Open Challenges in Representation Engineering

Jan Wehner and Daniel Tan

3 Apr 2025 19:21 UTC

12 points

0 comments5 min readLW link

Why Have Sentence Lengths Decreased?

Arjun Panickssery3 Apr 2025 17:50 UTC

73 points

15 comments4 min readLW link

I “invented” semimeasure theory and all I got was imprecise probability theory

Cole Wyeth3 Apr 2025 16:33 UTC

13 points

0 comments6 min readLW link

AI 2027: What Superintelligence Looks Like

Daniel Kokotajlo, Thomas Larsen, elifland, Scott Alexander, Jonas V and romeo

3 Apr 2025 16:23 UTC

272 points

21 comments41 min readLW link

(ai-2027.com)

Consider Breathwork

Jonathan Moregård3 Apr 2025 14:46 UTC

4 points

0 comments1 min readLW link

(honestliving.substack.com)

What are the differences between a singularity, an intelligence explosion, and a hard takeoff?

Vishakha and Algon

3 Apr 2025 10:37 UTC

5 points

0 comments2 min readLW link

(aisafety.info)

Emergent Misalignment and Emergent Alignment

Alvin Ånestrand3 Apr 2025 8:04 UTC

4 points

0 comments8 min readLW link

On Pseudo-Principality: Reclaiming “Whataboutism” as a Test for Counterfeit Principles

Alex_Steiner3 Apr 2025 1:37 UTC

12 points

4 comments7 min readLW link

Technical Claims

Vladimir_Nesov3 Apr 2025 0:30 UTC

14 points

0 comments1 min readLW link

The AI Adoption Gap: Preparing the US Government for Advanced AI

Lizka2 Apr 2025 23:46 UTC

12 points

2 comments17 min readLW link

(www.forethought.org)

Suggesting some revisions to Graham’s hierarchy of disagreement

Sniffnoy2 Apr 2025 22:25 UTC

8 points

1 comment2 min readLW link

The Care and Feeding of Mythological Intelligences

Jack2 Apr 2025 22:05 UTC

6 points

0 comments6 min readLW link

(elsworth.phd)

The Rise of Hyperpalatability

Jack2 Apr 2025 20:18 UTC

31 points

6 comments6 min readLW link

LLM Alignment Experiment: Effect of roles and optimism on probabilities

sjay82 Apr 2025 17:44 UTC

1 point

0 comments2 min readLW link

Why you can add moral value, and if an AI has moral weights for these moral values, those might be off

Wes R2 Apr 2025 17:43 UTC

−5 points

1 comment10 min readLW link

(docs.google.com)

How To Believe False Things

Eneasz2 Apr 2025 16:28 UTC

64 points

10 comments3 min readLW link

How prediction markets can create harmful outcomes: a case study

B Jacobs2 Apr 2025 15:37 UTC

28 points

2 comments5 min readLW link

(bobjacobs.substack.com)

Non-Monotonic Infra-Bayesian Physicalism

Marcus Ogren2 Apr 2025 12:14 UTC

16 points

0 comments18 min readLW link

Ghiblification is good, actually

Ozyrus2 Apr 2025 10:48 UTC

16 points

1 comment2 min readLW link

Show, not tell: GPT-4o is more opinionated in images than in text

Daniel Tan and eggsyntax

2 Apr 2025 8:51 UTC

72 points

17 comments3 min readLW link