Is ChatGPT ac­tu­ally fixed now?

sjadlerMay 8, 2025, 11:34 PM
17 points
0 comments1 min readLW link
(stevenadler.substack.com)

Post EAG Lon­don AI x-Safety Co-work­ing Retreat

plexMay 8, 2025, 11:00 PM
10 points
0 comments1 min readLW link

a brief cri­tique of reduction

Vadim GolubMay 8, 2025, 10:43 PM
−17 points
4 comments2 min readLW link

Video & tran­script: Challenges for Safe & Benefi­cial Brain-Like AGI

Steven ByrnesMay 8, 2025, 9:11 PM
24 points
0 comments18 min readLW link

Ap­pendix: In­ter­pretable by De­sign—Con­straint Sets with Disjoint Limit Points

Ronak_MehtaMay 8, 2025, 9:09 PM
2 points
0 comments2 min readLW link

In­ter­pretable by De­sign—Con­straint Sets with Disjoint Limit Points

Ronak_MehtaMay 8, 2025, 9:08 PM
23 points
1 comment9 min readLW link
(ronakrm.github.io)

Is there a Half-Life for the Suc­cess Rates of AI Agents?

Matrice JacobineMay 8, 2025, 8:10 PM
8 points
0 comments1 min readLW link
(www.tobyord.com)

Misal­ign­ment and Strate­gic Un­der­perfor­mance: An Anal­y­sis of Sand­bag­ging and Ex­plo­ra­tion Hacking

May 8, 2025, 7:06 PM
75 points
1 comment15 min readLW link

Be­hold the Pale Child (es­cap­ing Moloch’s Mad Maze)

rogersbaconMay 8, 2025, 4:36 PM
8 points
16 comments11 min readLW link
(www.secretorum.life)

An al­ign­ment safety case sketch based on debate

May 8, 2025, 3:02 PM
57 points
19 comments25 min readLW link
(arxiv.org)

Mechanis­tic In­ter­pretabil­ity Via Learn­ing Differ­en­tial Equa­tions: AI Safety Camp Pro­ject In­ter­me­di­ate Re­port.

May 8, 2025, 2:45 PM
6 points
0 comments7 min readLW link

AI #115: The Evil Ap­pli­ca­tions Division

ZviMay 8, 2025, 1:40 PM
32 points
3 comments62 min readLW link
(thezvi.wordpress.com)

The Stegano­graphic Po­ten­tials of Lan­guage Models

May 8, 2025, 11:23 AM
9 points
0 comments1 min readLW link

Our bet on whether the AI mar­ket will crash

May 8, 2025, 9:56 AM
22 points
2 comments1 min readLW link

Con­cept-an­chored rep­re­sen­ta­tion en­g­ineer­ing for alignment

Sandy FraserMay 8, 2025, 8:59 AM
5 points
0 comments3 min readLW link

Orthog­o­nal­ity Th­e­sis in lay­man’s terms.

Michael (@lethal_ai)May 8, 2025, 8:31 AM
1 point
0 comments2 min readLW link

Arkose may be clos­ing, but you can help

Victoria BrookMay 8, 2025, 7:28 AM
8 points
0 comments2 min readLW link

Heal­ing pow­ers of med­i­ta­tion or the role of at­ten­tion in hu­moral reg­u­la­tion.

Yaroslav GranowskiMay 8, 2025, 6:48 AM
7 points
0 comments1 min readLW link

Ori­ent­ing Toward Wizard Power

johnswentworthMay 8, 2025, 5:23 AM
512 points
135 comments5 min readLW link

Re­la­tional Align­ment: Trust, Re­pair, and the Emo­tional Work of AI

Priyanka BharadwajMay 8, 2025, 2:44 AM
3 points
0 comments3 min readLW link

There’s more low-hang­ing fruit in in­ter­dis­ci­plinary work thanks to LLMs

ChristianKlMay 7, 2025, 7:48 PM
26 points
2 comments1 min readLW link

OpenAI Claims Non­profit Will Re­tain Nom­i­nal Control

ZviMay 7, 2025, 7:40 PM
65 points
4 comments11 min readLW link
(thezvi.wordpress.com)

So­cial sta­tus games might have “com­pute weight class” in the future

RaemonMay 7, 2025, 6:56 PM
31 points
5 comments2 min readLW link

Events of Low Prob­a­bil­ity: Buri­dan’s Principle

Nikita GladkovMay 7, 2025, 6:46 PM
12 points
0 comments10 min readLW link

[Question] Which jour­nal­ists would you give quotes to? [one jour­nal­ist per com­ment, agree vote for trust­wor­thy]

Nathan YoungMay 7, 2025, 6:39 PM
12 points
26 comments1 min readLW link

Progress = Fewer Bad Moments

ChipmonkMay 7, 2025, 5:33 PM
24 points
9 comments2 min readLW link
(chrislakin.blog)

Please Donate to CAIP (Post 1 of 6 on AI Gover­nance)

Mass_DriverMay 7, 2025, 5:13 PM
113 points
20 comments33 min readLW link

UK AISI’s Align­ment Team: Re­search Agenda

May 7, 2025, 4:33 PM
111 points
2 comments11 min readLW link

Four Pre­dic­tions About OpenAI’s Plans To Re­tain Non­profit Control

garrisonMay 7, 2025, 3:48 PM
12 points
0 commentsLW link
(www.obsolete.pub)

A Dis­ci­plined Way to Avoid Wireheading

amitlevy49May 7, 2025, 3:20 PM
18 points
6 comments5 min readLW link
(ivy0.substack.com)

Reflec­tions on Com­pat­i­bil­ism, On­tolog­i­cal Trans­la­tions, and the Ar­tifi­cial Divine

Mahdi ComplexMay 7, 2025, 12:16 PM
2 points
1 comment22 min readLW link

The His­tor­i­cal Par­allels: Pre­limi­nary Reflection

EQMay 7, 2025, 8:06 AM
3 points
0 comments9 min readLW link
(eqmind.substack.com)

Euro­pean Links (07.05.25)

Martin SustrikMay 7, 2025, 4:20 AM
10 points
0 comments2 min readLW link
(250bpm.substack.com)

[Question] Chess—“Elo” of ran­dom play?

Shankar SivarajanMay 7, 2025, 2:18 AM
10 points
16 comments1 min readLW link

$500 + $500 Bounty Prob­lem: Does An (Ap­prox­i­mately) Deter­minis­tic Max­i­mal Re­dund Always Ex­ist?

May 6, 2025, 11:05 PM
72 points
11 comments3 min readLW link

The The­ory Be­hind Loss Curves

James CamachoMay 6, 2025, 10:22 PM
16 points
3 comments4 min readLW link
(github.com)

Nega­tive Re­sults on Group SAEs

Josh EngelsMay 6, 2025, 9:49 PM
70 points
3 comments8 min readLW link

ACX At­lanta May 2025 Meetup

Steve FrenchMay 6, 2025, 9:00 PM
2 points
0 comments1 min readLW link

[Question] What kind of policy by an AGI would make peo­ple happy?

StanislavKrymMay 6, 2025, 6:05 PM
1 point
2 comments1 min readLW link

Paper High­lights, April ’25

gasteigerjoMay 6, 2025, 2:22 PM
4 points
0 comments7 min readLW link
(aisafetyfrontier.substack.com)

Zucker­berg’s Dystopian AI Vision

ZviMay 6, 2025, 1:50 PM
61 points
7 comments11 min readLW link
(thezvi.wordpress.com)

Will pro­tein de­sign tools solve the snake an­tivenom short­age?

Abhishaike MahajanMay 6, 2025, 1:11 PM
30 points
0 comments17 min readLW link
(www.owlposting.com)

Utah Court Case Over State Law Re­gard­ing “Per­son­hood” for Non­hu­man Intelligences

Stephen MartinMay 6, 2025, 12:54 PM
10 points
3 comments2 min readLW link

Global Risks Weekly Roundup #18/​2025: US tar­iff short­ages, mil­i­tary polic­ing, Gaza famine.

NunoSempereMay 6, 2025, 10:39 AM
31 points
2 comments3 min readLW link
(blog.sentinel-team.org)

OpenAI’s Jig May Be Up

ValeMay 6, 2025, 8:51 AM
3 points
2 comments3 min readLW link

My Rea­sons for Us­ing Anki

Parker ConleyMay 6, 2025, 7:01 AM
10 points
0 comments3 min readLW link
(parconley.com)

It’s ‘Well, ac­tu­ally...’ all the way down

benwrMay 6, 2025, 5:44 AM
40 points
34 comments1 min readLW link
(www.benwr.net)

Five Hinge‑Ques­tions That De­cide Whether AGI Is Five Years Away or Twenty

charlieoneillMay 6, 2025, 2:48 AM
125 points
17 comments5 min readLW link

Non­profit to re­tain con­trol of OpenAI

ArchimedesMay 5, 2025, 11:41 PM
37 points
1 comment1 min readLW link
(openai.com)

Un­ex­pected Con­scious Entities

Gunnar_ZarnckeMay 5, 2025, 10:14 PM
34 points
6 comments6 min readLW link