Limit­ing fac­tors to pre­dict AI take-off speed

Alfonso Pérez EscuderoMay 31, 2023, 11:19 PM
1 point
0 comments6 min readLW link

Un­pre­dictabil­ity and the In­creas­ing Difficulty of AI Align­ment for In­creas­ingly In­tel­li­gent AI

Max_He-HoMay 31, 2023, 10:25 PM
5 points
2 comments20 min readLW link

Shut­down-Seek­ing AI

Simon GoldsteinMay 31, 2023, 10:19 PM
50 points
32 comments15 min readLW link

Full Au­toma­tion is Un­likely and Un­nec­es­sary for Ex­plo­sive Growth

aogMay 31, 2023, 9:55 PM
28 points
3 comments5 min readLW link

LessWrong Com­mu­nity Week­end 2023 Up­dates: Keynote Speaker Mal­colm Ocean, Re­main­ing Tick­ets and More

Henry ProwbellMay 31, 2023, 9:53 PM
23 points
0 comments2 min readLW link

The Div­ine Move Para­dox & Think­ing as a Species

Christopher James HartMay 31, 2023, 9:38 PM
9 points
8 comments3 min readLW link

In­tent-al­igned AI sys­tems de­plete hu­man agency: the need for agency foun­da­tions re­search in AI safety

catubcMay 31, 2023, 9:18 PM
26 points
4 comments11 min readLW link

[Question] How much over­lap is there be­tween the util­ity func­tion of GPT-n and GPT-(n+1), as­sum­ing both are near AGI?

PhosphorousMay 31, 2023, 8:28 PM
2 points
0 comments2 min readLW link

My AI-risk cartoon

preMay 31, 2023, 7:46 PM
6 points
0 comments1 min readLW link

Eval­u­a­tion Ev­i­dence Re­con­struc­tions of Mock Crimes Sub­mis­sion 3

Alan E DunneMay 31, 2023, 7:03 PM
−1 points
0 comments3 min readLW link

Im­prov­ing Math­e­mat­i­cal Rea­son­ing with-Pro­cess Supervision

p.b.May 31, 2023, 7:00 PM
14 points
3 comments1 min readLW link
(openai.com)

The Crux List

ZviMay 31, 2023, 6:30 PM
72 points
19 comments33 min readLW link
(thezvi.wordpress.com)

Stages of Survival

ZviMay 31, 2023, 6:30 PM
44 points
0 comments17 min readLW link
(thezvi.wordpress.com)

Types and De­grees of Alignment

ZviMay 31, 2023, 6:30 PM
36 points
10 comments8 min readLW link
(thezvi.wordpress.com)

To Pre­dict What Hap­pens, Ask What Happens

ZviMay 31, 2023, 6:30 PM
81 points
0 comments9 min readLW link
(thezvi.wordpress.com)

A push to­wards in­ter­ac­tive trans­former decoding

R0bkMay 31, 2023, 5:56 PM
3 points
0 comments2 min readLW link
(github.com)

Neu­roevolu­tion, So­cial In­tel­li­gence, and Logic

vinnik.dmitry07May 31, 2023, 5:54 PM
1 point
0 comments10 min readLW link

Con­trast Pairs Drive the Em­piri­cal Perfor­mance of Con­trast Con­sis­tent Search (CCS)

Scott EmmonsMay 31, 2023, 5:09 PM
97 points
1 comment6 min readLW link1 review

Cos­mopoli­tan val­ues don’t come free

So8resMay 31, 2023, 3:58 PM
137 points
85 comments1 min readLW link

[Question] Ar­gu­ments Against Fos­sil Fu­ture?

SableMay 31, 2023, 1:41 PM
13 points
29 comments1 min readLW link

On Ob­jec­tive Ethics, and a bit about boats

EndlessBlueMay 31, 2023, 11:40 AM
−7 points
3 comments2 min readLW link

Against Con­flat­ing Ex­per­tise: Dist­in­guish­ing AI Devel­op­ment from AI Im­pli­ca­tion Analysis

RatiosMay 31, 2023, 9:50 AM
13 points
4 comments1 min readLW link

A rough model for P(AI doom)

Michael TontchevMay 31, 2023, 8:58 AM
0 points
1 comment2 min readLW link

[Question] What’s the con­sen­sus on porn?

FinalFormal2May 31, 2023, 3:15 AM
5 points
19 comments1 min readLW link

Product En­dorse­ment: Food for sleep interruptions

ElizabethMay 31, 2023, 1:50 AM
45 points
7 comments1 min readLW link
(acesounderglass.com)

Op­ti­mal Clothing

Gordon Seidoh WorleyMay 31, 2023, 1:00 AM
31 points
8 comments6 min readLW link

Ab­strac­tion is Big­ger than Nat­u­ral Abstraction

Nicholas / Heather KrossMay 31, 2023, 12:00 AM
18 points
0 comments5 min readLW link
(www.thinkingmuchbetter.com)

Hu­mans, chim­panzees and other animals

gjmMay 30, 2023, 11:53 PM
21 points
18 comments1 min readLW link

The case for re­mov­ing al­ign­ment and ML re­search from the train­ing dataset

berenMay 30, 2023, 8:54 PM
48 points
8 comments5 min readLW link

Why Job Dis­place­ment Pre­dic­tions are Wrong: Ex­pla­na­tions of Cog­ni­tive Automation

Moritz WallawitschMay 30, 2023, 8:43 PM
−4 points
0 comments8 min readLW link

PaLM-2 & GPT-4 in “Ex­trap­o­lat­ing GPT-N perfor­mance”

Lukas FinnvedenMay 30, 2023, 6:33 PM
57 points
6 comments6 min readLW link

Why I don’t think that the prob­a­bil­ity that AGI kills ev­ery­one is roughly 1 (but rather around 0.995).

BastumannenMay 30, 2023, 5:54 PM
−6 points
0 comments2 min readLW link

AI X-risk is a pos­si­ble solu­tion to the Fermi Paradox

magic9mushroomMay 30, 2023, 5:42 PM
5 points
22 comments2 min readLW link2 reviews

LIMA: Less Is More for Alignment

Ulisse MiniMay 30, 2023, 5:10 PM
16 points
6 comments1 min readLW link
(arxiv.org)

Boomerang—pro­to­col to dis­solve some com­mit­ment races

Filip SondejMay 30, 2023, 4:21 PM
37 points
10 comments8 min readLW link

An­nounc­ing Apollo Research

May 30, 2023, 4:17 PM
217 points
11 comments8 min readLW link

Ad­vice for new al­ign­ment peo­ple: Info Max

Jonas HallgrenMay 30, 2023, 3:42 PM
23 points
4 comments5 min readLW link

[Question] Who is li­able for AI?

jmhMay 30, 2023, 1:54 PM
14 points
4 comments1 min readLW link

AI Safety Newslet­ter #8: Rogue AIs, how to screen for AI risks, and grants for re­search on demo­cratic gov­er­nance of AI

May 30, 2023, 11:52 AM
20 points
0 comments6 min readLW link
(newsletter.safe.ai)

The bul­ls­eye frame­work: My case against AI doom

titotalMay 30, 2023, 11:52 AM
89 points
35 commentsLW link

State­ment on AI Ex­tinc­tion—Signed by AGI Labs, Top Aca­demics, and Many Other Notable Figures

Dan HMay 30, 2023, 9:05 AM
382 points
78 comments1 min readLW link1 review
(www.safe.ai)

The­o­ret­i­cal Limi­ta­tions of Au­tore­gres­sive Models

Gabriel WuMay 30, 2023, 2:37 AM
20 points
1 comment10 min readLW link
(gabrieldwu.github.io)

A book re­view for “An­i­mal Weapons” and cross-ap­ply­ing the les­sons to x-risk

Habeeb AbdulfatahMay 30, 2023, 12:58 AM
−6 points
1 comment1 min readLW link
(www.super-linear.org)

Without a tra­jec­tory change, the de­vel­op­ment of AGI is likely to go badly

Max HMay 29, 2023, 11:42 PM
16 points
2 comments13 min readLW link

Win­ners-take-how-much?

YonatanKMay 29, 2023, 9:56 PM
3 points
2 comments3 min readLW link

Re­ply to a fer­til­ity doc­tor con­cern­ing poly­genic em­bryo screening

GeneSmithMay 29, 2023, 9:50 PM
59 points
6 comments8 min readLW link

Sen­tience matters

So8resMay 29, 2023, 9:25 PM
143 points
96 comments2 min readLW link

Wikipe­dia as an in­tro­duc­tion to the al­ign­ment problem

SoerenMindMay 29, 2023, 6:43 PM
83 points
10 comments1 min readLW link
(en.wikipedia.org)

[Question] What are some of the best in­tro­duc­tions/​break­downs of AI ex­is­ten­tial risk for those un­fa­mil­iar?

Isaac KingMay 29, 2023, 5:04 PM
17 points
2 comments1 min readLW link

Creat­ing Flash­cards with LLMs

Diogo CruzMay 29, 2023, 4:55 PM
15 points
3 comments9 min readLW link