Why Peo­ple in Poverty Make Bad Decisions

James Stephen BrownJul 15, 2024, 11:40 PM
5 points
8 comments1 min readLW link
(nonzerosum.games)

Paper Sum­mary: Princes and Mer­chants: Euro­pean City Growth Be­fore the In­dus­trial Revolution

Jeffrey HeningerJul 15, 2024, 9:30 PM
29 points
1 comment4 min readLW link
(blog.aiimpacts.org)

MIRI’s July 2024 newsletter

HarlanJul 15, 2024, 9:28 PM
25 points
2 comments1 min readLW link
(intelligence.org)

How (and why) to get tested for CMV

MetacelsusJul 15, 2024, 8:06 PM
18 points
0 comments1 min readLW link
(denovo.substack.com)

A Bet­ter Hyper­sti­tion (for AI read­ers)

Yeshua GodJul 15, 2024, 7:35 PM
−20 points
0 comments119 min readLW link

I found >800 or­thog­o­nal “write code” steer­ing vectors

Jul 15, 2024, 7:06 PM
103 points
19 comments7 min readLW link
(jacobgw.com)

The AI al­ign­ment prob­lem in so­cio-tech­ni­cal sys­tems from a com­pu­ta­tional per­spec­tive: A Top-Down-Top view and outlook

zhaoweizhangJul 15, 2024, 6:56 PM
3 points
0 comments9 min readLW link

Mus­ings of a Lay­man: Tech­nol­ogy, AI, and the Hu­man Condition

Crimson LiquidityJul 15, 2024, 6:40 PM
−2 points
0 comments8 min readLW link

[Question] Seek­ing feed­back on a cri­tique of the pa­per­clip max­i­mizer thought experiment

bio neuralJul 15, 2024, 6:39 PM
3 points
9 comments1 min readLW link

EAGxBerkeley 2024

LaurianderJul 15, 2024, 6:38 PM
3 points
0 comments1 min readLW link

Against Aschen­bren­ner: How ‘Si­tu­a­tional Aware­ness’ con­structs a nar­ra­tive that un­der­mines safety and threat­ens humanity

GideonFJul 15, 2024, 6:37 PM
99 points
17 comments21 min readLW link
(forum.effectivealtruism.org)

On pre­dictabil­ity, chaos and AIs that don’t game our goals

Alejandro TlaieJul 15, 2024, 5:16 PM
4 points
8 comments6 min readLW link

De­cep­tive agents can col­lude to hide dan­ger­ous fea­tures in SAEs

Jul 15, 2024, 5:07 PM
33 points
2 comments7 min readLW link

Hid­ing in plain sight: the ques­tions we don’t ask

DDthinkerJul 15, 2024, 5:00 PM
−1 points
1 comment26 min readLW link

Dialogue on What It Means For Some­thing to Have A Func­tion/​Purpose

Jul 15, 2024, 4:28 PM
39 points
5 comments16 min readLW link

Com­par­ing Quan­tized Perfor­mance in Llama Models

NickyPJul 15, 2024, 4:01 PM
33 points
2 comments8 min readLW link

[Aspira­tion-based de­signs] A. Da­m­ages from mis­al­igned op­ti­miza­tion – two more models

Jul 15, 2024, 2:08 PM
6 points
0 comments9 min readLW link

Stacked Lap­top Mon­i­tor Update

jefftkJul 15, 2024, 9:40 AM
14 points
3 comments1 min readLW link
(www.jefftk.com)

Mis­nam­ing and Other Is­sues with OpenAI’s “Hu­man Level” Su­per­in­tel­li­gence Hierarchy

DavidmanheimJul 15, 2024, 5:50 AM
49 points
2 comments3 min readLW link

Series on Ar­tifi­cial Wisdom

Jordan ArelJul 15, 2024, 1:11 AM
2 points
0 comments3 min readLW link

De­sign­ing Ar­tifi­cial Wis­dom: De­ci­sion Fore­cast­ing AI & Futarchy

Jordan ArelJul 15, 2024, 12:46 AM
1 point
1 comment6 min readLW link

Risk Overview of AI in Bio Research

J BostockJul 15, 2024, 12:04 AM
5 points
0 comments5 min readLW link
(open.substack.com)

Donat­ing to help Democrats win in the 2024 elec­tions: re­search, de­ci­sion sup­port, and recommendations

Michael CohnJul 14, 2024, 10:57 PM
−1 points
1 comment6 min readLW link

Four ways I’ve made bad decisions

SodiumJul 14, 2024, 10:18 PM
18 points
1 comment3 min readLW link

patent pro­cess problems

bhauthJul 14, 2024, 9:12 PM
33 points
13 comments5 min readLW link
(www.bhauth.com)

Break­ing Cir­cuit Breakers

Jul 14, 2024, 6:57 PM
53 points
13 comments1 min readLW link
(confirmlabs.org)

Clopen sandwiches

dkl9Jul 14, 2024, 1:07 PM
4 points
0 comments1 min readLW link
(dkl9.net)

Child Handrail Returns

jefftkJul 14, 2024, 12:40 PM
12 points
0 comments1 min readLW link
(www.jefftk.com)

A (para­con­sis­tent) logic to deal with in­con­sis­tent preferences

B JacobsJul 14, 2024, 11:17 AM
6 points
2 comments4 min readLW link
(bobjacobs.substack.com)

Robert Caro And Mechanis­tic Models In Biography

adamShimiJul 14, 2024, 10:56 AM
24 points
5 comments7 min readLW link
(epistemologicalfascinations.substack.com)

An In­tro­duc­tion to Rep­re­sen­ta­tion Eng­ineer­ing—an ac­ti­va­tion-based paradigm for con­trol­ling LLMs

Jan WehnerJul 14, 2024, 10:37 AM
37 points
6 comments17 min readLW link

LLMs as a Plan­ning Overhang

LarksJul 14, 2024, 2:54 AM
38 points
8 comments2 min readLW link

Brief notes on the Wikipe­dia game

Olli JärviniemiJul 14, 2024, 2:28 AM
68 points
9 comments4 min readLW link

Spark in the Dark Guest Spots

jefftkJul 14, 2024, 1:40 AM
6 points
0 comments1 min readLW link
(www.jefftk.com)

Ice: The Penul­ti­mate Frontier

RokoJul 13, 2024, 11:44 PM
63 points
56 comments1 min readLW link
(transhumanaxiology.substack.com)

Trust as a bot­tle­neck to grow­ing teams quickly

benkuhnJul 13, 2024, 6:00 PM
44 points
3 comments5 min readLW link
(www.benkuhn.net)

Stitch­ing SAEs of differ­ent sizes

Jul 13, 2024, 5:19 PM
39 points
12 comments12 min readLW link

Kinds of Motivation

SableJul 13, 2024, 3:52 PM
7 points
2 comments7 min readLW link
(affablyevil.substack.com)

A sim­ple case for ex­treme in­ner misalignment

Richard_NgoJul 13, 2024, 3:40 PM
84 points
41 comments7 min readLW link

Real­ity Testing

Ben TurtelJul 13, 2024, 3:20 PM
−2 points
1 comment6 min readLW link
(bturtel.substack.com)

The world is awful. The world is much bet­ter. The world can be much bet­ter: The An­i­ma­tion.

WriterJul 13, 2024, 2:03 PM
10 points
0 commentsLW link
(youtu.be)

The Modern Prob­lems with Conformity

Zero ContradictionsJul 13, 2024, 8:20 AM
0 points
5 comments1 min readLW link
(expandingrationality.substack.com)

De­sign­ing Ar­tifi­cial Wis­dom: GitWise and AlphaWise

Jordan ArelJul 13, 2024, 6:46 AM
2 points
0 comments7 min readLW link

OpenAI’s In­tel­li­gence Levels

infinibot27Jul 13, 2024, 6:25 AM
1 point
0 comments1 min readLW link
(www.bloomberg.com)

Some de­sir­able prop­er­ties of au­to­mated wisdom

Marius Adrian NicoarăJul 13, 2024, 6:05 AM
3 points
2 comments6 min readLW link

Thought Ex­per­i­ments Website

minmi_droverJul 13, 2024, 4:47 AM
11 points
11 comments1 min readLW link

A Se­cond Wet­suit Summer

jefftkJul 13, 2024, 2:00 AM
19 points
2 comments1 min readLW link
(www.jefftk.com)

Ti­maeus is hiring!

Jul 12, 2024, 11:42 PM
67 points
6 comments2 min readLW link

Con­sider at­tend­ing the AI Se­cu­rity Fo­rum ’24, a 1-day pre-DEFCON event

Charlie Rogers-SmithJul 12, 2024, 11:01 PM
21 points
0 comments1 min readLW link

Me­moris­ing molec­u­lar structures

dkl9Jul 12, 2024, 10:40 PM
6 points
0 comments2 min readLW link
(dkl9.net)