All 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 202120222023 2024 2025

All Jan Feb Mar Apr MayJunJul Aug Sep Oct Nov Dec

All 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 161718 19 20 21 22 23 24 25 26 27 28 29 30

A transparency and interpretability tech tree

evhubJun 16, 2022, 11:44 PM

163 points

11 comments18 min readLW link 1 review

BBC Future covers progress studies

jasoncrawfordJun 16, 2022, 10:44 PM

21 points

6 comments3 min readLW link

(rootsofprogress.org)

Humans are very reliable agents

alyssavanceJun 16, 2022, 10:02 PM

269 points

35 comments3 min readLW link

Towards Gears-Level Understanding of Agency

Thane RuthenisJun 16, 2022, 10:00 PM

25 points

4 comments18 min readLW link

A possible AI-inoculation due to early “robot uprising”

ShmiJun 16, 2022, 9:21 PM

16 points

2 comments1 min readLW link

AI Risk, as Seen on Snapchat

dkirmaniJun 16, 2022, 7:31 PM

23 points

8 comments1 min readLW link

[Link] “The madness of reduced medical diagnostics” by Dynomight

KennyJun 16, 2022, 7:20 PM

16 points

25 comments1 min readLW link

Breaking Down Goal-Directed Behaviour

Oliver SourbutJun 16, 2022, 6:45 PM

11 points

1 comment2 min readLW link

Perils of optimizing in social contexts

owencbJun 16, 2022, 5:40 PM

50 points

1 comment2 min readLW link

Don’t Over-Optimize Things

owencbJun 16, 2022, 4:33 PM

27 points

6 comments4 min readLW link

[Question] Security analysis of ‘cloud chemistry labs’?

KennyJun 16, 2022, 4:06 PM

6 points

2 comments1 min readLW link

Covid 6/16/22: Do Not Hand it to Them

ZviJun 16, 2022, 2:40 PM

29 points

5 comments7 min readLW link

(thezvi.wordpress.com)

[Question] Is there a worked example of Georgian taxes?

DagonJun 16, 2022, 2:07 PM

8 points

12 comments1 min readLW link

Against Active Shooter Drills

ZviJun 16, 2022, 1:40 PM

91 points

30 comments7 min readLW link

(thezvi.wordpress.com)

Ten experiments in modularity, which we’d like you to run!

CallumMcDougall, Lucius Bushnaq and Avery

Jun 16, 2022, 9:17 AM

62 points

3 comments9 min readLW link

[Question] What if LaMDA is indeed sentient / self-aware / worth having rights?

RomanSJun 16, 2022, 9:10 AM

22 points

13 comments1 min readLW link

Lifeguards

Orpheus16Jun 15, 2022, 11:03 PM

12 points

3 comments2 min readLW link

(forum.effectivealtruism.org)

Rationality Vienna Hike

Laszlo_TreszkaiJun 15, 2022, 10:11 PM

3 points

0 comments1 min readLW link

Contra Hofstadter on GPT-3 Nonsense

ricticJun 15, 2022, 9:53 PM

237 points

24 comments2 min readLW link

Progress links and tweets, 2022-06-13

jasoncrawfordJun 15, 2022, 7:47 PM

12 points

0 comments1 min readLW link

(rootsofprogress.org)

I applied for a MIRI job in 2020. Here’s what happened next.

ViktoriaMalyasovaJun 15, 2022, 7:37 PM

86 points

17 comments7 min readLW link

Contextual Evil

ACrackedPotJun 15, 2022, 7:32 PM

1 point

12 comments2 min readLW link

Multigate Priors

Adam JermynJun 15, 2022, 7:30 PM

4 points

0 comments3 min readLW link

FYI: I’m working on a book about the threat of AGI/ASI for a general audience. I hope it will be of value to the cause and the community

Darren McKeeJun 15, 2022, 6:08 PM

43 points

15 comments2 min readLW link

[Question] What are all the AI Alignment and AI Safety Communication Hubs?

Gunnar_ZarnckeJun 15, 2022, 4:16 PM

27 points

5 comments1 min readLW link

Georgism, in theory

Stuart_ArmstrongJun 15, 2022, 3:20 PM

40 points

22 comments4 min readLW link

Berlin AI Safety Open Meetup June 2022

pranomostroJun 15, 2022, 2:33 PM

12 points

0 comments1 min readLW link

A central AI alignment problem: capabilities generalization, and the sharp left turn

So8resJun 15, 2022, 1:10 PM

272 points

55 comments10 min readLW link 1 review

Our mental building blocks are more different than I thought

Marius HobbhahnJun 15, 2022, 11:07 AM

50 points

11 comments14 min readLW link

[Question] Has there been any work on attempting to use Pascal’s Mugging to make an AGI behave?

Chris_LeongJun 15, 2022, 8:33 AM

7 points

17 comments1 min readLW link

Alignment Risk Doesn’t Require Superintelligence

JustisMillsJun 15, 2022, 3:12 AM

35 points

4 comments2 min readLW link

A Butterfly’s View of Probability

Gabriel WuJun 15, 2022, 2:14 AM

29 points

17 comments11 min readLW link

[Question] Favourite new AI productivity tools?

Gabe MJun 15, 2022, 1:08 AM

14 points

5 comments1 min readLW link

Will vague “AI sentience” concerns do more for AI safety than anything else we might do?

Aryeh EnglanderJun 14, 2022, 11:53 PM

15 points

2 comments1 min readLW link

Yes, AI research will be substantially curtailed if a lab causes a major disaster

lcJun 14, 2022, 10:17 PM

103 points

31 comments2 min readLW link

Slow motion videos as AI risk intuition pumps

Andrew_CritchJun 14, 2022, 7:31 PM

241 points

41 comments2 min readLW link 1 review

Cryptographic Life: How to transcend in a sub-lightspeed world via Homomorphic encryption

GololJun 14, 2022, 7:22 PM

1 point

0 comments3 min readLW link

Blake Richards on Why he is Skeptical of Existential Risk from AI

Michaël TrazziJun 14, 2022, 7:09 PM

41 points

12 comments4 min readLW link

(theinsideview.ai)

[Question] How Do You Quantify [Physics Interfacing] Real World Capabilities?

DragonGodJun 14, 2022, 2:49 PM

17 points

1 comment4 min readLW link

Was the Industrial Revolution The Industrial Revolution?

Davis KedroskyJun 14, 2022, 2:48 PM

29 points

0 comments12 min readLW link

(daviskedrosky.substack.com)

Investigating causal understanding in LLMs

Marius Hobbhahn and Tom Lieberum

Jun 14, 2022, 1:57 PM

28 points

6 comments13 min readLW link

Why multi-agent safety is important

Akbir KhanJun 14, 2022, 9:23 AM

10 points

2 comments10 min readLW link

[Question] Was Eliezer Yudkowsky right to give himself 10% to succeed with HPMoR in 2010?

momom2Jun 14, 2022, 7:00 AM

2 points

2 comments1 min readLW link

Resources I send to AI researchers about AI safety

Vael GatesJun 14, 2022, 2:24 AM

69 points

12 comments1 min readLW link

Vael Gates: Risks from Advanced AI (June 2022)

Vael GatesJun 14, 2022, 12:54 AM

38 points

2 comments30 min readLW link

Cambridge LW Meetup: Personal Finance

Tony WangJun 14, 2022, 12:12 AM

3 points

0 comments1 min readLW link

OpenAI: GPT-based LLMs show ability to discriminate between its own wrong answers, but inability to explain how/why it makes that discrimination, even as model scales

Aditya JainJun 13, 2022, 11:33 PM

14 points

5 comments1 min readLW link

(openai.com)

[Question] Who said something like “The fact that putting 2 apples next to 2 other apples leads to there being 4 apples there has nothing to do with the fact that 2 + 2 = 4”?

hunterglenn13 Jun 2022 22:23 UTC

1 point

2 comments1 min readLW link

Continuity Assumptions

Jan_Kulveit13 Jun 2022 21:31 UTC

44 points

13 comments4 min readLW link

Crypto-fed Computation

aaguirre13 Jun 2022 21:20 UTC

24 points

7 comments7 min readLW link