AXRP Epi­sode 18 - Con­cept Ex­trap­o­la­tion with Stu­art Armstrong

DanielFilanSep 3, 2022, 11:12 PM
12 points
1 comment39 min readLW link

An Up­date on Academia vs. In­dus­try (one year into my fac­ulty job)

David Scott Krueger (formerly: capybaralet)Sep 3, 2022, 8:43 PM
122 points
18 comments4 min readLW link

[Question] Re­quest for Align­ment Re­search Pro­ject Recommendations

Rauno ArikeSep 3, 2022, 3:29 PM
10 points
2 comments1 min readLW link

Three sce­nar­ios of pseudo-al­ign­ment

Eleni AngelouSep 3, 2022, 12:47 PM
9 points
0 comments3 min readLW link

Bugs or Fea­tures?

qbolecSep 3, 2022, 7:04 AM
73 points
9 comments2 min readLW link

[Ex­plo­ra­tory] Seper­ate ex­plo­ra­tory writ­ing from pub­lic writing

Johannes C. MayerSep 3, 2022, 2:57 AM
6 points
2 comments1 min readLW link

We may be able to see sharp left turns coming

Sep 3, 2022, 2:55 AM
54 points
29 comments1 min readLW link

[Ex­plo­ra­tory] Ex­plo­ra­tory Writ­ing Info

Johannes C. MayerSep 3, 2022, 2:50 AM
3 points
3 comments1 min readLW link

[Question] Can some­one ex­plain to me why most re­searchers think al­ign­ment is prob­a­bly some­thing that is hu­manly tractable?

iamthouthouartiSep 3, 2022, 1:12 AM
32 points
11 comments1 min readLW link

Be­havi­our Man­i­folds and the Hes­sian of the To­tal Loss—Notes and Criticism

carboniferous_umbraculum Sep 3, 2022, 12:15 AM
35 points
5 comments6 min readLW link

Sticky goals: a con­crete ex­per­i­ment for un­der­stand­ing de­cep­tive alignment

evhubSep 2, 2022, 9:57 PM
39 points
13 comments3 min readLW link

Agency en­g­ineer­ing: is AI-al­ign­ment “to hu­man in­tent” enough?

catubcSep 2, 2022, 6:14 PM
9 points
10 comments6 min readLW link

Hanover, Ger­many—ACX Mee­tups Every­where 2022

eikowagenknechtSep 2, 2022, 5:31 PM
2 points
0 comments1 min readLW link

Laz­i­ness in AI

Richard HenageSep 2, 2022, 5:04 PM
13 points
5 comments1 min readLW link

Ex­port­ing Han­gouts History

jefftkSep 2, 2022, 3:00 PM
20 points
0 comments2 min readLW link
(www.jefftk.com)

Simulators

janusSep 2, 2022, 12:45 PM
631 points
168 comments41 min readLW link8 reviews
(generative.ink)

Lev­el­ling Up in AI Safety Re­search Engineering

Gabe MSep 2, 2022, 4:59 AM
58 points
9 comments17 min readLW link

Stop Dis­cour­ag­ing Microwave For­mula Preparation

jefftkSep 2, 2022, 2:10 AM
68 points
12 comments2 min readLW link
(www.jefftk.com)

A Richly In­ter­ac­tive AGI Align­ment Chart

lisperatiSep 2, 2022, 12:44 AM
14 points
6 comments1 min readLW link

Ap­pendix: How to run a suc­cess­ful Ham­ming circle

CFAR!DuncanSep 2, 2022, 12:22 AM
41 points
6 comments7 min readLW link

Re­place­ment for PONR concept

Daniel KokotajloSep 2, 2022, 12:09 AM
58 points
6 comments2 min readLW link

AI co­or­di­na­tion needs clear wins

evhubSep 1, 2022, 11:41 PM
147 points
16 comments2 min readLW link1 review

Short story spec­u­lat­ing on pos­si­ble ram­ifi­ca­tions of AI on the art world

YitzSep 1, 2022, 9:15 PM
30 points
8 comments3 min readLW link
(archiveofourown.org)

Why was progress so slow in the past?

jasoncrawfordSep 1, 2022, 8:26 PM
54 points
31 comments6 min readLW link
(rootsofprogress.org)

AI Safety and Neigh­bor­ing Com­mu­ni­ties: A Quick-Start Guide, as of Sum­mer 2022

Sam BowmanSep 1, 2022, 7:15 PM
76 points
2 comments7 min readLW link

Gra­di­ent Hacker De­sign Prin­ci­ples From Biology

johnswentworthSep 1, 2022, 7:03 PM
60 points
13 comments3 min readLW link

Book re­view: Put Your Ass Where Your Heart Wants to Be

RuhulSep 1, 2022, 6:21 PM
1 point
2 comments10 min readLW link

A Sur­vey of Foun­da­tional Meth­ods in In­verse Re­in­force­ment Learning

adamkSep 1, 2022, 6:21 PM
27 points
0 comments12 min readLW link

I Tripped and Be­came GPT! (And How This Up­dated My Timelines)

FrankophoneSep 1, 2022, 5:56 PM
31 points
0 comments4 min readLW link

[Question] Fixed point the­ory (lo­cally (α,β,ψ) dom­i­nated con­trac­tive con­di­tion)

muzammilSep 1, 2022, 5:56 PM
0 points
3 comments1 min readLW link

Align­ment is hard. Com­mu­ni­cat­ing that, might be harder

Eleni AngelouSep 1, 2022, 4:57 PM
7 points
8 comments3 min readLW link

Covid 9/​1/​22: Meet the New Booster

ZviSep 1, 2022, 2:00 PM
41 points
6 comments14 min readLW link
(thezvi.wordpress.com)

A Starter-kit for Ra­tion­al­ity Space

Jesse HooglandSep 1, 2022, 1:04 PM
43 points
0 comments1 min readLW link
(github.com)

Pon­der­ing the paucity of vol­canic pro­fan­ity post Pom­peii perusal

CraigMichaelSep 1, 2022, 9:29 AM
21 points
2 comments15 min readLW link

In­fra-Ex­er­cises, Part 1

Sep 1, 2022, 5:06 AM
62 points
10 comments1 min readLW link

Strat­egy For Con­di­tion­ing Gen­er­a­tive Models

Sep 1, 2022, 4:34 AM
31 points
4 comments18 min readLW link

Safety Com­mit­tee Resources

jefftkSep 1, 2022, 2:30 AM
22 points
2 comments1 min readLW link
(www.jefftk.com)

Progress links and tweets, 2022-08-31

jasoncrawfordAug 31, 2022, 9:54 PM
13 points
4 comments1 min readLW link
(rootsofprogress.org)

Enantiodromia

ChristianKlAug 31, 2022, 9:13 PM
38 points
7 comments3 min readLW link

[Question] Sup­pos­ing Europe is headed for a se­ri­ous en­ergy crisis this win­ter, what can/​should one do as an in­di­vi­d­ual to pre­pare?

Erich_GrunewaldAug 31, 2022, 7:28 PM
18 points
13 comments1 min readLW link

New 80,000 Hours prob­lem pro­file on ex­is­ten­tial risks from AI

Benjamin HiltonAug 31, 2022, 5:36 PM
28 points
6 comments7 min readLW link
(80000hours.org)

Grand Theft Education

ZviAug 31, 2022, 11:50 AM
66 points
18 comments20 min readLW link
(thezvi.wordpress.com)

How much im­pact can any one man have?

GregorDeVillainAug 31, 2022, 10:26 AM
9 points
3 comments4 min readLW link

[Question] How might we make bet­ter use of AI ca­pa­bil­ities re­search for al­ign­ment pur­poses?

Jemal YoungAug 31, 2022, 4:19 AM
11 points
4 comments1 min readLW link

[Question] AI Box Ex­per­i­ment: Are peo­ple still in­ter­ested?

DoubleAug 31, 2022, 3:04 AM
30 points
13 comments1 min readLW link

OC ACX/​LW in New­port Beach

Michael MichalchikAug 31, 2022, 2:56 AM
1 point
1 comment1 min readLW link

Sur­vey of NLP Re­searchers: NLP is con­tribut­ing to AGI progress; ma­jor catas­tro­phe plausible

Sam BowmanAug 31, 2022, 1:39 AM
91 points
6 comments2 min readLW link

And the word was “God”

pchvykovAug 30, 2022, 9:13 PM
−22 points
4 comments3 min readLW link

Wor­lds Where Iter­a­tive De­sign Fails

johnswentworthAug 30, 2022, 8:48 PM
208 points
30 comments10 min readLW link1 review

In­ner Align­ment via Superpowers

Aug 30, 2022, 8:01 PM
37 points
13 comments4 min readLW link