All 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021 2022 20232024

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

momom219 Sep 2024 22:27 UTC

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

19 Sep 2024 22:22 UTC

47 points

47 comments5 min readLW link

Interested in Cognitive Bootcamp?

Raemon19 Sep 2024 22:12 UTC

48 points

0 comments2 min readLW link

Just How Good Are Modern Chess Computers?

nem19 Sep 2024 18:57 UTC

10 points

1 comment6 min readLW link

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

tailcalled19 Sep 2024 18:56 UTC

32 points

10 comments6 min readLW link

AISafety.info: What are Inductive Biases?

Algon19 Sep 2024 17:26 UTC

11 points

4 comments2 min readLW link

(aisafety.info)

Physics of Language models (part 2.1)

Nathan Helm-Burger19 Sep 2024 16:48 UTC

9 points

2 comments1 min readLW link

(youtu.be)

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

DMMF19 Sep 2024 16:37 UTC

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 Cirstea19 Sep 2024 16:13 UTC

21 points

1 comment1 min readLW link

(arxiv.org)

Laziness death spirals

PatrickDFarley19 Sep 2024 15:58 UTC

247 points

35 comments8 min readLW link

[Intuitive self-models] 1. Preliminaries

Steven Byrnes19 Sep 2024 13:45 UTC

88 points

20 comments15 min readLW link

AI #82: The Governor Ponders

Zvi19 Sep 2024 13:30 UTC

50 points

8 comments27 min readLW link

(thezvi.wordpress.com)

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

Martin Sustrik19 Sep 2024 4:20 UTC

16 points

7 comments2 min readLW link

(250bpm.substack.com)

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

Ruby19 Sep 2024 1:35 UTC

43 points

12 comments1 min readLW link

The Obliqueness Thesis

jessicata19 Sep 2024 0:26 UTC

77 points

17 comments17 min readLW link

How to choose what to work on

jasoncrawford18 Sep 2024 20:39 UTC

22 points

6 comments4 min readLW link

(blog.rootsofprogress.org)

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

kqr18 Sep 2024 18:44 UTC

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

18 Sep 2024 18:33 UTC

74 points

20 comments7 min readLW link

Endogenous Growth and Human Intelligence

Nicholas D.18 Sep 2024 14:05 UTC

3 points

0 comments2 min readLW link

Inquisitive vs. adversarial rationality

gb18 Sep 2024 13:50 UTC

6 points

9 comments2 min readLW link

Pronouns are Annoying

ymeskhout18 Sep 2024 13:30 UTC

15 points

21 comments4 min readLW link

(www.ymeskhout.com)

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

titotal18 Sep 2024 13:07 UTC

78 points

3 comments1 min readLW link

(open.substack.com)

Knowledge’s practicability

Ted Nguyễn18 Sep 2024 2:31 UTC

−5 points

0 comments7 min readLW link

(tednguyen.substack.com)

Skills from a year of Purposeful Rationality Practice

Raemon18 Sep 2024 2:05 UTC

185 points

18 comments7 min readLW link

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

Elizabeth17 Sep 2024 23:48 UTC

29 points

6 comments1 min readLW link

Superposition through Active Learning Lens

akankshanc17 Sep 2024 17:32 UTC

1 point

0 comments10 min readLW link

Survey—Psychological Impact of Long-Term AI Engagement

Manuela García17 Sep 2024 17:31 UTC

2 points

0 comments1 min readLW link

Survey—Psychological Impact of Long-Term AI Engagement

Manuela García17 Sep 2024 17:31 UTC

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?

Noosphere8917 Sep 2024 17:28 UTC

1 point

22 comments1 min readLW link

How harmful is music, really?

dkl917 Sep 2024 14:53 UTC

10 points

6 comments3 min readLW link

(dkl9.net)

Monthly Roundup #22: September 2024

Zvi17 Sep 2024 12:20 UTC

35 points

10 comments45 min readLW link

(thezvi.wordpress.com)

I finally got ChatGPT to sound like me

lsusr17 Sep 2024 9:39 UTC

46 points

18 comments6 min readLW link

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

Louka Ewington-Pitsos17 Sep 2024 3:52 UTC

6 points

2 comments7 min readLW link

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

xhq17 Sep 2024 0:25 UTC

−11 points

3 comments14 min readLW link

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

M. Y. Zuo16 Sep 2024 21:01 UTC

−3 points

12 comments1 min readLW link

[Question] Does life actually locally increase entropy?

tailcalled16 Sep 2024 20:30 UTC

10 points

27 comments1 min readLW link

Book review: Xenosystems

jessicata16 Sep 2024 20:17 UTC

49 points

18 comments37 min readLW link

(unstableontology.com)

MIRI’s September 2024 newsletter

Harlan16 Sep 2024 18:15 UTC

46 points

0 comments1 min readLW link

(intelligence.org)

Generative ML in chemistry is bottlenecked by synthesis

Abhishaike Mahajan16 Sep 2024 16:31 UTC

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

16 Sep 2024 16:07 UTC

55 points

7 comments31 min readLW link

GPT-o1

Zvi16 Sep 2024 13:40 UTC

86 points

34 comments46 min readLW link

(thezvi.wordpress.com)

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

Daniel C16 Sep 2024 12:39 UTC

6 points

0 comments1 min readLW link

Longevity and the Mind

George3d616 Sep 2024 9:43 UTC

5 points

2 comments10 min readLW link

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

Ape in the coat16 Sep 2024 8:11 UTC

32 points

28 comments2 min readLW link

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

Abhimanyu Pallavi Sudhir16 Sep 2024 1:04 UTC

5 points

1 comment5 min readLW link

Hyperpolation

Gunnar_Zarncke15 Sep 2024 21:37 UTC

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 Zoellner15 Sep 2024 21:24 UTC

52 points

16 comments1 min readLW link

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

Vladimir_Nesov15 Sep 2024 21:03 UTC

27 points

11 comments1 min readLW link

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

Lorec15 Sep 2024 20:49 UTC

3 points

9 comments3 min readLW link

Why I funded PIBBSS

Ryan Kidd15 Sep 2024 19:56 UTC

115 points

21 comments3 min readLW link