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 AprMayJun

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 31

“Superhuman” Isn’t Well Specified

JustisMillsMay 3, 2025, 11:42 PM

32 points

9 comments3 min readLW link

(justismills.substack.com)

Navigating burnout

gwMay 3, 2025, 10:07 PM

73 points

1 comment9 min readLW link

(www.georgeyw.com)

What is your favorite podcast?

ChristianKlMay 3, 2025, 9:25 PM

32 points

9 comments1 min readLW link

[Question] Does translating a post with an LLM affect its rating?

ReverendBayesMay 3, 2025, 2:45 PM

9 points

9 comments2 min readLW link

SimpleStories: A Better Synthetic Dataset and Tiny Models for Interpretability

Lennart FinkeMay 3, 2025, 2:04 PM

13 points

0 comments1 min readLW link

What’s up with AI’s vision

Joachim BartosikMay 3, 2025, 1:23 PM

12 points

19 comments1 min readLW link

Sparsity is the enemy of feature extraction (ft. absorption)

7vik, chanind and Adrià Garriga-alonso

May 3, 2025, 10:13 AM

31 points

0 comments6 min readLW link

Exploring out-of-context reasoning (OOCR) fine-tuning in LLMs to increase test-phase awareness

Sanyu RajakumarMay 3, 2025, 3:33 AM

8 points

0 comments6 min readLW link

Prison Journal: Building Better Thinking Skills—Altruistic Person Saved > 100 Gorillas saved

P. JoãoMay 3, 2025, 1:34 AM

−30 points

2 comments1 min readLW link

Updates from Comments on “AI 2027 is a Bet Against Amdahl’s Law”

snewmanMay 2, 2025, 11:52 PM

40 points

2 comments13 min readLW link

Attend SPAR’s virtual demo day! (career fair + talks)

agucovaMay 2, 2025, 11:45 PM

9 points

0 comments LW link

(demoday.sparai.org)

Why does METR score o3 as effective for such a long time duration despite overall poor scores?

Cole WyethMay 2, 2025, 10:58 PM

19 points

3 comments1 min readLW link

Short story: Who is nancygonzalez8451097

Anders LindströmMay 2, 2025, 9:01 PM

13 points

2 comments5 min readLW link

Interim Research Report: Mechanisms of Awareness

Josh Engels, Neel Nanda and Senthooran Rajamanoharan

May 2, 2025, 8:29 PM

43 points

6 comments8 min readLW link

Agents, Tools, and Simulators

WillPetillo, Sean Herrington, Adebayo Mubarak, Cancus and Spencer Ames

May 2, 2025, 8:19 PM

12 points

0 comments10 min readLW link

Obstacles in ARC’s agenda: Low Probability Estimation

David MatolcsiMay 2, 2025, 7:38 PM

43 points

0 comments6 min readLW link

What’s going on with AI progress and trends? (As of 5/2025)

ryan_greenblattMay 2, 2025, 7:00 PM

71 points

7 comments8 min readLW link

When AI Optimizes for the Wrong Thing

Anthony FoxMay 2, 2025, 6:00 PM

5 points

0 comments1 min readLW link

Alignment Structure Direction—Recursive Adversarial Oversight(RAO)

Jayden ShepardMay 2, 2025, 5:51 PM

2 points

0 comments2 min readLW link

AI Welfare Risks

Adrià MoretMay 2, 2025, 5:49 PM

5 points

0 comments1 min readLW link

(philpapers.org)

Philosoplasticity: On the Inevitable Drift of Meaning in Recursive Self-Interpreting Systems

Maikol CoinMay 2, 2025, 5:46 PM

−1 points

0 comments4 min readLW link

Supermen of the (Not so Far) Future

TerriLeafMay 2, 2025, 3:55 PM

9 points

0 comments4 min readLW link

Steering Language Models in Multiple Directions Simultaneously

lukemarks, Narmeen and Amirali Abdullah

May 2, 2025, 3:27 PM

18 points

0 comments7 min readLW link

AI Incident Monitoring: A Brief Analysis

Spencer AmesMay 2, 2025, 3:06 PM

3 points

0 comments5 min readLW link

RA x ControlAI video: What if AI just keeps getting smarter?

WriterMay 2, 2025, 2:19 PM

100 points

17 comments9 min readLW link

OpenAI Preparedness Framework 2.0

ZviMay 2, 2025, 1:10 PM

60 points

1 comment23 min readLW link

(thezvi.wordpress.com)

Ex-OpenAI employee amici leave to file denied in Musk v OpenAI case?

TFDMay 2, 2025, 12:27 PM

4 points

6 comments2 min readLW link

(www.thefloatingdroid.com)

Roads are at maximum efficiency always

HrussMay 2, 2025, 10:29 AM

1 point

3 comments1 min readLW link

The Continuum Fallacy and its Relatives

Zero ContradictionsMay 2, 2025, 2:58 AM

4 points

2 comments4 min readLW link

(thewaywardaxolotl.blogspot.com)

Memory Decoding Journal Club: Motor learning selectively strengthens cortical and striatal synapses of motor engram neurons

Devin WardMay 1, 2025, 11:52 PM

1 point

0 comments1 min readLW link

My Research Process: Understanding and Cultivating Research Taste

Neel NandaMay 1, 2025, 11:08 PM

26 points

1 comment9 min readLW link

AI Governance to Avoid Extinction: The Strategic Landscape and Actionable Research Questions

peterbarnett and Aaron_Scher

May 1, 2025, 10:46 PM

105 points

7 comments8 min readLW link

(techgov.intelligence.org)

How to specify an alignment target

Richard JugginsMay 1, 2025, 9:11 PM

14 points

2 comments12 min readLW link

Obstacles in ARC’s agenda: Mechanistic Anomaly Detection

David MatolcsiMay 1, 2025, 8:51 PM

42 points

1 comment11 min readLW link

AI-Generated GitHub repo backdated with junk then filled with my systems work. Has anyone seen this before?

rguntherMay 1, 2025, 8:14 PM

7 points

1 comment1 min readLW link

What is Inadequate about Bayesianism for AI Alignment: Motivating Infra-Bayesianism

Brittany GelbMay 1, 2025, 7:06 PM

17 points

0 comments7 min readLW link

Can LLMs Simulate Internal Evaluation? A Case Study in Self-Generated Recommendations

The Neutral MindMay 1, 2025, 7:04 PM

4 points

0 comments2 min readLW link

Superhuman Coders in AI 2027 - Not So Fast

dschwarz and FutureSearch

May 1, 2025, 6:56 PM

62 points

0 comments5 min readLW link

AI #114: Liars, Sycophants and Cheaters

ZviMay 1, 2025, 2:00 PM

40 points

5 comments63 min readLW link

(thezvi.wordpress.com)

Slowdown After 2028: Compute, RLVR Uncertainty, MoE Data Wall

Vladimir_NesovMay 1, 2025, 1:54 PM

175 points

22 comments5 min readLW link

Anthropomorphizing AI might be good, actually

Seth HerdMay 1, 2025, 1:50 PM

35 points

6 comments3 min readLW link

Dont focus on updating P doom

AlgonMay 1, 2025, 11:10 AM

7 points

3 comments2 min readLW link

Prioritizing Work

jefftkMay 1, 2025, 2:00 AM

106 points

11 comments1 min readLW link

(www.jefftk.com)

Don’t rely on a “race to the top”

sjadlerMay 1, 2025, 12:33 AM

4 points

0 comments1 min readLW link

Meta-Technicalities: Safeguarding Values in Formal Systems

LTMApr 30, 2025, 11:43 PM

2 points

0 comments3 min readLW link

(routecause.substack.com)

Obstacles in ARC’s agenda: Finding explanations

David MatolcsiApr 30, 2025, 11:03 PM

122 points

10 comments17 min readLW link

GPT-4o Responds to Negative Feedback

Zvi30 Apr 2025 20:20 UTC

45 points

2 comments18 min readLW link

(thezvi.wordpress.com)

State of play of AI progress (and related brakes on an intelligence explosion) [Linkpost]

Noosphere8930 Apr 2025 19:58 UTC

7 points

0 comments5 min readLW link

(www.interconnects.ai)

Don’t accuse your interlocutor of being insufficiently truth-seeking

TFD30 Apr 2025 19:38 UTC

30 points

15 comments2 min readLW link

(www.thefloatingdroid.com)

How can we solve diffuse threats like research sabotage with AI control?

Vivek Hebbar30 Apr 2025 19:23 UTC

52 points

1 comment8 min readLW link