All 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021 2022 2023 20242025

All Jan Feb Mar Apr May Jun Jul Aug SepOct

All1 2 3 4 5 6 7 8

Omelas Is Perfectly Misread

Tobias H2 Oct 2025 23:11 UTC

197 points

49 comments5 min readLW link

Nice-ish, smooth takeoff (with imperfect safeguards) probably kills most “classic humans” in a few decades.

Raemon2 Oct 2025 21:03 UTC

143 points

19 comments12 min readLW link

The Origami Men

Tomás B.6 Oct 2025 15:25 UTC

136 points

8 comments16 min readLW link

Checking in on AI-2027

Baybar2 Oct 2025 18:46 UTC

119 points

21 comments4 min readLW link

Do One New Thing A Day To Solve Your Problems

Algon3 Oct 2025 17:08 UTC

102 points

5 comments2 min readLW link

Where does Sonnet 4.5′s desire to “not get too comfortable” come from?

Kaj_Sotala4 Oct 2025 10:19 UTC

91 points

16 comments64 min readLW link

Gradual Disempowerment Monthly Roundup

Raymond Douglas6 Oct 2025 15:36 UTC

90 points

7 comments6 min readLW link

Making Your Pain Worse can Get You What You Want

Logan Riggs5 Oct 2025 0:19 UTC

76 points

3 comments3 min readLW link

Bending The Curve

Zvi7 Oct 2025 20:00 UTC

72 points

6 comments21 min readLW link

(thezvi.wordpress.com)

Eliciting secret knowledge from language models

Bartosz Cywiński, Arthur Conmy and Sam Marks

2 Oct 2025 20:57 UTC

67 points

3 comments2 min readLW link

(arxiv.org)

The Counterfactual Quiet AGI Timeline

Davidmanheim5 Oct 2025 9:09 UTC

64 points

5 comments9 min readLW link

Maybe social media algorithms don’t suck

Algon5 Oct 2025 18:47 UTC

64 points

18 comments3 min readLW link

“Intelligence” → “Relentless, Creative Resourcefulness”

Raemon7 Oct 2025 0:28 UTC

63 points

28 comments17 min readLW link

“Pessimization” is Just Ordinary Failure

J Bostock1 Oct 2025 13:48 UTC

56 points

2 comments6 min readLW link

Recent AI Experiences

abramdemski3 Oct 2025 19:32 UTC

54 points

1 comment6 min readLW link

Petri: An open-source auditing tool to accelerate AI safety research

Sam Marks7 Oct 2025 20:39 UTC

54 points

0 comments1 min readLW link

(alignment.anthropic.com)

</rant> </uncharitable> </psychologizing>

Raemon1 Oct 2025 21:20 UTC

53 points

11 comments2 min readLW link

We’ve automated x-risk-pilling people

Mikhail Samin3 Oct 2025 10:26 UTC

51 points

27 comments1 min readLW link

(whycare.aisgf.us)

How to Feel More Alive

Logan Riggs2 Oct 2025 15:45 UTC

47 points

2 comments4 min readLW link

LLMs one-box when in a “hostile telepath” version of Newcomb’s Paradox, except for the one that beat the predictor

Kaj_Sotala6 Oct 2025 8:44 UTC

47 points

6 comments17 min readLW link

IABIED and Memetic Engineering

Error3 Oct 2025 1:01 UTC

47 points

5 comments4 min readLW link

No, That’s Not What the Flight Costs

Max Niederman2 Oct 2025 17:55 UTC

45 points

15 comments1 min readLW link

(maxniederman.com)

Lectures on statistical learning theory for alignment researchers

Vanessa Kosoy1 Oct 2025 8:36 UTC

41 points

1 comment1 min readLW link

(www.youtube.com)

Claude Sonnet 4.5 Is A Very Good Model

Zvi1 Oct 2025 18:00 UTC

40 points

2 comments24 min readLW link

(thezvi.wordpress.com)

Sora and The Big Bright Screen Slop Machine

Zvi3 Oct 2025 11:40 UTC

38 points

1 comment35 min readLW link

(thezvi.wordpress.com)

You Should Get a Reusable Mask

jefftk8 Oct 2025 2:40 UTC

37 points

3 comments1 min readLW link

(www.jefftk.com)

Some Biology Related Things I Found Interesting

Morpheus2 Oct 2025 12:18 UTC

37 points

9 comments2 min readLW link

Replacing RL w/ Parameter-based Evolutionary Strategies

Logan Riggs8 Oct 2025 1:02 UTC

37 points

3 comments3 min readLW link

Do Things for as Many Reasons as Possible

Philipreal6 Oct 2025 0:28 UTC

35 points

1 comment2 min readLW link

Antisocial media: AI’s killer app?

David Scott Krueger (formerly: capybaralet)3 Oct 2025 0:00 UTC

35 points

8 comments5 min readLW link

(therealartificialintelligence.substack.com)

[Question] Generalization and the Multiple Stage Fallacy?

Zack_M_Davis7 Oct 2025 6:20 UTC

34 points

6 comments3 min readLW link

AI #136: A Song and Dance

Zvi2 Oct 2025 13:10 UTC

33 points

3 comments47 min readLW link

(thezvi.wordpress.com)

AI and Cheap Weapons

Felix C.1 Oct 2025 17:31 UTC

31 points

3 comments23 min readLW link

Base64Bench: How good are LLMs at base64, and why care about it?

richbc5 Oct 2025 18:07 UTC

31 points

6 comments11 min readLW link

We won’t get AIs smart enough to solve alignment but too dumb to rebel

Joe Rogero6 Oct 2025 21:49 UTC

28 points

16 comments5 min readLW link

LLMs are badly misaligned

Joe Rogero5 Oct 2025 14:00 UTC

27 points

25 comments3 min readLW link

Making Sense of Consciousness Part 6: Perceptions of Disembodiment

sarahconstantin3 Oct 2025 20:40 UTC

27 points

0 comments8 min readLW link

(sarahconstantin.substack.com)

How the NanoGPT Speedrun WR dropped by 20% in 3 months

larry-dial5 Oct 2025 1:05 UTC

26 points

9 comments9 min readLW link

Medical Roundup #5

Zvi6 Oct 2025 15:10 UTC

26 points

2 comments26 min readLW link

(thezvi.wordpress.com)

Excerpts from my neuroscience to-do list

Steven Byrnes6 Oct 2025 21:05 UTC

26 points

1 comment4 min readLW link

Telling the Difference Between Memories & Logical Guesses

Logan Riggs7 Oct 2025 5:46 UTC

25 points

3 comments4 min readLW link

What I’ve Learnt About How to Sleep

Algon4 Oct 2025 20:52 UTC

25 points

7 comments2 min readLW link

But what kind of stuff can you just do?

Bastiaan1 Oct 2025 16:58 UTC

25 points

5 comments1 min readLW link

Anthropic’s JumpReLU training method is really good

chanind and Adrià Garriga-alonso

3 Oct 2025 15:23 UTC

22 points

0 comments2 min readLW link

Goodness is harder to achieve than competence

Joe Rogero3 Oct 2025 21:32 UTC

22 points

0 comments3 min readLW link

Good is a smaller target than smart

Joe Rogero3 Oct 2025 21:04 UTC

21 points

0 comments2 min readLW link

Intent alignment seems incoherent

Joe Rogero7 Oct 2025 23:01 UTC

20 points

1 comment6 min readLW link

The quotation mark

Maxwell Peterson5 Oct 2025 23:23 UTC

19 points

8 comments13 min readLW link

My Brush with Superhuman Persuasion

Ben S.1 Oct 2025 17:50 UTC

18 points

13 comments9 min readLW link

(thebsdetector.substack.com)

Research Robots: When AIs Experiment on Us

Shoshannah Tekofsky7 Oct 2025 12:10 UTC

18 points

0 comments7 min readLW link

(theaidigest.org)