All 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021 2022 202320242025

All Jan Feb Mar Apr May Jun Jul Aug Sep OctNovDec

All 1 2 345 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30

Drug development costs can range over two orders of magnitude

rossryNov 3, 2024, 11:13 PM

38 points

6 votes

Overall karma indicates overall quality.

0 comments11 min readLW link

Redefining Tolerance: Beyond Popper’s Paradox

mindprisonNov 3, 2024, 10:23 PM

−1 points

7 votes

Overall karma indicates overall quality.

0 comments3 min readLW link

Goal: Understand Intelligence

Johannes C. MayerNov 3, 2024, 9:20 PM

14 points

14 votes

Overall karma indicates overall quality.

19 comments1 min readLW link

Current safety training techniques do not fully transfer to the agent setting

Simon Lermen and Govind Pimpale

Nov 3, 2024, 7:24 PM

158 points

62 votes

Overall karma indicates overall quality.

9 comments5 min readLW link

Why our politicians aren’t Median

Yair HalberstadtNov 3, 2024, 2:03 PM

72 points

35 votes

Overall karma indicates overall quality.

15 comments3 min readLW link

Human Biodiversity (Part 4: Astral Codex Ten)

Evan_GaensbauerNov 3, 2024, 4:20 AM

−13 points

18 votes

Overall karma indicates overall quality.

6 comments1 min readLW link

(reflectivealtruism.com)

Understanding incomparability versus incommensurability in relation to RLHF

artemiocobbNov 2, 2024, 10:57 PM

1 point

3 votes

Overall karma indicates overall quality.

1 comment2 min readLW link

electric turbofans

bhauthNov 2, 2024, 10:50 PM

63 points

34 votes

Overall karma indicates overall quality.

2 comments5 min readLW link

(bhauth.com)

Reality as Category-Theoretic State Machines: A Mathematical Framework

Wenitte ApiouNov 2, 2024, 9:04 PM

−8 points

7 votes

Overall karma indicates overall quality.

0 comments2 min readLW link

The Median Researcher Problem

johnswentworthNov 2, 2024, 8:16 PM

157 points

130 votes

Overall karma indicates overall quality.

70 comments1 min readLW link

Testing “True” Language Understanding in LLMs: A Simple Proposal

MtryaSamNov 2, 2024, 7:12 PM

9 points

5 votes

Overall karma indicates overall quality.

2 comments2 min readLW link

Testing “True” Language Understanding in LLMs: A Simple Proposal

MtryaSamNov 2, 2024, 7:12 PM

−3 points

3 votes

Overall karma indicates overall quality.

0 comments2 min readLW link

Fragile, Robust, and Antifragile Preference Satisfaction

adamShimiNov 2, 2024, 5:25 PM

19 points

9 votes

Overall karma indicates overall quality.

0 comments5 min readLW link

(formethods.substack.com)

Higher Order Signs, Hallucination and Schizophrenia

Nicolas VillarrealNov 2, 2024, 4:33 PM

4 points

9 votes

Overall karma indicates overall quality.

0 comments13 min readLW link

(nicolasdvillarreal.substack.com)

[Question] Is OpenAI net negative for AI Safety?

Lysandre TerrisseNov 2, 2024, 4:18 PM

4 points

4 votes

Overall karma indicates overall quality.

0 comments1 min readLW link

Two arguments against longtermist thought experiments

momom2Nov 2, 2024, 10:22 AM

15 points

9 votes

Overall karma indicates overall quality.

5 comments3 min readLW link

Both-Sidesism—When Fair & Balanced Goes Wrong

James Stephen BrownNov 2, 2024, 3:04 AM

3 points

30 votes

Overall karma indicates overall quality.

15 comments6 min readLW link

(nonzerosum.games)

What can we learn from insecure domains?

Logan ZoellnerNov 1, 2024, 11:53 PM

14 points

15 votes

Overall karma indicates overall quality.

21 comments1 min readLW link

Science advances one funeral at a time

Cameron Berg, Judd Rosenblatt, Diogo de Lucena and Trent Hodgeson

Nov 1, 2024, 11:06 PM

100 points

46 votes

Overall karma indicates overall quality.

9 comments2 min readLW link

The Cartesian Crisis

mindprisonNov 1, 2024, 11:02 PM

−5 points

4 votes

Overall karma indicates overall quality.

2 comments2 min readLW link

Hypothesis on Composition Circuits in Vision Transformers

phenomanonNov 1, 2024, 10:16 PM

2 points

2 votes

Overall karma indicates overall quality.

0 comments3 min readLW link

SAE Probing: What is it good for?

Subhash Kantamneni, Josh Engels, Senthooran Rajamanoharan and Neel Nanda

Nov 1, 2024, 7:23 PM

34 points

13 votes

Overall karma indicates overall quality.

0 comments11 min readLW link

[Question] Set Theory Multiverse vs Mathematical Truth—Philosophical Discussion

Wenitte ApiouNov 1, 2024, 6:56 PM

8 points

5 votes

Overall karma indicates overall quality.

25 comments1 min readLW link

Educational CAI: Aligning a Language Model with Pedagogical Theories

Bharath PuranamNov 1, 2024, 6:55 PM

5 points

3 votes

Overall karma indicates overall quality.

1 comment13 min readLW link

Prediction markets and Taxes

Edmund NelsonNov 1, 2024, 5:39 PM

11 points

7 votes

Overall karma indicates overall quality.

8 comments1 min readLW link

Dentistry, Oral Surgeons, and the Inefficiency of Small Markets

GeneSmithNov 1, 2024, 5:26 PM

86 points

58 votes

Overall karma indicates overall quality.

18 comments5 min readLW link

Live Machinery: An Interface Design Philosophy for Wholesome AI Futures

SahilNov 1, 2024, 5:24 PM

48 points

25 votes

Overall karma indicates overall quality.

3 comments35 min readLW link

Seeking Collaborators

abramdemskiNov 1, 2024, 5:13 PM

62 points

23 votes

Overall karma indicates overall quality.

15 comments7 min readLW link

Complete Feedback

abramdemskiNov 1, 2024, 4:58 PM

25 points

10 votes

Overall karma indicates overall quality.

8 comments3 min readLW link

Levers for Biological Progress—A Response to “Machines of Loving Grace”

Niko_McCartyNov 1, 2024, 4:35 PM

17 points

5 votes

Overall karma indicates overall quality.

0 comments20 min readLW link

(www.asimov.press)

2024 Unofficial LW Community Census, Request for Comments

ScrewtapeNov 1, 2024, 4:34 PM

23 points

12 votes

Overall karma indicates overall quality.

32 comments3 min readLW link

[Question] When engaging with a large amount of resources during a literature review, how do you prevent yourself from becoming overwhelmed?

corruptedCatapillarNov 1, 2024, 7:29 AM

25 points

10 votes

Overall karma indicates overall quality.

2 comments3 min readLW link

(draft) Cyborg software should be open (?)

AtillaYasarNov 1, 2024, 7:24 AM

4 points

4 votes

Overall karma indicates overall quality.

5 comments3 min readLW link

Another UFO Bet

codyzNov 1, 2024, 1:55 AM

9 points

9 votes

Overall karma indicates overall quality.

11 comments1 min readLW link

Trading Candy

jefftkNov 1, 2024, 1:10 AM

28 points

13 votes

Overall karma indicates overall quality.

4 comments1 min readLW link

(www.jefftk.com)

JargonBot Beta Test

RaemonNov 1, 2024, 1:05 AM

84 points

38 votes

Overall karma indicates overall quality.

55 comments6 min readLW link

GPT-4o Guardrails Gone: Data Poisoning & Jailbreak-Tuning

ChengCheng, Brendan Murphy, AdamGleave and Kellin Pelrine

Nov 1, 2024, 12:10 AM

18 points

8 votes

Overall karma indicates overall quality.

0 comments6 min readLW link

(far.ai)

The slingshot helps with learning

Wilson WuOct 31, 2024, 11:18 PM

33 points

11 votes

Overall karma indicates overall quality.

0 comments8 min readLW link

Toward Safety Case Inspired Basic Research

Lucas Teixeira, Lauren Greenspan, Dmitry Vaintrob and Eric Winsor

Oct 31, 2024, 11:06 PM

55 points

16 votes

Overall karma indicates overall quality.

3 comments13 min readLW link

Spooky Recommendation System Scaling

phdeadOct 31, 2024, 10:00 PM

11 points

4 votes

Overall karma indicates overall quality.

0 comments4 min readLW link

‘Meta’, ‘mesa’, and mountains

LorecOct 31, 2024, 5:25 PM

1 point

3 votes

Overall karma indicates overall quality.

0 comments3 min readLW link

Toward Safety Cases For AI Scheming

Mikita Balesni and Marius Hobbhahn

Oct 31, 2024, 5:20 PM

60 points

23 votes

Overall karma indicates overall quality.

1 comment2 min readLW link

AI #88: Thanks for the Memos

ZviOct 31, 2024, 3:00 PM

46 points

17 votes

Overall karma indicates overall quality.

5 comments77 min readLW link

(thezvi.wordpress.com)

The Compendium, A full argument about extinction risk from AGI

adamShimi, Gabriel Alfour, Connor Leahy, Chris Scammell and Andrea_Miotti

Oct 31, 2024, 12:01 PM

196 points

87 votes

Overall karma indicates overall quality.

52 comments2 min readLW link

(www.thecompendium.ai)

Some Preliminary Notes on the Promise of a Wisdom Explosion

Chris_LeongOct 31, 2024, 9:21 AM

2 points

2 votes

Overall karma indicates overall quality.

0 comments1 min readLW link

(aiimpacts.org)

What TMS is like

SableOct 31, 2024, 12:44 AM

218 points

117 votes

Overall karma indicates overall quality.

23 comments6 min readLW link

(affablyevil.substack.com)

AI Safety at the Frontier: Paper Highlights, October ’24

gasteigerjoOct 31, 2024, 12:09 AM

3 points

1 vote

Overall karma indicates overall quality.

0 comments9 min readLW link

(aisafetyfrontier.substack.com)

Standard SAEs Might Be Incoherent: A Choosing Problem & A “Concise” Solution

Kola AyonrindeOct 30, 2024, 10:50 PM

27 points

11 votes

Overall karma indicates overall quality.

0 comments12 min readLW link

Generic advice caveats

Saul MunnOct 30, 2024, 9:03 PM

27 points

13 votes

Overall karma indicates overall quality.

1 comment3 min readLW link

(www.brasstacks.blog)

I turned decision theory problems into memes about trolleys

TapataktOct 30, 2024, 8:13 PM

105 points

54 votes

Overall karma indicates overall quality.

23 comments1 min readLW link

Keyboard shortcuts

Keys shown in yellow (e.g., ]) are accesskeys, and require a browser-specific modifier key (or keys).

Keys shown in grey (e.g., ?) do not require any modifier keys.

General
? Show keyboard shortcuts
Esc Hide keyboard shortcuts

Site navigation
h Go to Home (a.k.a. “Frontpage”) view
f Go to Featured (a.k.a. “Curated”) view
a Go to All (a.k.a. “Community”) view
m Go to Meta view
v Go to Tags view
c Go to Recent Comments view
r Go to Archive view
q Go to Sequences view
t Go to About page
u Go to User or Login page
o Go to Inbox page

Page navigation
, Jump up to top of page
. Jump down to bottom of page
/ Jump to top of comments section
s Search

Page actions
n New post or comment
e Edit current post

Post/comment list views
. Focus next entry in list
, Focus previous entry in list
; Cycle between links in focused entry
Enter Go to currently focused entry
Esc Unfocus currently focused entry
] Go to next page
[ Go to previous page
\ Go to first page
e Edit currently focused post

Editor
k Bold text
i Italic text
l Insert hyperlink
q Blockquote text

Appearance
= Increase text size
- Decrease text size
0 Reset to default text size
′ Cycle through content width settings
1 Switch to default theme [A]
2 Switch to dark theme [B]
3 Switch to grey theme [C]
4 Switch to ultramodern theme [D]
5 Switch to simple theme [E]
6 Switch to brutalist theme [F]
7 Switch to ReadTheSequences theme [G]
8 Switch to classic Less Wrong theme [H]
9 Switch to modern Less Wrong theme [I]
; Open theme tweaker
Enter Save changes and close theme tweaker
Esc Close theme tweaker (without saving)

Slide shows
l Start/resume slideshow
Esc Exit slideshow
→↓ Next slide
←↑ Previous slide
Space Reset slide zoom

Miscellaneous
x Switch to next view on user page
z Switch to previous view on user page
` Toggle compact comment list view
g Toggle anti-kibitzer