Sin­gu­lar learn­ing the­ory and bridg­ing from ML to brain emulations

Nov 1, 2023, 9:31 PM
26 points
16 comments29 min readLW link

My thoughts on the so­cial re­sponse to AI risk

Matthew BarnettNov 1, 2023, 9:17 PM
157 points
37 comments10 min readLW link

Re­ac­tions to the Ex­ec­u­tive Order

ZviNov 1, 2023, 8:40 PM
77 points
4 comments29 min readLW link
(thezvi.wordpress.com)

Dario Amodei’s pre­pared re­marks from the UK AI Safety Sum­mit, on An­thropic’s Re­spon­si­ble Scal­ing Policy

Zac Hatfield-DoddsNov 1, 2023, 6:10 PM
83 points
1 comment4 min readLW link
(www.anthropic.com)

Book Re­view: Deter­mined by Sapolsky

Kailuo WangNov 1, 2023, 5:37 PM
1 point
0 comments7 min readLW link

AI Align­ment: A Com­pre­hen­sive Survey

Stephen McAleerNov 1, 2023, 5:35 PM
20 points
1 comment1 min readLW link
(arxiv.org)

A list of all the dead­lines in Bi­den’s Ex­ec­u­tive Order on AI

Valentin BaltadzhievNov 1, 2023, 5:14 PM
26 points
2 comments11 min readLW link

2023 LessWrong Com­mu­nity Cen­sus, Re­quest for Comments

ScrewtapeNov 1, 2023, 4:32 PM
43 points
37 comments2 min readLW link

[Question] Snap­shot of nar­ra­tives and frames against reg­u­lat­ing AI

Jan_KulveitNov 1, 2023, 4:30 PM
36 points
19 comments3 min readLW link

Com­men­sal Institutions

SableNov 1, 2023, 4:01 PM
8 points
12 comments4 min readLW link
(affablyevil.substack.com)

ChatGPT’s On­tolog­i­cal Land­scape

Bill BenzonNov 1, 2023, 3:12 PM
7 points
0 comments4 min readLW link

On the Ex­ec­u­tive Order

ZviNov 1, 2023, 2:20 PM
100 points
4 comments30 min readLW link
(thezvi.wordpress.com)

Chi­nese sci­en­tists ac­knowl­edge xrisk & call for in­ter­na­tional reg­u­la­tory body [Linkpost]

Orpheus16Nov 1, 2023, 1:28 PM
44 points
4 comments1 min readLW link
(www.ft.com)

[Question] Fore­cast­ing Ques­tions: What do you want to pre­dict on AI?

Nathan YoungNov 1, 2023, 1:17 PM
7 points
2 comments1 min readLW link

Mis­sion Im­pos­si­ble: Dead Reck­on­ing Part 1 AI Takeaways

ZviNov 1, 2023, 12:52 PM
47 points
13 comments6 min readLW link

Ro­bust­ness of Con­trast-Con­sis­tent Search to Ad­ver­sar­ial Prompting

Nov 1, 2023, 12:46 PM
18 points
1 comment7 min readLW link

The Bletch­ley Dec­la­ra­tion on AI Safety

Hauke HillebrandtNov 1, 2023, 11:44 AM
17 points
0 commentsLW link
(www.gov.uk)

Bay Win­ter Sols­tice 2023: Song & speech auditions

tcheasdfjklNov 1, 2023, 4:17 AM
17 points
2 comments1 min readLW link

On Hav­ing No Clue

Chris_LeongNov 1, 2023, 1:36 AM
20 points
11 comments1 min readLW link

Balanc­ing Se­cu­rity Mind­set with Col­lab­o­ra­tive Re­search: A Proposal

MadHatterNov 1, 2023, 12:46 AM
9 points
3 comments4 min readLW link

Com­pu­ta­tional Ap­proaches to Pathogen Detection

jefftkNov 1, 2023, 12:30 AM
32 points
5 comments5 min readLW link
(www.jefftk.com)

Thoughts on the AI Safety Sum­mit com­pany policy re­quests and responses

So8resOct 31, 2023, 11:54 PM
169 points
14 comments10 min readLW link

AISN #25: White House Ex­ec­u­tive Order on AI, UK AI Safety Sum­mit, and Progress on Vol­un­tary Eval­u­a­tions of AI Risks

Dan HOct 31, 2023, 7:34 PM
35 points
1 comment6 min readLW link
(newsletter.safe.ai)

If AIs be­come self-aware, what re­li­gion will they have?

mnvrOct 31, 2023, 5:29 PM
−17 points
3 comments4 min readLW link

Self-Blinded L-Thea­nine RCT

niplavOct 31, 2023, 3:24 PM
53 points
12 comments3 min readLW link

AI Safety 101 - Chap­ter 5.2 - Un­re­stricted Ad­ver­sar­ial Training

Charbel-RaphaëlOct 31, 2023, 2:34 PM
17 points
0 comments19 min readLW link

Prevent­ing Lan­guage Models from hid­ing their reasoning

Oct 31, 2023, 2:34 PM
119 points
15 comments12 min readLW link1 review

AI Safety 101 - Chap­ter 5.1 - Debate

Charbel-RaphaëlOct 31, 2023, 2:29 PM
15 points
0 comments13 min readLW link

M&A in AI

Hauke HillebrandtOct 31, 2023, 12:20 PM
2 points
0 commentsLW link

Urg­ing an In­ter­na­tional AI Treaty: An Open Letter

Olli JärviniemiOct 31, 2023, 11:26 AM
48 points
2 comments1 min readLW link
(aitreaty.org)

[Closed] Agent Foun­da­tions track in MATS

Vanessa KosoyOct 31, 2023, 8:12 AM
54 points
1 comment1 min readLW link
(www.matsprogram.org)

In­trin­sic Drives and Ex­trin­sic Mi­suse: Two In­ter­twined Risks of AI

jsteinhardtOct 31, 2023, 5:10 AM
40 points
0 comments12 min readLW link
(bounded-regret.ghost.io)

Fo­cus on ex­is­ten­tial risk is a dis­trac­tion from the real is­sues. A false fallacy

Nik SamoylovOct 30, 2023, 11:42 PM
−19 points
11 comments2 min readLW link

Will re­leas­ing the weights of large lan­guage mod­els grant wide­spread ac­cess to pan­demic agents?

jefftkOct 30, 2023, 6:22 PM
47 points
25 commentsLW link
(arxiv.org)

[Linkpost] Two ma­jor an­nounce­ments in AI gov­er­nance today

AngélinaOct 30, 2023, 5:28 PM
1 point
1 comment1 min readLW link
(www.whitehouse.gov)

Grokking Beyond Neu­ral Networks

Jack MillerOct 30, 2023, 5:28 PM
10 points
0 comments2 min readLW link
(arxiv.org)

Re­sponse to “Co­or­di­nated paus­ing: An eval­u­a­tion-based co­or­di­na­tion scheme for fron­tier AI de­vel­op­ers”

Matthew WeardenOct 30, 2023, 5:27 PM
5 points
2 comments6 min readLW link
(matthewwearden.co.uk)

Jailbreak and Guard Aligned Lan­guage Models with Only Few In-Con­text Demonstrations

Zeming WeiOct 30, 2023, 5:22 PM
3 points
1 comment1 min readLW link

5 Rea­sons Why Govern­ments/​Mili­taries Already Want AI for In­for­ma­tion Warfare

trevorOct 30, 2023, 4:30 PM
32 points
0 comments10 min readLW link

[Linkpost] Bi­den-Har­ris Ex­ec­u­tive Order on AI

berenOct 30, 2023, 3:20 PM
3 points
0 comments1 min readLW link

AI Align­ment [progress] this Week (10/​29/​2023)

Logan ZoellnerOct 30, 2023, 3:02 PM
15 points
4 comments6 min readLW link
(midwitalignment.substack.com)

Im­prov­ing the Welfare of AIs: A Nearcasted Proposal

ryan_greenblattOct 30, 2023, 2:51 PM
114 points
9 comments20 min readLW link1 review

Pres­i­dent Bi­den Is­sues Ex­ec­u­tive Order on Safe, Se­cure, and Trust­wor­thy Ar­tifi­cial Intelligence

Tristan WilliamsOct 30, 2023, 11:15 AM
171 points
39 commentsLW link
(www.whitehouse.gov)

GPT-2 XL’s ca­pac­ity for co­her­ence and on­tol­ogy clustering

MiguelDevOct 30, 2023, 9:24 AM
6 points
2 comments41 min readLW link

Char­bel-Raphaël and Lu­cius dis­cuss interpretability

Oct 30, 2023, 5:50 AM
111 points
7 comments21 min readLW link

Multi-Win­ner 3-2-1 Voting

Yoav RavidOct 30, 2023, 3:31 AM
14 points
6 comments3 min readLW link

math ter­minol­ogy as convolution

bhauthOct 30, 2023, 1:05 AM
34 points
1 comment4 min readLW link
(www.bhauth.com)

Grokking, mem­o­riza­tion, and gen­er­al­iza­tion — a discussion

Oct 29, 2023, 11:17 PM
75 points
11 comments23 min readLW link

Comp Sci in 2027 (Short story by Eliezer Yud­kowsky)

sudoOct 29, 2023, 11:09 PM
203 points
24 comments10 min readLW link1 review
(nitter.net)

Math­e­mat­i­cally-Defined Op­ti­miza­tion Cap­tures A Lot of Use­ful Information

J BostockOct 29, 2023, 5:17 PM
19 points
0 comments2 min readLW link