All 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021 2022 2023 202420252026

All Jan Feb Mar AprMayJun Jul Aug Sep Oct Nov Dec

All 1 2 3 4 5 6 7 8 9 10 111213 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31

Absolute Zero: Alpha Zero for LLM

alapmi11 May 2025 20:42 UTC

23 points

16 comments1 min readLW link

AGI will result from an ecosystem not a single firm

hamish_low11 May 2025 20:06 UTC

6 points

1 comment6 min readLW link

(cambrianr.substack.com)

Thou shalt not command an alighned AI

Martin Vlach11 May 2025 20:02 UTC

0 points

4 comments1 min readLW link

[Question] How do I design long prompts for thinking zero shot systems with distinct equally distributed prompt sections (mission, goals, memories, how-to-respond,… etc) and how to maintain llm coherence?

ollie_11 May 2025 19:32 UTC

2 points

5 comments1 min readLW link

a confusion about preference orderings

nostalgebraist11 May 2025 19:30 UTC

93 points

39 comments11 min readLW link

[Book Translation] Three Days in Dwarfland

Viliam11 May 2025 17:54 UTC

27 points

6 comments1 min readLW link

Better Air Purifiers

jefftk11 May 2025 16:50 UTC

81 points

21 comments3 min readLW link

(www.jefftk.com)

Aligning Agents, Tools, and Simulators

WillPetillo, Sean Herrington, Spencer Ames, Adebayo Mubarak and Can Narin

11 May 2025 7:59 UTC

24 points

2 comments6 min readLW link

Consider not donating under $100 to political candidates

DanielFilan11 May 2025 3:20 UTC

141 points

33 comments1 min readLW link

(danielfilan.com)

Somerville Porchfest 2025

jefftk11 May 2025 2:00 UTC

15 points

1 comment2 min readLW link

(www.jefftk.com)

It’s Okay to Feel Bad for a Bit

moridinamael10 May 2025 23:24 UTC

149 points

34 comments3 min readLW link

G.D. as Capitalist Evolution, and the claim for humanity’s (temporary) upper hand

Martin Vlach10 May 2025 21:18 UTC

8 points

3 comments1 min readLW link

Book Review: “Encounters with Einstein” by Heisenberg

Baram Sosis10 May 2025 20:55 UTC

31 points

6 comments7 min readLW link

Where is the YIMBY movement for healthcare?

jasoncrawford10 May 2025 20:36 UTC

20 points

10 comments2 min readLW link

(newsletter.rootsofprogress.org)

Become a Superintelligence Yourself

Yaroslav Granowski10 May 2025 20:20 UTC

2 points

1 comment5 min readLW link

A Look Inside a Frequentist

Eggs10 May 2025 15:18 UTC

5 points

10 comments3 min readLW link

Open-source weaponry

samuelshadrach10 May 2025 13:11 UTC

3 points

0 comments3 min readLW link

(samuelshadrach.com)

Glass box learners want to be black box

Cole Wyeth10 May 2025 11:05 UTC

49 points

10 comments4 min readLW link

Takes and loose predictions on AI progress and some key problems

zef10 May 2025 10:11 UTC

5 points

0 comments5 min readLW link

(halcyoncyborg.substack.com)

Corbent – A Master Plan for Next‑Generation Direct Air Capture

Rudaiba10 May 2025 4:09 UTC

11 points

15 comments19 min readLW link

What if we just…didn’t build AGI? An Argument Against Inevitability

Nate Sharpe10 May 2025 3:37 UTC

9 points

7 comments14 min readLW link

(natezsharpe.substack.com)

Mind the Coherence Gap: Lessons from Steering Llama with Goodfire

eitan sprejer9 May 2025 21:29 UTC

4 points

1 comment6 min readLW link

My Experience With EMDR

Sable9 May 2025 21:25 UTC

22 points

0 comments11 min readLW link

(affablyevil.substack.com)

AI’s Hidden Game: Understanding Strategic Deception in AI and Why It Matters for Our Future

EmilyinAI9 May 2025 20:01 UTC

4 points

0 comments6 min readLW link

Muddling Through Some Thoughts on the Nature of Historiography

E.G. Blee-Goldman9 May 2025 19:04 UTC

2 points

0 comments4 min readLW link

A Guide to AI 2027

koenrane9 May 2025 17:14 UTC

0 points

1 comment28 min readLW link

Let’s stop making “Intelligence scale” graphs with humans and AI

Expertium9 May 2025 16:01 UTC

3 points

15 comments1 min readLW link

Slow corporations as an intuition pump for AI R&D automation

ryan_greenblatt and elifland

9 May 2025 14:49 UTC

91 points

25 comments9 min readLW link

Cheaters Gonna Cheat Cheat Cheat Cheat Cheat

Zvi9 May 2025 14:30 UTC

55 points

4 comments22 min readLW link

(thezvi.wordpress.com)

Humans vs LLM, memes as theorems

Yaroslav Granowski9 May 2025 13:26 UTC

1 point

0 comments1 min readLW link

Moving towards a question-based planning framework, instead of task lists

casualphysicsenjoyer9 May 2025 12:18 UTC

4 points

1 comment8 min readLW link

(substack.com)

Jim Babcock’s Mainline Doom Scenario: Human-Level AI Can’t Control Its Successor

Liron and jimrandomh

9 May 2025 5:20 UTC

30 points

4 comments62 min readLW link

(www.youtube.com)

Attend the 2025 Reproductive Frontiers Summit, June 10-12

TsviBT and Rachel Reid

9 May 2025 5:17 UTC

59 points

0 comments3 min readLW link

Interest In Conflict Is Instrumentally Convergent

Screwtape9 May 2025 2:16 UTC

66 points

58 comments10 min readLW link

Is ChatGPT actually fixed now?

sjadler8 May 2025 23:34 UTC

17 points

0 comments1 min readLW link

(stevenadler.substack.com)

Post EAG London AI x-Safety Co-working Retreat

plex8 May 2025 23:00 UTC

10 points

0 comments1 min readLW link

a brief critique of reduction

Vadim Golub8 May 2025 22:43 UTC

−17 points

4 comments2 min readLW link

Video & transcript: Challenges for Safe & Beneficial Brain-Like AGI

Steven Byrnes8 May 2025 21:11 UTC

27 points

0 comments18 min readLW link

Appendix: Interpretable by Design—Constraint Sets with Disjoint Limit Points

Ronak_Mehta8 May 2025 21:09 UTC

2 points

0 comments2 min readLW link

Interpretable by Design—Constraint Sets with Disjoint Limit Points

Ronak_Mehta8 May 2025 21:08 UTC

24 points

2 comments9 min readLW link

(ronakrm.github.io)

Is there a Half-Life for the Success Rates of AI Agents?

Matrice Jacobine8 May 2025 20:10 UTC

8 points

0 comments1 min readLW link

(www.tobyord.com)

Misalignment and Strategic Underperformance: An Analysis of Sandbagging and Exploration Hacking

Buck and Julian Stastny

8 May 2025 19:06 UTC

80 points

3 comments15 min readLW link

Behold the Pale Child (escaping Moloch’s Mad Maze)

rogersbacon8 May 2025 16:36 UTC

8 points

16 comments11 min readLW link

(www.secretorum.life)

An alignment safety case sketch based on debate

Marie_DB, Jacob Pfau, Benjamin Hilton and Geoffrey Irving

8 May 2025 15:02 UTC

59 points

21 comments25 min readLW link

(arxiv.org)

Mechanistic Interpretability Via Learning Differential Equations: AI Safety Camp Project Intermediate Report.

Valentin2026, ayoakin, Eduard Kovalets, tz3r0n4r, Soumyadeep Bose, Utkarsh Priyadarshi, Varun Piram and Axel Ahlqvist

8 May 2025 14:45 UTC

8 points

0 comments7 min readLW link

AI #115: The Evil Applications Division

Zvi8 May 2025 13:40 UTC

32 points

3 comments62 min readLW link

(thezvi.wordpress.com)

The Steganographic Potentials of Language Models

Artem Karpov, Tinuade and SCho

8 May 2025 11:23 UTC

9 points

0 comments1 min readLW link

Our bet on whether the AI market will crash

Remmelt and mabramov

8 May 2025 9:56 UTC

25 points

2 comments1 min readLW link

Concept-anchored representation engineering for alignment

Sandy Fraser8 May 2025 8:59 UTC

5 points

0 comments3 min readLW link

Orthogonality Thesis in layman’s terms.

Michael (@lethal_ai)8 May 2025 8:31 UTC

1 point

0 comments2 min readLW link