Ver­nor Vinge, who coined the term “Tech­nolog­i­cal Sin­gu­lar­ity”, dies at 79

Kaj_SotalaMar 21, 2024, 10:14 PM
150 points
25 comments1 min readLW link
(arstechnica.com)

ChatGPT can learn in­di­rect control

Raymond DouglasMar 21, 2024, 9:11 PM
213 points
27 comments1 min readLW link

“Deep Learn­ing” Is Func­tion Approximation

Zack_M_DavisMar 21, 2024, 5:50 PM
98 points
28 comments10 min readLW link
(zackmdavis.net)

A Teacher vs. Every­one Else

ronak69Mar 21, 2024, 5:45 PM
41 points
8 comments2 min readLW link

Static vs Dy­namic Alignment

Gracie GreenMar 21, 2024, 5:44 PM
5 points
0 comments12 min readLW link

On green

Joe CarlsmithMar 21, 2024, 5:38 PM
269 points
35 comments31 min readLW link

Com­par­ing Align­ment to other AGI in­ter­ven­tions: Ex­ten­sions and analysis

Martín SotoMar 21, 2024, 5:30 PM
7 points
0 comments4 min readLW link

The Com­cast Problem

RamblinDashMar 21, 2024, 4:46 PM
1 point
15 comments1 min readLW link

Vi­pas­sana Med­i­ta­tion and Ac­tive In­fer­ence: A Frame­work for Un­der­stand­ing Suffer­ing and its Cessation

sturbMar 21, 2024, 12:32 PM
50 points
8 comments19 min readLW link

AI #56: Black­well That Ends Well

ZviMar 21, 2024, 12:10 PM
34 points
16 comments68 min readLW link
(thezvi.wordpress.com)

An Afford­able CO2 Monitor

Pretentious PenguinMar 21, 2024, 3:06 AM
28 points
1 comment1 min readLW link

Deep­Mind: Eval­u­at­ing Fron­tier Models for Danger­ous Capabilities

Zach Stein-PerlmanMar 21, 2024, 3:00 AM
61 points
8 comments1 min readLW link
(arxiv.org)

Where are the Con­tra Dances?

jefftkMar 21, 2024, 2:00 AM
9 points
0 comments1 min readLW link
(www.jefftk.com)

Slim overview of work one could do to make AI go bet­ter (and a grab-bag of other ca­reer con­sid­er­a­tions)

Chi NguyenMar 20, 2024, 11:17 PM
9 points
1 commentLW link

How does AI solve prob­lems?

Dom PolsinelliMar 20, 2024, 10:29 PM
2 points
0 comments7 min readLW link

What I Learned (Con­clu­sion To “The Sense Of Phys­i­cal Ne­ces­sity”)

LoganStrohlMar 20, 2024, 9:24 PM
34 points
0 comments3 min readLW link

Stage­wise Devel­op­ment in Neu­ral Networks

Mar 20, 2024, 7:54 PM
96 points
1 comment11 min readLW link

On the Glad­stone Report

ZviMar 20, 2024, 7:50 PM
64 points
11 comments40 min readLW link
(thezvi.wordpress.com)

Nat­u­ral La­tents: The Concepts

Mar 20, 2024, 6:21 PM
90 points
18 comments19 min readLW link

Com­par­ing Align­ment to other AGI in­ter­ven­tions: Ba­sic model

Martín SotoMar 20, 2024, 6:17 PM
12 points
4 comments7 min readLW link

New re­port: Safety Cases for AI

joshcMar 20, 2024, 4:45 PM
89 points
14 comments1 min readLW link
(twitter.com)

User-in­cli­na­tion-guess­ing al­gorithms: reg­is­ter­ing a goal

ProgramCrafterMar 20, 2024, 3:55 PM
2 points
0 comments2 min readLW link

My MATS Sum­mer 2023 experience

James ChuaMar 20, 2024, 11:26 AM
29 points
0 comments3 min readLW link
(jameschua.net)

[Question] What are the weirdest things a hu­man may want for their own sake?

Mateusz BagińskiMar 20, 2024, 11:15 AM
7 points
16 comments1 min readLW link

[Question] Best *or­ga­ni­za­tion* red-pill books and posts?

lemonhopeMar 20, 2024, 7:01 AM
10 points
2 comments1 min readLW link

Par­ent-Friendly Dance Weekends

jefftkMar 20, 2024, 2:10 AM
16 points
0 comments2 min readLW link
(www.jefftk.com)

[Question] “I Can’t Believe It Both Is and Is Not En­cephal­itis!” Or: What do you do when the ev­i­dence is crazy?

ErhannisMar 19, 2024, 10:08 PM
20 points
3 comments11 min readLW link

Delta’s of Change

Jonas KgomoMar 19, 2024, 9:03 PM
1 point
0 comments4 min readLW link

In­creas­ing IQ by 10 Points is Possible

George3d6Mar 19, 2024, 8:48 PM
23 points
51 comments5 min readLW link
(morelucid.substack.com)

Are ex­treme prob­a­bil­ities for P(doom) epistem­i­cally jus­tifed?

Mar 19, 2024, 8:32 PM
20 points
12 comments7 min readLW link

Have I Solved the Two En­velopes Prob­lem Once and For All?

JackOfAllTradesMar 19, 2024, 7:57 PM
−6 points
5 comments3 min readLW link

[Question] How can one be less wrong, if their con­ver­sa­tion part­ner loses the in­ter­est on dis­cussing the topic with them?

OokerMar 19, 2024, 6:11 PM
−10 points
3 comments1 min readLW link

Carlo: un­cer­tainty anal­y­sis in Google Sheets

ProbabilityEnjoyerMar 19, 2024, 5:59 PM
6 points
0 comments1 min readLW link
(carlo.app)

NAIRA—An ex­er­cise in reg­u­la­tory, com­pet­i­tive safety gov­er­nance [AI Gover­nance In­sti­tu­tional De­sign idea]

HerambMar 19, 2024, 5:43 PM
2 points
0 comments1 min readLW link
(forum.effectivealtruism.org)

AI Safety Eval­u­a­tions: A Reg­u­la­tory Review

Mar 19, 2024, 3:05 PM
22 points
1 comment11 min readLW link

Mechanism for fea­ture learn­ing in neu­ral net­works and back­prop­a­ga­tion-free ma­chine learn­ing models

Matt GoldenbergMar 19, 2024, 2:55 PM
8 points
1 comment1 min readLW link
(www.science.org)

Monthly Roundup #16: March 2024

ZviMar 19, 2024, 1:10 PM
33 points
4 comments55 min readLW link
(thezvi.wordpress.com)

Ex­per­i­men­ta­tion (Part 7 of “The Sense Of Phys­i­cal Ne­ces­sity”)

LoganStrohlMar 18, 2024, 9:25 PM
33 points
0 comments10 min readLW link

INTERVIEW: Round 2 - StakeOut.AI w/​ Dr. Peter Park

jacobhaimesMar 18, 2024, 9:21 PM
5 points
0 comments1 min readLW link
(into-ai-safety.github.io)

Neu­ro­science and Alignment

Garrett BakerMar 18, 2024, 9:09 PM
40 points
25 comments2 min readLW link

GPT, the mag­i­cal col­lab­o­ra­tion zone, Lex Frid­man and Sam Altman

Bill BenzonMar 18, 2024, 8:04 PM
3 points
1 comment3 min readLW link

Mea­sur­ing Co­her­ence of Poli­cies in Toy Environments

Mar 18, 2024, 5:59 PM
59 points
9 comments14 min readLW link

AtP*: An effi­cient and scal­able method for lo­cal­iz­ing LLM be­havi­our to components

Mar 18, 2024, 5:28 PM
19 points
0 comments1 min readLW link
(arxiv.org)

Com­mu­nity Notes by X

NicholasKeesMar 18, 2024, 5:13 PM
127 points
15 comments7 min readLW link

[Question] Is the Basilisk pre­tend­ing to be hid­den in this simu­la­tion so that it can check what I would do if con­di­tioned by a world with­out the Basilisk?

maybefbiMar 18, 2024, 4:05 PM
−18 points
1 comment1 min readLW link

On Devin

ZviMar 18, 2024, 1:20 PM
148 points
34 comments11 min readLW link
(thezvi.wordpress.com)

RLLMv10 experiment

MiguelDevMar 18, 2024, 8:32 AM
5 points
0 comments2 min readLW link

Join the AI Eval­u­a­tion Tasks Bounty Hackathon

Esben KranMar 18, 2024, 8:15 AM
12 points
1 commentLW link

5 Physics Problems

Mar 18, 2024, 8:05 AM
60 points
0 comments15 min readLW link

In­fer­ring the model di­men­sion of API-pro­tected LLMs

Ege ErdilMar 18, 2024, 6:19 AM
34 points
3 comments4 min readLW link
(arxiv.org)