link

Hert­ford, Sour­but (ra­tio­nal­ity les­sons from Univer­sity Challenge)

Oliver SourbutSep 4, 2023, 6:44 PM
28 points
7 comments14 min readLW link
(www.oliversourbut.net)

a rant on poli­ti­cian-en­g­ineer coal­i­tional conflict

bhauthSep 4, 2023, 5:15 PM
64 points
12 comments4 min readLW link

How Fo­rumMag­num builds com­mu­ni­ties of inquiry

Jim FisherSep 4, 2023, 4:52 PM
33 points
21 comments5 min readLW link

In­ter­pret­ing a ma­trix-val­ued word em­bed­ding with a math­e­mat­i­cally proven char­ac­ter­i­za­tion of all optima

Joseph Van NameSep 4, 2023, 4:19 PM
3 points
4 comments12 min readLW link

Hard Ques­tions Are Lan­guage Bugs

George3d6Sep 4, 2023, 2:44 PM
30 points
13 comments7 min readLW link
(ontologi.cc)

De­fund­ing My Mistake

ymeskhoutSep 4, 2023, 2:43 PM
178 points
41 comments6 min readLW link

The om­ni­zoid—Heighn FDT De­bate #1: Why FDT Isn’t Crazy

HeighnSep 4, 2023, 12:57 PM
24 points
4 comments6 min readLW link

Paper: On mea­sur­ing situ­a­tional aware­ness in LLMs

Sep 4, 2023, 12:54 PM
109 points
16 comments5 min readLW link
(arxiv.org)

Im­pend­ing AGI doesn’t make ev­ery­thing else unimportant

Igor IvanovSep 4, 2023, 12:34 PM
29 points
12 comments5 min readLW link

Open Thread – Au­tumn 2023

RaemonSep 3, 2023, 10:54 PM
26 points
111 comments1 min readLW link

What must be the case that ChatGPT would have mem­o­rized “To be or not to be”? – Three kinds of con­cep­tual ob­jects for LLMs

Bill BenzonSep 3, 2023, 6:39 PM
19 points
0 comments12 min readLW link

Fun­da­men­tal ques­tion: What de­ter­mines a mind’s effects?

TsviBTSep 3, 2023, 5:15 PM
15 points
4 comments13 min readLW link

An em­bed­ding de­coder model, trained with a differ­ent ob­jec­tive on a differ­ent dataset, can de­code an­other model’s em­bed­dings sur­pris­ingly accurately

Logan ZoellnerSep 3, 2023, 11:34 AM
20 points
1 comment1 min readLW link

Series of ab­surd up­grades in na­ture’s great search

lemonhopeSep 3, 2023, 9:35 AM
15 points
8 comments1 min readLW link

Con­ser­va­tion of Ex­pected Ev­i­dence and Ran­dom Sam­pling in Anthropics

Ape in the coatSep 3, 2023, 6:55 AM
9 points
9 comments7 min readLW link

The goal of physics

Jim PivarskiSep 2, 2023, 11:08 PM
46 points
4 comments5 min readLW link

Will value of paid sex drop right be­fore the end of the world?

azamatvalievSep 2, 2023, 7:03 PM
−13 points
0 comments4 min readLW link

PIBBSS Sum­mer Sym­po­sium 2023

Sep 2, 2023, 5:22 PM
25 points
2 comments3 min readLW link

The small­est pos­si­ble but­ton (or: moth traps!)

Neil Sep 2, 2023, 3:24 PM
122 points
18 comments3 min readLW link
(neilwarren.substack.com)

Steven Har­nad: Sym­bol ground­ing and the struc­ture of dictionaries

Bill BenzonSep 2, 2023, 12:28 PM
5 points
3 comments2 min readLW link

Is Me­taethics Un­nec­es­sary Given In­tent-Aligned AI?

Caleb BiddulphSep 2, 2023, 9:48 AM
10 points
0 comments7 min readLW link

Ra­tional Agents Co­op­er­ate in the Pri­soner’s Dilemma

Isaac KingSep 2, 2023, 6:15 AM
17 points
68 comments12 min readLW link

[Linkpost] Large lan­guage mod­els con­verge to­ward hu­man-like con­cept organization

Bogdan Ionut CirsteaSep 2, 2023, 6:00 AM
22 points
1 comment1 min readLW link

Plum Cook­ing Temperature

jefftkSep 2, 2023, 1:30 AM
11 points
0 comments1 min readLW link
(www.jefftk.com)

[Question] What did you learn from leaked doc­u­ments?

wassnameSep 2, 2023, 1:28 AM
15 points
10 comments1 min readLW link

One Minute Every Moment

abramdemskiSep 1, 2023, 8:23 PM
125 points
23 comments3 min readLW link

Ten­sor Trust: An on­line game to un­cover prompt in­jec­tion vulnerabilities

Sep 1, 2023, 7:31 PM
30 points
0 comments5 min readLW link
(tensortrust.ai)

Re­pro­duc­ing ARC Evals’ re­cent re­port on lan­guage model agents

Thomas BroadleySep 1, 2023, 4:52 PM
104 points
17 comments3 min readLW link
(thomasbroadley.com)

[Question] Why aren’t more peo­ple in AIS fa­mil­iar with PDP?

PrometheusSep 1, 2023, 3:27 PM
4 points
9 comments1 min readLW link

AGI isn’t just a technology

Seth HerdSep 1, 2023, 2:35 PM
18 points
12 comments2 min readLW link

Can an LLM iden­tify ring-com­po­si­tion in a liter­ary text? [ChatGPT]

Bill BenzonSep 1, 2023, 2:18 PM
4 points
2 comments11 min readLW link

What is OpenAI’s plan for mak­ing AI Safer?

brookSep 1, 2023, 11:15 AM
6 points
0 comments4 min readLW link
(aisafetyexplained.substack.com)

Progress links di­gest, 2023-09-01: How an­cient peo­ple ma­nipu­lated wa­ter, and more

jasoncrawfordSep 1, 2023, 4:33 AM
13 points
4 comments6 min readLW link
(rootsofprogress.org)

A Golden Age of Build­ing? Ex­cerpts and les­sons from Em­pire State, Pen­tagon, Skunk Works and SpaceX

Bird ConceptSep 1, 2023, 4:03 AM
188 points
26 comments24 min readLW link1 review

[Question] Would AI ex­perts ever agree that AGI sys­tems have at­tained “con­scious­ness”?

Super AGISep 1, 2023, 3:57 AM
−16 points
6 comments1 min readLW link

Meta Ques­tions about Metaphilosophy

Wei DaiSep 1, 2023, 1:17 AM
161 points
80 comments3 min readLW link

[Linkpost] Michael Niel­sen re­marks on ‘Op­pen­heimer’

22tomAug 31, 2023, 3:46 PM
78 points
7 comments2 min readLW link
(michaelnotebook.com)

My thoughts on AI and per­sonal fu­ture plan af­ter learn­ing about AI Safety for 4 months

Ziyue WangAug 31, 2023, 3:32 PM
7 points
0 comments4 min readLW link

Which Ques­tions Are An­thropic Ques­tions?

dadadarrenAug 31, 2023, 3:15 PM
16 points
13 comments3 min readLW link

The Tree of Life, and a Note on Job

Bill BenzonAug 31, 2023, 2:03 PM
13 points
7 comments4 min readLW link

Clean­ing a SoundCraft Mixer

jefftkAug 31, 2023, 1:20 PM
11 points
0 comments1 min readLW link
(www.jefftk.com)

AI #27: Por­tents of Gemini

ZviAug 31, 2023, 12:40 PM
54 points
37 comments47 min readLW link
(thezvi.wordpress.com)

[CANCELLED DUE TO ILLNESS] San Fran­cisco ACX Meetup “First Satur­day”

guenaelAug 31, 2023, 12:34 PM
1 point
0 comments1 min readLW link

Long-Term Fu­ture Fund Ask Us Any­thing (Septem­ber 2023)

Aug 31, 2023, 12:28 AM
33 points
6 comments1 min readLW link
(forum.effectivealtruism.org)

Re­sponses to ap­par­ent ra­tio­nal­ist con­fu­sions about game /​ de­ci­sion theory

Anthony DiGiovanniAug 30, 2023, 10:02 PM
142 points
20 comments12 min readLW link1 review

In­vuln­er­a­ble In­com­plete Prefer­ences: A For­mal Statement

SCPAug 30, 2023, 9:59 PM
134 points
39 comments35 min readLW link

Re­port on Fron­tier Model Training

YafahEdelmanAug 30, 2023, 8:02 PM
122 points
21 comments21 min readLW link
(docs.google.com)