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 Jul Aug Sep Oct Nov

All 1 234 5 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

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

snewmanMay 2, 2025, 11:52 PM

42 points

16 votes

Overall karma indicates overall quality.

2 comments13 min readLW link

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

agucovaMay 2, 2025, 11:45 PM

9 points

3 votes

Overall karma indicates overall quality.

0 comments2 min readLW 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

8 votes

Overall karma indicates overall quality.

3 comments1 min readLW link

Short story: Who is nancygonzalez8451097

Anders LindströmMay 2, 2025, 9:01 PM

13 points

7 votes

Overall karma indicates overall quality.

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

15 votes

Overall karma indicates overall quality.

6 comments8 min readLW link

Agents, Tools, and Simulators

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

May 2, 2025, 8:19 PM

14 points

9 votes

Overall karma indicates overall quality.

5 comments10 min readLW link

Obstacles in ARC’s agenda: Low Probability Estimation

David MatolcsiMay 2, 2025, 7:38 PM

44 points

15 votes

Overall karma indicates overall quality.

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

75 points

30 votes

Overall karma indicates overall quality.

8 comments8 min readLW link

When AI Optimizes for the Wrong Thing

Anthony FoxMay 2, 2025, 6:00 PM

5 points

3 votes

Overall karma indicates overall quality.

0 comments1 min readLW link

Alignment Structure Direction—Recursive Adversarial Oversight(RAO)

Jayden ShepardMay 2, 2025, 5:51 PM

2 points

4 votes

Overall karma indicates overall quality.

0 comments2 min readLW link

AI Welfare Risks

Adrià MoretMay 2, 2025, 5:49 PM

6 points

4 votes

Overall karma indicates overall quality.

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

0 points

6 votes

Overall karma indicates overall quality.

0 comments4 min readLW link

Supermen of the (Not so Far) Future

TerriLeafMay 2, 2025, 3:55 PM

9 points

9 votes

Overall karma indicates overall quality.

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

8 votes

Overall karma indicates overall quality.

0 comments7 min readLW link

AI Incident Monitoring: A Brief Analysis

Spencer AmesMay 2, 2025, 3:06 PM

3 points

2 votes

Overall karma indicates overall quality.

0 comments5 min readLW link

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

WriterMay 2, 2025, 2:19 PM

100 points

38 votes

Overall karma indicates overall quality.

18 comments9 min readLW link

OpenAI Preparedness Framework 2.0

ZviMay 2, 2025, 1:10 PM

61 points

25 votes

Overall karma indicates overall quality.

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

2 votes

Overall karma indicates overall quality.

6 comments2 min readLW link

(www.thefloatingdroid.com)

Roads are at maximum efficiency always

HrussMay 2, 2025, 10:29 AM

1 point

4 votes

Overall karma indicates overall quality.

3 comments1 min readLW link

The Continuum Fallacy and its Relatives

Zero ContradictionsMay 2, 2025, 2:58 AM

4 points

5 votes

Overall karma indicates overall quality.

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

1 vote

Overall karma indicates overall quality.

0 comments1 min readLW link

My Research Process: Understanding and Cultivating Research Taste

Neel NandaMay 1, 2025, 11:08 PM

27 points

10 votes

Overall karma indicates overall quality.

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

31 votes

Overall karma indicates overall quality.

7 comments8 min readLW link

(techgov.intelligence.org)

How to specify an alignment target

Richard JugginsMay 1, 2025, 9:11 PM

14 points

6 votes

Overall karma indicates overall quality.

2 comments12 min readLW link

Obstacles in ARC’s agenda: Mechanistic Anomaly Detection

David MatolcsiMay 1, 2025, 8:51 PM

42 points

14 votes

Overall karma indicates overall quality.

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

11 votes

Overall karma indicates overall quality.

1 comment1 min readLW link

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

Brittany GelbMay 1, 2025, 7:06 PM

48 points

11 votes

Overall karma indicates overall quality.

1 comment7 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

2 votes

Overall karma indicates overall quality.

0 comments2 min readLW link

Superhuman Coders in AI 2027 - Not So Fast

dschwarz and FutureSearch

May 1, 2025, 6:56 PM

67 points

27 votes

Overall karma indicates overall quality.

0 comments5 min readLW link

AI #114: Liars, Sycophants and Cheaters

ZviMay 1, 2025, 2:00 PM

40 points

15 votes

Overall karma indicates overall quality.

6 comments63 min readLW link

(thezvi.wordpress.com)

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

Vladimir_NesovMay 1, 2025, 1:54 PM

196 points

79 votes

Overall karma indicates overall quality.

25 comments5 min readLW link

Anthropomorphizing AI might be good, actually

Seth HerdMay 1, 2025, 1:50 PM

35 points

17 votes

Overall karma indicates overall quality.

6 comments3 min readLW link

Dont focus on updating P doom

AlgonMay 1, 2025, 11:10 AM

7 points

3 votes

Overall karma indicates overall quality.

3 comments2 min readLW link

Prioritizing Work

jefftkMay 1, 2025, 2:00 AM

108 points

63 votes

Overall karma indicates overall quality.

11 comments1 min readLW link

(www.jefftk.com)

Don’t rely on a “race to the top”

sjadlerMay 1, 2025, 12:33 AM

10 points

4 votes

Overall karma indicates overall quality.

0 comments1 min readLW link

Meta-Technicalities: Safeguarding Values in Formal Systems

LTMApr 30, 2025, 11:43 PM

2 points

1 vote

Overall karma indicates overall quality.

0 comments3 min readLW link

(routecause.substack.com)

Obstacles in ARC’s agenda: Finding explanations

David MatolcsiApr 30, 2025, 11:03 PM

123 points

37 votes

Overall karma indicates overall quality.

10 comments17 min readLW link

GPT-4o Responds to Negative Feedback

ZviApr 30, 2025, 8:20 PM

45 points

17 votes

Overall karma indicates overall quality.

2 comments18 min readLW link

(thezvi.wordpress.com)

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

Noosphere89Apr 30, 2025, 7:58 PM

7 points

1 vote

Overall karma indicates overall quality.

0 comments5 min readLW link

(www.interconnects.ai)

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

TFDApr 30, 2025, 7:38 PM

30 points

13 votes

Overall karma indicates overall quality.

15 comments2 min readLW link

(www.thefloatingdroid.com)

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

Vivek HebbarApr 30, 2025, 7:23 PM

52 points

15 votes

Overall karma indicates overall quality.

1 comment8 min readLW link

[Question] Can Narrowing One’s Reference Class Undermine the Doomsday Argument?

Iannoose n.Apr 30, 2025, 6:24 PM

2 points

2 votes

Overall karma indicates overall quality.

1 comment1 min readLW link

[Question] Does there exist an interactive reasoning map tool that lets users visually lay out claims, assign probabilities and confidence levels, and dynamically adjust their beliefs based on weighted influences between connected assertions?

Zack FriedmanApr 30, 2025, 6:22 PM

5 points

4 votes

Overall karma indicates overall quality.

4 comments1 min readLW link

Distilling the Internal Model Principle part II

JoseFaustinoApr 30, 2025, 5:56 PM

15 points

11 votes

Overall karma indicates overall quality.

0 comments19 min readLW link

Research Priorities for Hardware-Enabled Mechanisms (HEMs)

aogApr 30, 2025, 5:43 PM

17 points

3 votes

Overall karma indicates overall quality.

2 comments15 min readLW link

(www.longview.org)

Video and transcript of talk on automating alignment research

Joe CarlsmithApr 30, 2025, 5:43 PM

27 points

5 votes

Overall karma indicates overall quality.

0 comments24 min readLW link

(joecarlsmith.com)

Can we safely automate alignment research?

Joe CarlsmithApr 30, 2025, 5:37 PM

47 points

16 votes

Overall karma indicates overall quality.

29 comments48 min readLW link

(joecarlsmith.com)

Investigating task-specific prompts and sparse autoencoders for activation monitoring

Henk TillmanApr 30, 2025, 5:09 PM

23 points

7 votes

Overall karma indicates overall quality.

0 comments1 min readLW link

(arxiv.org)

European Links (30.04.25)

Martin SustrikApr 30, 2025, 3:40 PM

15 points

6 votes

Overall karma indicates overall quality.

1 comment8 min readLW link

(250bpm.substack.com)

Scaling Laws for Scalable Oversight

Subhash Kantamneni, Josh Engels, David Baek and Max Tegmark

Apr 30, 2025, 12:13 PM

37 points

17 votes

Overall karma indicates overall quality.

1 comment9 min readLW link

Keyboard shortcuts

Keys shown in yellow (e.g., ]) are accesskeys, and require a browser-specific modifier key (or keys).

Keys shown in grey (e.g., ?) do not require any modifier keys.

General
? Show keyboard shortcuts
Esc Hide keyboard shortcuts

Site navigation
h Go to Home (a.k.a. “Frontpage”) view
f Go to Featured (a.k.a. “Curated”) view
a Go to All (a.k.a. “Community”) view
m Go to Meta view
v Go to Tags view
c Go to Recent Comments view
r Go to Archive view
q Go to Sequences view
t Go to About page
u Go to User or Login page
o Go to Inbox page

Page navigation
, Jump up to top of page
. Jump down to bottom of page
/ Jump to top of comments section
s Search

Page actions
n New post or comment
e Edit current post

Post/comment list views
. Focus next entry in list
, Focus previous entry in list
; Cycle between links in focused entry
Enter Go to currently focused entry
Esc Unfocus currently focused entry
] Go to next page
[ Go to previous page
\ Go to first page
e Edit currently focused post

Editor
k Bold text
i Italic text
l Insert hyperlink
q Blockquote text

Appearance
= Increase text size
- Decrease text size
0 Reset to default text size
′ Cycle through content width settings
1 Switch to default theme [A]
2 Switch to dark theme [B]
3 Switch to grey theme [C]
4 Switch to ultramodern theme [D]
5 Switch to simple theme [E]
6 Switch to brutalist theme [F]
7 Switch to ReadTheSequences theme [G]
8 Switch to classic Less Wrong theme [H]
9 Switch to modern Less Wrong theme [I]
; Open theme tweaker
Enter Save changes and close theme tweaker
Esc Close theme tweaker (without saving)

Slide shows
l Start/resume slideshow
Esc Exit slideshow
→↓ Next slide
←↑ Previous slide
Space Reset slide zoom

Miscellaneous
x Switch to next view on user page
z Switch to previous view on user page
` Toggle compact comment list view
g Toggle anti-kibitzer