Gen­er­a­tive, Epi­sodic Ob­jec­tives for Safe AI

Michael GlassOct 5, 2022, 11:18 PM
11 points
3 comments8 min readLW link

Depen­dency Tree For The Devel­op­ment Of Plate Tectonics

ElizabethOct 5, 2022, 10:40 PM
41 points
3 comments4 min readLW link
(acesounderglass.com)

[Question] How does an­thropic rea­son­ing and illu­sion­ism/​elimini­tivism in­ter­act?

ShiroeOct 5, 2022, 10:31 PM
5 points
18 comments1 min readLW link

[Question] Find­ing Great Tutors

Ulisse MiniOct 5, 2022, 10:08 PM
27 points
5 comments1 min readLW link

Progress links and tweets, 2022-10-05

jasoncrawfordOct 5, 2022, 7:24 PM
9 points
1 comment1 min readLW link
(rootsofprogress.org)

A blog post is a very long and com­plex search query to find fas­ci­nat­ing peo­ple and make them route in­ter­est­ing stuff to your inbox

Henrik KarlssonOct 5, 2022, 7:07 PM
89 points
12 comments11 min readLW link
(escapingflatland.substack.com)

Neu­ral Tan­gent Ker­nel Distillation

Oct 5, 2022, 6:11 PM
76 points
20 comments8 min readLW link

Track­ing Com­pute Stocks and Flows: Case Stud­ies?

CullenOct 5, 2022, 5:57 PM
11 points
5 comments1 min readLW link

[Linkpost] “Blueprint for an AI Bill of Rights”—Office of Science and Tech­nol­ogy Policy, USA (2022)

Fer32dwt34r3dfszOct 5, 2022, 4:42 PM
9 points
4 comments2 min readLW link
(www.whitehouse.gov)

Paper: Dis­cov­er­ing novel al­gorithms with AlphaTen­sor [Deep­mind]

LawrenceCOct 5, 2022, 4:20 PM
82 points
18 comments1 min readLW link
(www.deepmind.com)

Reflec­tion Mechanisms as an Align­ment tar­get: A fol­low-up survey

Oct 5, 2022, 2:03 PM
15 points
2 comments7 min readLW link

Char­i­ta­ble Reads of Anti-AGI-X-Risk Ar­gu­ments, Part 1

sstichOct 5, 2022, 5:03 AM
3 points
4 comments3 min readLW link

Sleep Training

jefftkOct 5, 2022, 2:10 AM
36 points
4 comments2 min readLW link
(www.jefftk.com)

Looping

Jarred FilmerOct 5, 2022, 1:47 AM
56 points
6 comments2 min readLW link

How are you deal­ing with on­tol­ogy iden­ti­fi­ca­tion?

Erik JennerOct 4, 2022, 11:28 PM
34 points
10 comments3 min readLW link

Smoke with­out fire is scary

Adam JermynOct 4, 2022, 9:08 PM
52 points
22 comments4 min readLW link

Dep­re­cated: Some hu­mans are fit­ness maximizers

Shoshannah TekofskyOct 4, 2022, 7:38 PM
6 points
22 comments6 min readLW link

Will you let your kid play foot­ball?

5houtOct 4, 2022, 5:48 PM
14 points
12 comments2 min readLW link

Quick notes on “mir­ror neu­rons”

Steven ByrnesOct 4, 2022, 5:39 PM
39 points
2 comments2 min readLW link

Fea­ture re­quest: Filter by read/​ upvoted

Nathan YoungOct 4, 2022, 5:17 PM
16 points
5 comments1 min readLW link

Lay­ers Of Mind

PeteGOct 4, 2022, 4:52 PM
−8 points
4 comments2 min readLW link

[Question] Does Google still hire peo­ple via their foo­bar challenge?

AlgonOct 4, 2022, 3:39 PM
11 points
7 comments1 min readLW link

Rus­sia will do a nu­clear test

sanxiynOct 4, 2022, 2:59 PM
3 points
7 comments1 min readLW link

Paper+Sum­mary: OMNIGROK: GROKKING BEYOND ALGORITHMIC DATA

Marius HobbhahnOct 4, 2022, 7:22 AM
46 points
11 comments1 min readLW link
(arxiv.org)

Frontline of AGI Align­ment

SD MarlowOct 4, 2022, 3:47 AM
−10 points
0 comments1 min readLW link
(robothouse.substack.com)

Ad­ver­sar­ial vs Col­lab­o­ra­tive Contexts

jefftkOct 4, 2022, 2:40 AM
31 points
4 comments2 min readLW link
(www.jefftk.com)

Hu­mans aren’t fit­ness maximizers

So8resOct 4, 2022, 1:31 AM
50 points
46 comments5 min readLW link

Self-defeat­ing con­spir­acy the­o­rists and their theories

M. Y. ZuoOct 4, 2022, 12:48 AM
5 points
12 comments3 min readLW link

No free lunch the­o­rem is irrelevant

CatneeOct 4, 2022, 12:21 AM
18 points
7 comments1 min readLW link

The Village and the River Mon­sters… Or: Less Fight­ing, More Brainstorming

ExCephOct 3, 2022, 11:01 PM
7 points
29 comments8 min readLW link
(ginnungagapfoundation.wordpress.com)

Re­call and Re­gur­gi­ta­tion in GPT2

Megan KinnimentOct 3, 2022, 7:35 PM
43 points
1 comment26 min readLW link

Iver­mectin: Much Less Than You Needed To Know

George3d6Oct 3, 2022, 3:02 PM
31 points
10 comments1 min readLW link
(doyourownresearch.substack.com)

If you want to learn tech­ni­cal AI safety, here’s a list of AI safety courses, read­ing lists, and resources

KatWoodsOct 3, 2022, 12:43 PM
12 points
3 comments1 min readLW link

Oc­to­ber Bu­dapest Less Wrong/​ACX meetup

Richard HorvathOct 3, 2022, 10:53 AM
2 points
0 comments1 min readLW link

A re­view of the Bio-An­chors report

jylin04Oct 3, 2022, 10:27 AM
45 points
4 comments1 min readLW link
(docs.google.com)

Data for IRL: What is needed to learn hu­man val­ues?

Jan WehnerOct 3, 2022, 9:23 AM
18 points
6 comments12 min readLW link

Statis­tics for ob­jects with shared identities

Q HomeOct 3, 2022, 9:21 AM
2 points
7 comments4 min readLW link

Vi­su­al­iz­ing Learned Rep­re­sen­ta­tions of Rice Disease

muhia_beeOct 3, 2022, 9:09 AM
7 points
0 comments4 min readLW link
(indecisive-sand-24a.notion.site)

[Question] Is there a cul­ture over­hang?

Aleksi LiimatainenOct 3, 2022, 7:26 AM
18 points
4 comments1 min readLW link

Baby Mon­i­tor with Delay

jefftkOct 3, 2022, 1:40 AM
12 points
13 comments1 min readLW link
(www.jefftk.com)

What makes a prob­a­bil­ity ques­tion “well-defined”? (Part I)

Noah TopperOct 2, 2022, 9:05 PM
14 points
4 comments7 min readLW link
(naivebayes.substack.com)

Against the weird­ness heuristic

Eleni AngelouOct 2, 2022, 7:41 PM
17 points
3 comments2 min readLW link

Easy fix­ing Voting

Charbel-RaphaëlOct 2, 2022, 5:03 PM
12 points
2 comments1 min readLW link

Monthly Shorts 9/​22, and An Es­say in Defense of Technodeterminism

CelerOct 2, 2022, 4:10 PM
10 points
1 comment5 min readLW link
(keller.substack.com)

Open & Wel­come Thread—Oct 2022

niplavOct 2, 2022, 11:04 AM
17 points
62 comments1 min readLW link

[Question] Any fur­ther work on AI Safety Suc­cess Sto­ries?

KriegerOct 2, 2022, 9:53 AM
8 points
6 comments1 min readLW link

Re­quest for feed­back on sam­ple blurbs for the EA fan­tasy novel I wrote

Timothy UnderwoodOct 2, 2022, 9:01 AM
8 points
0 comments2 min readLW link

Paper: Large Lan­guage Models Can Self-im­prove [Linkpost]

Evan R. MurphyOct 2, 2022, 1:29 AM
52 points
15 comments1 min readLW link
(openreview.net)

Four us­ages of “loss” in AI

TurnTroutOct 2, 2022, 12:52 AM
46 points
18 comments4 min readLW link

The com­pu­ta­tional com­plex­ity of progress

MalmesburyOct 1, 2022, 9:50 PM
21 points
2 comments10 min readLW link