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 AugSepOct Nov Dec

All 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 192021 22 23 24 25 26 27 28 29 30

Piling bounded arguments

momom2Sep 19, 2024, 10:27 PM

7 points

0 comments4 min readLW link

We Don’t Know Our Own Values, but Reward Bridges The Is-Ought Gap

johnswentworth and David Lorell

Sep 19, 2024, 10:22 PM

49 points

48 comments5 min readLW link

Interested in Cognitive Bootcamp?

RaemonSep 19, 2024, 10:12 PM

48 points

0 comments2 min readLW link

Just How Good Are Modern Chess Computers?

nemSep 19, 2024, 6:57 PM

10 points

1 comment6 min readLW link

RLHF is the worst possible thing done when facing the alignment problem

tailcalledSep 19, 2024, 6:56 PM

32 points

10 comments6 min readLW link

AISafety.info: What are Inductive Biases?

AlgonSep 19, 2024, 5:26 PM

11 points

4 comments2 min readLW link

(aisafety.info)

Physics of Language models (part 2.1)

Nathan Helm-BurgerSep 19, 2024, 4:48 PM

9 points

2 comments1 min readLW link

(youtu.be)

Why good things often don’t lead to better outcomes

DMMFSep 19, 2024, 4:37 PM

16 points

1 comment4 min readLW link

(danfrank.ca)

To CoT or not to CoT? Chain-of-thought helps mainly on math and symbolic reasoning

Bogdan Ionut CirsteaSep 19, 2024, 4:13 PM

21 points

1 comment1 min readLW link

(arxiv.org)

Laziness death spirals

PatrickDFarleySep 19, 2024, 3:58 PM

276 points

40 comments8 min readLW link

[Intuitive self-models] 1. Preliminaries

Steven ByrnesSep 19, 2024, 1:45 PM

91 points

23 comments15 min readLW link

AI #82: The Governor Ponders

ZviSep 19, 2024, 1:30 PM

50 points

8 comments27 min readLW link

(thezvi.wordpress.com)

Slave Morality: A place for every man and every man in his place

Martin SustrikSep 19, 2024, 4:20 AM

16 points

7 comments2 min readLW link

(250bpm.substack.com)

Which LessWrong/Alignment topics would you like to be tutored in? [Poll]

RubySep 19, 2024, 1:35 AM

43 points

12 comments1 min readLW link

The Obliqueness Thesis

jessicataSep 19, 2024, 12:26 AM

95 points

19 comments17 min readLW link

How to choose what to work on

jasoncrawfordSep 18, 2024, 8:39 PM

22 points

6 comments4 min readLW link

(blog.rootsofprogress.org)

Intention-to-Treat (Re: How harmful is music, really?)

kqrSep 18, 2024, 6:44 PM

11 points

0 comments5 min readLW link

(entropicthoughts.com)

The case for a negative alignment tax

Cameron Berg, Judd Rosenblatt, Diogo de Lucena and AE Studio

Sep 18, 2024, 6:33 PM

75 points

20 comments7 min readLW link

Endogenous Growth and Human Intelligence

Nicholas D.Sep 18, 2024, 2:05 PM

3 points

0 comments2 min readLW link

Inquisitive vs. adversarial rationality

gbSep 18, 2024, 1:50 PM

6 points

9 comments2 min readLW link

Pronouns are Annoying

ymeskhoutSep 18, 2024, 1:30 PM

15 points

23 comments4 min readLW link

(www.ymeskhout.com)

Is “superhuman” AI forecasting BS? Some experiments on the “539″ bot from the Centre for AI Safety

titotalSep 18, 2024, 1:07 PM

79 points

3 comments LW link

(open.substack.com)

Knowledge’s practicability

Ted NguyễnSep 18, 2024, 2:31 AM

−5 points

0 comments7 min readLW link

(tednguyen.substack.com)

Skills from a year of Purposeful Rationality Practice

RaemonSep 18, 2024, 2:05 AM

190 points

18 comments7 min readLW link

[Question] Where to find reliable reviews of AI products?

ElizabethSep 17, 2024, 11:48 PM

29 points

6 comments1 min readLW link

Superposition through Active Learning Lens

akankshancSep 17, 2024, 5:32 PM

1 point

0 comments10 min readLW link

Survey—Psychological Impact of Long-Term AI Engagement

Manuela GarcíaSep 17, 2024, 5:31 PM

2 points

0 comments1 min readLW link

Survey—Psychological Impact of Long-Term AI Engagement

Manuela GarcíaSep 17, 2024, 5:31 PM

1 point

1 comment1 min readLW link

[Question] What does it mean for an event or observation to have probability 0 or 1 in Bayesian terms?

Noosphere89Sep 17, 2024, 5:28 PM

1 point

22 comments1 min readLW link

How harmful is music, really?

dkl9Sep 17, 2024, 2:53 PM

10 points

6 comments3 min readLW link

(dkl9.net)

Monthly Roundup #22: September 2024

ZviSep 17, 2024, 12:20 PM

35 points

10 comments45 min readLW link

(thezvi.wordpress.com)

I finally got ChatGPT to sound like me

lsusrSep 17, 2024, 9:39 AM

47 points

18 comments6 min readLW link

Food, Prison & Exotic Animals: Sparse Autoencoders Detect 6.5x Performing Youtube Thumbnails

Louka Ewington-PitsosSep 17, 2024, 3:52 AM

6 points

2 comments7 min readLW link

Head in the Cloud: Why an Upload of Your Mind is Not You

xhqSep 17, 2024, 12:25 AM

−11 points

3 comments14 min readLW link

[Question] How does someone prove that their general intelligence is above average?

M. Y. ZuoSep 16, 2024, 9:01 PM

−3 points

12 comments1 min readLW link

[Question] Does life actually locally increase entropy?

tailcalledSep 16, 2024, 8:30 PM

10 points

27 comments1 min readLW link

Book review: Xenosystems

jessicataSep 16, 2024, 8:17 PM

50 points

18 comments37 min readLW link

(unstableontology.com)

MIRI’s September 2024 newsletter

HarlanSep 16, 2024, 6:15 PM

46 points

0 comments1 min readLW link

(intelligence.org)

Generative ML in chemistry is bottlenecked by synthesis

Abhishaike MahajanSep 16, 2024, 4:31 PM

38 points

2 comments14 min readLW link

(www.owlposting.com)

Secret Collusion: Will We Know When to Unplug AI?

schroederdewitt, srm, MikhailB, Lewis Hammond, chansmi and sofmonk

Sep 16, 2024, 4:07 PM

61 points

8 comments31 min readLW link

GPT-o1

ZviSep 16, 2024, 1:40 PM

86 points

34 comments46 min readLW link

(thezvi.wordpress.com)

[Question] Can subjunctive dependence emerge from a simplicity prior?

Daniel CSep 16, 2024, 12:39 PM

11 points

0 comments1 min readLW link

Longevity and the Mind

George3d6Sep 16, 2024, 9:43 AM

5 points

2 comments10 min readLW link

[Question] What’s the Deal with Logical Uncertainty?

Ape in the coatSep 16, 2024, 8:11 AM

32 points

29 comments2 min readLW link

Reinforcement Learning from Information Bazaar Feedback, and other uses of information markets

Abhimanyu Pallavi SudhirSep 16, 2024, 1:04 AM

5 points

1 comment5 min readLW link

Hyperpolation

Gunnar_ZarnckeSep 15, 2024, 9:37 PM

22 points

6 comments1 min readLW link

(arxiv.org)

[Question] If I wanted to spend WAY more on AI, what would I spend it on?

Logan ZoellnerSep 15, 2024, 9:24 PM

53 points

16 comments1 min readLW link

Superintelligence Can’t Solve the Problem of Deciding What You’ll Do

Vladimir_NesovSep 15, 2024, 9:03 PM

27 points

11 comments1 min readLW link

For Limited Superintelligences, Epistemic Exclusion is Harder than Robustness to Logical Exploitation

LorecSep 15, 2024, 8:49 PM

3 points

9 comments3 min readLW link

Why I funded PIBBSS

Ryan KiddSep 15, 2024, 7:56 PM

115 points

21 comments3 min readLW link