Po­ten­tial Align­ment men­tal tool: Keep­ing track of the types

Donald HobsonNov 22, 2021, 8:05 PM
29 points
1 comment2 min readLW link

Yud­kowsky and Chris­ti­ano dis­cuss “Take­off Speeds”

Eliezer YudkowskyNov 22, 2021, 7:35 PM
210 points
176 comments60 min readLW link1 review

Mo­rally un­der­defined situ­a­tions can be deadly

Stuart_ArmstrongNov 22, 2021, 2:48 PM
17 points
8 comments2 min readLW link

A Bayesian Ag­gre­ga­tion Paradox

JsevillamolNov 22, 2021, 10:39 AM
87 points
23 comments7 min readLW link

[Question] Do fac­tored sets elu­ci­date any­thing about how to up­date ev­ery­day be­liefs?

TekhneMakreNov 22, 2021, 6:51 AM
5 points
1 comment1 min readLW link

Even if you’re right, you’re wrong

DanielFilanNov 22, 2021, 5:40 AM
17 points
5 comments1 min readLW link
(danielfilan.com)

The Meta-Puzzle

DanielFilanNov 22, 2021, 5:30 AM
23 points
27 comments3 min readLW link
(danielfilan.com)

Some real ex­am­ples of gra­di­ent hacking

Oliver SourbutNov 22, 2021, 12:11 AM
15 points
8 comments2 min readLW link

“The Wis­dom of the Lazy Teacher”

Richard_KennawayNov 21, 2021, 9:11 PM
16 points
5 comments1 min readLW link

Vi­talik: Cryp­toe­co­nomics and X-Risk Re­searchers Should Listen to Each Other More

Emerson SpartzNov 21, 2021, 6:53 PM
47 points
9 comments5 min readLW link

Giv­ing Up On T-Mobile

jefftkNov 21, 2021, 4:50 PM
13 points
6 comments2 min readLW link
(www.jefftk.com)

From lan­guage to ethics by au­to­mated reasoning

Michele CampoloNov 21, 2021, 3:16 PM
4 points
4 comments6 min readLW link

Split and Commit

Duncan Sabien (Inactive)Nov 21, 2021, 6:27 AM
191 points
34 comments7 min readLW link1 review

What’s the weirdest way to win this game?

Adam ScherlisNov 21, 2021, 5:18 AM
9 points
5 comments1 min readLW link
(adam.scherlis.com)

Eat the cute an­i­mals instead

Andrew VlahosNov 21, 2021, 1:06 AM
−4 points
2 comments1 min readLW link

Chris Voss ne­go­ti­a­tion MasterClass: review

VipulNaikNov 20, 2021, 10:39 PM
70 points
15 comments33 min readLW link

ACX Mon­treal Meetup Dec 4 2021

ENov 20, 2021, 5:49 PM
8 points
0 comments1 min readLW link

The Maker of MIND

Tomás B.Nov 20, 2021, 4:28 PM
112 points
19 comments11 min readLW link

South Bay ACX/​LW Meetup—CHANGED LOCATION

ISNov 20, 2021, 2:42 PM
11 points
0 comments1 min readLW link

The Em­peror’s New Clothes: a story of mo­ti­vated stupidity

David Hugh-JonesNov 20, 2021, 1:24 PM
10 points
5 comments3 min readLW link
(wyclif.substack.com)

[Book Re­view] “Sorceror’s Ap­pren­tice” by Tahir Shah

lsusrNov 20, 2021, 11:29 AM
92 points
11 comments7 min readLW link

Com­pe­tence/​Confidence

Duncan Sabien (Inactive)Nov 20, 2021, 8:59 AM
60 points
19 comments1 min readLW link

Awe­some-github Post-Scarcity List

lorepieriNov 20, 2021, 8:47 AM
3 points
6 comments1 min readLW link

A Cer­tain For­mal­iza­tion of Cor­rigi­bil­ity Is VNM-Incoherent

TurnTroutNov 20, 2021, 12:30 AM
68 points
24 comments8 min readLW link

More de­tailed pro­posal for mea­sur­ing al­ign­ment of cur­rent models

Beth BarnesNov 20, 2021, 12:03 AM
31 points
0 comments8 min readLW link

Am­bi­tious Altru­is­tic Soft­ware Eng­ineer­ing Efforts: Op­por­tu­ni­ties and Benefits

ozziegooenNov 19, 2021, 5:55 PM
42 points
1 comment9 min readLW link
(forum.effectivealtruism.org)

[Question] Which booster shot to get and when?

NormanPerlmutterNov 19, 2021, 8:52 AM
22 points
17 comments2 min readLW link

Good­hart: Endgame

Charlie SteinerNov 19, 2021, 1:26 AM
25 points
3 comments8 min readLW link

Re­ac­tion and Re­ply to Sasha Chapin on Bad In-group Norms

Nicholas / Heather KrossNov 19, 2021, 1:13 AM
6 points
0 comments3 min readLW link
(www.thinkingmuchbetter.com)

[Question] Does any­one know what Marvin Min­sky is talk­ing about here?

delton137Nov 19, 2021, 12:56 AM
1 point
6 comments3 min readLW link

How To Get Into In­de­pen­dent Re­search On Align­ment/​Agency

johnswentworthNov 19, 2021, 12:00 AM
356 points
38 comments13 min readLW link2 reviews

“Ac­qui­si­tion of Chess Knowl­edge in AlphaZero”: prob­ing AZ over time

jsdNov 18, 2021, 11:24 PM
11 points
9 commentsLW link
(arxiv.org)

Ngo and Yud­kowsky on AI ca­pa­bil­ity gains

Nov 18, 2021, 10:19 PM
131 points
61 comments39 min readLW link1 review

Covid 11/​18: Paxlovid Re­mains Illegal

ZviNov 18, 2021, 3:50 PM
55 points
36 comments14 min readLW link
(thezvi.wordpress.com)

Satis­ficers Tend To Seek Power: In­stru­men­tal Con­ver­gence Via Retargetability

TurnTroutNov 18, 2021, 1:54 AM
85 points
8 comments17 min readLW link
(www.overleaf.com)

Fore­cast­ing: Zeroth and First Order

jsteinhardtNov 18, 2021, 1:30 AM
33 points
6 comments5 min readLW link
(bounded-regret.ghost.io)

Ex­pe­rience on Methotrexate

jefftkNov 17, 2021, 10:40 PM
13 points
0 comments2 min readLW link
(www.jefftk.com)

Ap­pli­ca­tions for AI Safety Camp 2022 Now Open!

adamShimiNov 17, 2021, 9:42 PM
47 points
3 comments1 min readLW link

[Question] Did EcoHealth cre­ate SARS-CoV-2?

jamalNov 17, 2021, 8:42 PM
3 points
7 comments1 min readLW link

On Rais­ing Awareness

Tomás B.Nov 17, 2021, 5:12 PM
21 points
10 comments3 min readLW link

Sasha Chapin on bad so­cial norms in ra­tio­nal­ity/​EA

Kaj_SotalaNov 17, 2021, 9:43 AM
51 points
22 comments5 min readLW link
(sashachapin.substack.com)

[Question] What are the mu­tual benefits of AGI-hu­man col­lab­o­ra­tion that would oth­er­wise be un­ob­tain­able?

M. Y. ZuoNov 17, 2021, 3:09 AM
1 point
4 comments1 min readLW link

Quadratic Vot­ing and Collusion

leogaoNov 17, 2021, 12:19 AM
41 points
24 comments2 min readLW link

Tak­ing a sim­plified model

dominicqNov 16, 2021, 10:21 PM
9 points
8 comments1 min readLW link

The Greedy Doc­tor Problem

JanNov 16, 2021, 10:06 PM
6 points
10 comments12 min readLW link
(universalprior.substack.com)

Equity pre­mium puzzles

Nov 16, 2021, 8:50 PM
20 points
4 comments12 min readLW link
(www.metaculus.com)

Why I am no longer driven

dominicqNov 16, 2021, 8:43 PM
71 points
16 comments4 min readLW link

Su­per in­tel­li­gent AIs that don’t re­quire alignment

Yair HalberstadtNov 16, 2021, 7:55 PM
10 points
2 comments6 min readLW link

Why Save The Drown­ing Child: Ethics Vs Theory

Raymond DouglasNov 16, 2021, 7:07 PM
17 points
12 comments4 min readLW link

Two Stupid AI Align­ment Ideas

aphyerNov 16, 2021, 4:13 PM
27 points
3 comments4 min readLW link