21 Jan 2025 21:32 UTC

135 points

15 comments2 min readLW link

(alignment.anthropic.com)

Veo-2 Can Produce Realistic Ads

Logan Riggs21 Jan 2025 19:13 UTC

14 points

0 comments1 min readLW link

Computational Limits on Efficiency

vibhumeh21 Jan 2025 18:29 UTC

8 points

1 comment5 min readLW link

Democratizing AI Governance: Balancing Expertise and Public Participation

Lucile Ter-Minassian21 Jan 2025 18:29 UTC

2 points

0 comments15 min readLW link

Hitler was not a monster

halgir21 Jan 2025 18:21 UTC

−12 points

5 comments1 min readLW link

Natural Intelligence is Overhyped

Collisteru21 Jan 2025 18:09 UTC

15 points

0 comments7 min readLW link

14+ AI Safety Advisors You Can Speak to – New AISafety.com Resource

Bryce Robertson and Søren Elverlin

21 Jan 2025 17:34 UTC

24 points

0 comments1 min readLW link

[Linkpost] Why AI Safety Camp struggles with fundraising (FBB #2)

gergogaspar21 Jan 2025 17:27 UTC

3 points

0 comments1 min readLW link

The Manhattan Trap: Why a Race to Artificial Superintelligence is Self-Defeating

Corin Katzke and GideonF

21 Jan 2025 16:57 UTC

92 points

11 comments2 min readLW link

(www.convergenceanalysis.org)

Links and short notes, 2025-01-20

jasoncrawford21 Jan 2025 16:10 UTC

8 points

0 comments1 min readLW link

(newsletter.rootsofprogress.org)

The Case Against AI Control Research

johnswentworth21 Jan 2025 16:03 UTC

433 points

85 comments6 min readLW link

Will AI Resilience protect Developing Nations?

edgecase6421 Jan 2025 15:31 UTC

4 points

0 comments8 min readLW link

Sleep, Diet, Exercise and GLP-1 Drugs

Zvi21 Jan 2025 12:20 UTC

41 points

6 comments18 min readLW link

(thezvi.wordpress.com)

We don’t want to post again “This might be the last AI Safety Camp”

Remmelt, Linda Linsefors and Robert Kralisch

21 Jan 2025 12:03 UTC

36 points

17 comments1 min readLW link

(manifund.org)

On Responsibility

silentbob21 Jan 2025 10:47 UTC

15 points

2 comments6 min readLW link

The ‘anti woke’ are positioned to win but can they capitalize?

Hzn21 Jan 2025 9:52 UTC

−8 points

0 comments2 min readLW link

Almost all growth is exponential growth

lemonhope21 Jan 2025 7:16 UTC

41 points

7 comments1 min readLW link

Arbitrage Drains Worse Markets to Feeds Better Ones

Cedar21 Jan 2025 3:44 UTC

25 points

1 comment1 min readLW link

On Contact, Part 1

james.lucassen21 Jan 2025 3:10 UTC

14 points

1 comment11 min readLW link

Retrospective: 12 [sic] Months Since MIRI

james.lucassen21 Jan 2025 2:52 UTC

68 points

0 comments9 min readLW link

Easily Evaluate SAE-Steered Models with EleutherAI Evaluation Harness

Matthew Khoriaty21 Jan 2025 2:02 UTC

8 points

0 comments3 min readLW link

Why We Need More Shovel-Ready AI Notkilleveryoneism Megaproject Proposals

Peter Berggren20 Jan 2025 22:38 UTC

36 points

1 comment6 min readLW link

Tips and Code for Empirical Research Workflows

John Hughes and Ethan Perez

20 Jan 2025 22:31 UTC

110 points

17 comments20 min readLW link

Lecture Series on Tiling Agents #2

abramdemski20 Jan 2025 21:02 UTC

16 points

0 comments1 min readLW link

Announcement: Learning Theory Online Course

Yegreg and Alex Flint

20 Jan 2025 19:55 UTC

63 points

33 comments4 min readLW link

The Hidden Status Game in Hospital Slacking

EpistemicExplorer20 Jan 2025 18:35 UTC

2 points

4 comments3 min readLW link

Monthly Roundup #26: January 2025

Zvi20 Jan 2025 15:30 UTC

34 points

15 comments43 min readLW link

(thezvi.wordpress.com)

Things I have been using LLMs for

Kaj_Sotala20 Jan 2025 14:20 UTC

51 points

13 comments7 min readLW link

(kajsotala.fi)

[Question] What are the chances that Superhuman Agents are already being tested on the internet?

artemium20 Jan 2025 11:09 UTC

3 points

1 comment1 min readLW link

Detroit Lions—over confidence is over rated?

Hzn20 Jan 2025 10:53 UTC

6 points

0 comments1 min readLW link

Logits, log-odds, and loss for parallel circuits

Dmitry Vaintrob20 Jan 2025 9:56 UTC

57 points

4 comments11 min readLW link

Worries about latent reasoning in LLMs

Caleb Biddulph20 Jan 2025 9:09 UTC

48 points

11 comments7 min readLW link

SIGMI Certification Criteria

a littoral wizard20 Jan 2025 2:41 UTC

6 points

0 comments1 min readLW link

AXRP Episode 38.5 - Adrià Garriga-Alonso on Detecting AI Scheming

DanielFilan20 Jan 2025 0:40 UTC

9 points

0 comments16 min readLW link

The Monster in Our Heads

testingthewaters19 Jan 2025 23:58 UTC

41 points

4 comments5 min readLW link

AI: How We Got Here—A Neuroscience Perspective

Mordechai Rorvig19 Jan 2025 23:51 UTC

5 points

0 comments2 min readLW link

(www.kickstarter.com)

Agent Foundations 2025 at CMU

Alexander Gietelink Oldenziel and windows

19 Jan 2025 23:48 UTC

90 points

10 comments1 min readLW link

Who is marketing AI alignment?

MrThink19 Jan 2025 21:37 UTC

23 points

4 comments1 min readLW link

Some lessons from the OpenAI-FrontierMath debacle

7vik19 Jan 2025 21:09 UTC

71 points

9 comments4 min readLW link

Maximally Eggy Crepes

jefftk19 Jan 2025 20:40 UTC

12 points

0 comments1 min readLW link

(www.jefftk.com)

The second bitter lesson — there’s a fundamental problem with aligning distributed AI

aelwood19 Jan 2025 19:00 UTC

−5 points

0 comments5 min readLW link

(pursuingreality.substack.com)

The Gentle Romance

Richard_Ngo19 Jan 2025 18:29 UTC

243 points

46 comments15 min readLW link

(www.asimov.press)

Is theory good or bad for AI safety?

Dmitry Vaintrob19 Jan 2025 10:32 UTC

29 points

1 comment5 min readLW link

[Question] What’s the Right Way to think about Information Theoretic quantities in Neural Networks?

Dalcy19 Jan 2025 8:04 UTC

45 points

13 comments3 min readLW link

Per Tribalismum ad Astra

Martin Sustrik19 Jan 2025 6:50 UTC

30 points

5 comments2 min readLW link

(250bpm.substack.com)

Five Recent AI Tutoring Studies

Arjun Panickssery19 Jan 2025 3:53 UTC

94 points

0 comments2 min readLW link

(arjunpanickssery.substack.com)

Does Society need a cultural outlet in turbulent political times?

Freya Mcneill19 Jan 2025 2:45 UTC

−3 points

0 comments7 min readLW link

On Thiel’s New American Regime

shawkisukkar19 Jan 2025 2:45 UTC

−3 points

0 comments5 min readLW link

(shawkisukkar.substack.com)

be the person that makes the meeting productive

Oldmanrahul18 Jan 2025 22:32 UTC

9 points

0 comments1 min readLW link

Beards and Masks?

jefftk18 Jan 2025 16:00 UTC

73 points

5 comments4 min readLW link

(www.jefftk.com)